SlideShare a Scribd company logo
Xavier Amatriain – August 2014 – KDD
The Recommender
Problem
Revisited
Xavier Amatriain
Research/Engineering Director @
Netflix
@xamat
Bamshad Mobasher
Professor @
DePaul University
Xavier Amatriain – August 2014 – KDD
Index
PART I - Xavier Amatriain (2h)
1. The Recommender Problem
2. Traditional Recommendation Methods
3. Beyond Traditional Methods
PART 2 - Bamshad Mobasher (1h)
1. Context-aware Recommendations
Xavier Amatriain – August 2014 – KDD
Recent/Upcoming Publications
● The Recommender Problem Revisited. KDD and Recsys 2014 Tutorial
● KDD: Big & Personal: data and models behind Netflix recommendations. 2013
● SIGKDD Explorations: Mining large streams of user data for personalized recommendations. 2012
● Recsys: Building industrial-scale real-world recommender systems. 2012
● Recys - Walk the Talk: Analyzing the relation between implicit and explicit feedback for preference elicitation. 2011
● SIGIR – Temporal behavior of CF. 2010
● Web Intelligence – Expert-based CF for music. 2010
● Recsys – Tensor Factorization. 2010
● Mobile HCI – Tourist Recommendation. 2010
● Recsys Handbook (book) – Data mining for recsys. 2010 & Recommender Systems in Industry. 2014
● SIGIR – Wisdom of the Few. 2009
● Recsys – Denoising by re-rating. 2009
● CARS – Implicit context-aware recommendations. 2009
● UMAP – I like it I like it not. 2009
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. Traditional Recommendation Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References

Recommended for you

Interactive Recommender Systems with Netflix and Spotify
Interactive Recommender Systems with Netflix and SpotifyInteractive Recommender Systems with Netflix and Spotify
Interactive Recommender Systems with Netflix and Spotify

Interactive recommender systems enable the user to steer the received recommendations in the desired direction through explicit interaction with the system. In the larger ecosystem of recommender systems used on a website, it is positioned between a lean-back recommendation experience and an active search for a specific piece of content. Besides this aspect, we will discuss several parts that are especially important for interactive recommender systems, including the following: design of the user interface and its tight integration with the algorithm in the back-end; computational efficiency of the recommender algorithm; as well as choosing the right balance between exploiting the feedback from the user as to provide relevant recommendations, and enabling the user to explore the catalog and steer the recommendations in the desired direction. In particular, we will explore the field of interactive video and music recommendations and their application at Netflix and Spotify. We outline some of the user-experiences built, and discuss the approaches followed to tackle the various aspects of interactive recommendations. We present our insights from user studies and A/B tests. The tutorial targets researchers and practitioners in the field of recommender systems, and will give the participants a unique opportunity to learn about the various aspects of interactive recommender systems in the video and music domain. The tutorial assumes familiarity with the common methods of recommender systems.

collaborative filteringmachine learningnetflix
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems

The document discusses recommender systems and describes several techniques used in collaborative filtering recommender systems including k-nearest neighbors (kNN), singular value decomposition (SVD), and similarity weights optimization (SWO). It provides examples of how these techniques work and compares kNN to SWO. The document aims to explain state-of-the-art recommender system methods.

Recommender systems: Content-based and collaborative filtering
Recommender systems: Content-based and collaborative filteringRecommender systems: Content-based and collaborative filtering
Recommender systems: Content-based and collaborative filtering

This document provides an overview of recommender systems, including content-based and collaborative filtering approaches. It discusses how content-based systems make recommendations based on item profiles and calculating similarity between user and item profiles. Collaborative filtering is described as finding similar users and making predictions based on their ratings. The document also covers evaluation metrics, complexity issues, and tips for building recommender systems.

Xavier Amatriain – August 2014 – KDD
1. The Recommender
Problem
Xavier Amatriain – August 2014 – KDD
The Age of Search has come to an end
• ... long live the Age of Recommendation!
• Chris Anderson in “The Long Tail”
• “We are leaving the age of information and entering the age
of recommendation”
• CNN Money, “The race to create a 'smart' Google”:
• “The Web, they say, is leaving the era of search and
entering one of discovery. What's the difference? Search is
what you do when you're looking for something. Discovery
is when something wonderful that you didn't know existed,
or didn't know how to ask for, finds you.”
Xavier Amatriain – August 2014 – KDD
Information overload
“People read around 10 MB worth of material a day, hear 400 MB a
day, and see 1 MB of information every second” - The Economist, November 2006
In 2015, consumption will raise to 74 GB a day - UCSD Study 2014
Xavier Amatriain – August 2014 – KDD
Everything is personalized

Recommended for you

Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems

Part of the course "Algorithmic Methods of Data Science". Sapienza University of Rome, 2015. http://aris.me/index.php/data-mining-ds-2015

recommendationsdata sciencedata mining
Recent Trends in Personalization at Netflix
Recent Trends in Personalization at NetflixRecent Trends in Personalization at Netflix
Recent Trends in Personalization at Netflix

Presentation at the Netflix Expo session at RecSys 2020 virtual conference on 2020-09-24. It provides an overview of recommendation and personalization at Netflix and then highlights some of the things we’ve been working on as well as some important open research questions in the field of recommendations.

machine learningnetflixpersonalization
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...

Tutorial Deepak Agarwal and me gave at the 2016 ACM Recommender Systems conference. These slides only cover my part.

machine learningrecommender systemsquora
Xavier Amatriain – August 2014 – KDD
The “Recommender problem”
● Traditional definition: Estimate a utility
function that automatically predicts how a
user will like an item.
● Based on:
○ Past behavior
○ Relations to other users
○ Item similarity
○ Context
○ …
Xavier Amatriain – August 2014 – KDD
Recommendation as data mining
The core of the
Recommendation
Engine can be
assimilated to a general
data mining problem
(Amatriain et al. Data Mining Methods for
Recommender Systems in Recommender
Systems Handbook)
Xavier Amatriain – August 2014 – KDD
Machine Learning + all those other
things
● User Interface
● System requirements (efficiency, scalability,
privacy....)
● Serendipity
● Diversity
● Awareness
● Explanations
● …
Xavier Amatriain – August 2014 – KDD
Serendipity
● Unsought finding
● Don't recommend items the user already knows
or would have found anyway.
● Expand the user's taste into neighboring areas
by improving the obvious
● Collaborative filtering can offer controllable
serendipity (e.g. controlling how many
neighbors to use in the recommendation)

Recommended for you

Tutorial: Context In Recommender Systems
Tutorial: Context In Recommender SystemsTutorial: Context In Recommender Systems
Tutorial: Context In Recommender Systems

This document provides an overview of a tutorial on context-aware recommender systems. The tutorial will cover traditional recommendation techniques, context-aware recommendation which incorporates additional contextual information such as time and location, and context suggestion. It includes an agenda with topics, background information on recommender systems and evaluation metrics, and descriptions of techniques for context-aware recommendation including context filtering and modeling.

context suggestionmatrix factorizationcarskit
Recommendation engines
Recommendation enginesRecommendation engines
Recommendation engines

This document provides an overview of recommendation engines and systems. It describes different types of recommendation approaches, including collaborative filtering, content-based filtering, and hybrid methods. It also discusses how recommendation algorithms work and are implemented in Apache Mahout, a machine learning library for developing scalable recommendation applications. Key recommendation techniques like item-based filtering and user-based filtering are explained.

recommendation enginecollaborative filteringjava
Sequential Decision Making in Recommendations
Sequential Decision Making in RecommendationsSequential Decision Making in Recommendations
Sequential Decision Making in Recommendations

In this talk, we talk about the algorithms for sequential decision making in recommendations and the challenges around it.

contextual banditsreinforcement learning
Xavier Amatriain – August 2014 – KDD
Explanation/Support for Recommendations
Social Support
Xavier Amatriain – August 2014 – KDD
Personalization awareness
Diversity
DadAll SonDaughterDad&Mom MomAll Daughter MomAll?
Diversity & Awareness
Xavier Amatriain – August 2014 – KDD
What works
● Depends on the domain and particular problem
● However, in the general case it has been
demonstrated that the best isolated approach is CF.
○ Other approaches can be hybridized to improve
results in specific cases (cold-start problem...)
● What matters:
○ Data preprocessing: outlier removal, denoising,
removal of global effects (e.g. individual user's
average)
○ “Smart” dimensionality reduction using MF/SVD
○ Combining methods through ensembles
Xavier Amatriain – August 2014 – KDD
Evolution of the Recommender Problem
Rating Ranking Page Optimization
4.7
Context-aware
Recommendations
Context

Recommended for you

Recommendation System
Recommendation SystemRecommendation System
Recommendation System

This document discusses recommendation systems and provides examples of different types of recommendation approaches. It introduces collaborative filtering and content-based filtering as the main recommendation techniques. For collaborative filtering, it provides an example of item-based collaborative filtering using the R programming language on a Last.fm music dataset. Content-based filtering recommends items based on their properties and features. Hybrid systems combine collaborative and content-based filtering to generate recommendations.

big datarecommendationsamazon
Recent Trends in Personalization: A Netflix Perspective
Recent Trends in Personalization: A Netflix PerspectiveRecent Trends in Personalization: A Netflix Perspective
Recent Trends in Personalization: A Netflix Perspective

Invited talk at the Adaptive and Multi-Task Learning (AMTL) workshop at ICML 2019 on 2019-06-15 in Long Beach, CA.

personalizationrecommender systemsmachine learning
An introduction to Recommender Systems
An introduction to Recommender SystemsAn introduction to Recommender Systems
An introduction to Recommender Systems

Recommender systems are software agents that analyze a user's preferences through transactions and provide personalized recommendations accordingly. There are several recommendation paradigms including non-personalized rules, personalized rules based on user data, and transaction-based collaborative filtering that learns from user interactions. Context-based recommender systems also consider additional information like time, location, or device to provide adaptive recommendations. Common techniques used in recommender systems include content-based filtering that recommends similar items, collaborative filtering that finds users with similar tastes, and demographic-based recommendations.

tensor factorizationrecommender systemsknn
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References
Xavier Amatriain – August 2014 – KDD
2. Traditional Approaches
Xavier Amatriain – August 2014 – KDD
2.1. Collaborative Filtering
Xavier Amatriain – August 2014 – KDD
The CF Ingredients
● List of m Users and a list of n Items
● Each user has a list of items with associated opinion
○ Explicit opinion - a rating score
○ Sometime the rating is implicitly – purchase records
or listen to tracks
● Active user for whom the CF prediction task is
performed
● Metric for measuring similarity between users
● Method for selecting a subset of neighbors
● Method for predicting a rating for items not currently
rated by the active user.

Recommended for you

AbemaTVにおける推薦システム
AbemaTVにおける推薦システムAbemaTVにおける推薦システム
AbemaTVにおける推薦システム

SENDAI X-TECH Innovation Project 2018-2019 「AbemaTVにおける推薦システム」の発表スライドです。

engineering
Recommender systems using collaborative filtering
Recommender systems using collaborative filteringRecommender systems using collaborative filtering
Recommender systems using collaborative filtering

This document summarizes a student project on implementing recommender systems. The project objectives were to design a website using user-based, item-based, and model-based collaborative filtering as well as MapReduce to generate movie recommendations. The system was tested on the MovieLens dataset using MAE and RMSE metrics, with user-based filtering found to have the best performance. The document outlines the technical aspects of the recommendation system including the technologies used, website architecture, and references.

btechrecommendersystems
Personalized Page Generation for Browsing Recommendations
Personalized Page Generation for Browsing RecommendationsPersonalized Page Generation for Browsing Recommendations
Personalized Page Generation for Browsing Recommendations

Talk from First Workshop on Recommendation Systems for TV and Online Video at RecSys 2014 in Foster City, CA on 2014-10-10 about how we personalize the layout of the Netflix homepage to make it easier for people to browse the recommendations to quickly find something to watch and enjoy.

collaborative filteringnetflixrecommendations
Xavier Amatriain – August 2014 – KDD
Personalized vs Non-Personalised CF
● CF recommendations are personalized:
prediction based only on similar users
● Non-personalized collaborative-based
recommendation: averagge the
recommendations of ALL the users
● How would the two approaches compare?
Xavier Amatriain – August 2014 – KDD
Personalized vs. Not Personalized
● Netflix Prize: it is very
simple to produce
“reasonable”
recommendations and
extremely difficult to
improve them to become
“great”
● But there is a huge
difference in business
value between reasonable
and great
0,1510,2230,0222811718164974424EachMovie
0,1790,2330,041100020939526040MovieLens
0,1520,2200,725351944910048483Jester
MAE
Pers
MAE
Non
Pers
density
total
ratings
itemsusersData Set
Xavier Amatriain – August 2014 – KDD
User-based Collaborative Filtering
Xavier Amatriain – August 2014 – KDD
Collaborative Filtering
The basic steps:
1. Identify set of ratings for the target/active user
2. Identify set of users most similar to the target/active user
according to a similarity function (neighborhood
formation)
3. Identify the products these similar users liked
4. Generate a prediction - rating that would be given by the
target user to the product - for each one of these products
5. Based on this predicted rating recommend a set of top N
products

Recommended for you

分散表現を用いたリアルタイム学習型セッションベース推薦システム
分散表現を用いたリアルタイム学習型セッションベース推薦システム分散表現を用いたリアルタイム学習型セッションベース推薦システム
分散表現を用いたリアルタイム学習型セッションベース推薦システム

梛木 佑真, 岡本 一志: 分散表現を用いたリアルタイム学習型セッションベース推薦システム, 第35回人工知能学会全国大会, 1I2-GS-4a-05, 2021.6

information recommendationcollaborative filteringsession data
Past, present, and future of Recommender Systems: an industry perspective
Past, present, and future of Recommender Systems: an industry perspectivePast, present, and future of Recommender Systems: an industry perspective
Past, present, and future of Recommender Systems: an industry perspective

Keynote for the ACM Intelligent User Interface conference in 2016 in Sonoma, CA. I start with the past by talking about the Recommender Problem, and the Netflix Prize. Then I go into the Present and the Future by talking about approaches that go beyond rating prediction and ranking and by finishing with some of the most important lessons learned over the years. Throughout my talk I put special emphasis on the relation between algorithms and the User Interface.

recommender systemsuser experiencemachine learning
Sistema de recomendações de Filmes do Netflix
Sistema de recomendações de Filmes do NetflixSistema de recomendações de Filmes do Netflix
Sistema de recomendações de Filmes do Netflix

1. Introduction: What is a Recommender System 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 3. Novel Methods 3.1. Learning to Rank 3.2. Context-aware Recommendations 3.2.1. Tensor Factorization 3.2.2. Factorization Machines 3.3. Deep Learning 3.4. Similarity 3.5. Social Recommendations 4. Hybrid Approaches 5. A practical example: Netflix 6. Conclusions 7. References

deep learningnovel methodssocial recommendations
Xavier Amatriain – August 2014 – KDD
UB Collaborative Filtering
● A collection of user ui
, i=1, …n and a collection
of products pj
, j=1, …, m
● An n × m matrix of ratings vij
, with vij
= ? if user
i did not rate product j
● Prediction for user i and product j is computed
as
• Similarity can be computed by Pearson correlation
or
or
Xavier Amatriain – August 2014 – KDD
User-based CF
Example
Xavier Amatriain – August 2014 – KDD
User-based CF
Example
Xavier Amatriain – August 2014 – KDD
User-based CF
Example

Recommended for you

Recommandation systems -
Recommandation systems - Recommandation systems -
Recommandation systems -

This document discusses recommender systems and approaches used at Netflix. It covers collaborative filtering using user-user and item-item methods, content-based recommendations using item attributes, and hybrid approaches. It provides examples of how Netflix uses collaborative filtering to generate personalized genre rows and social recommendations. Netflix combines many data sources and machine learning models to power its highly personalized recommendation engine.

recommended systems
Mini-training: Personalization & Recommendation Demystified
Mini-training: Personalization & Recommendation DemystifiedMini-training: Personalization & Recommendation Demystified
Mini-training: Personalization & Recommendation Demystified

Recommendations are everywhere : music, movies, books, social medias, e-commerce web sites… The Web is leaving the era of search and entering one of discovery. This quick introduction will help you to understand this vast topic and why you should use it.

personalizationrecommendercollaborative filtering
Intro to R and Data Mining 2012 09 27
Intro to R and Data Mining 2012 09 27Intro to R and Data Mining 2012 09 27
Intro to R and Data Mining 2012 09 27

R is a programming language and software environment for statistical analysis and graphics. It is used for data manipulation, statistics, and graphics. R allows users to create functions (like spells for wizards) and relies on functions developed by statistical researchers. While initially developed in the 1990s, R has grown significantly with over 800 add-on packages. Data mining involves exploring large datasets to discover patterns and make predictions. Common techniques in R include classification, clustering, association rule mining, and decision trees.

Xavier Amatriain – August 2014 – KDD
User-based CF
Example
Xavier Amatriain – August 2014 – KDD
User-based CF
Example
Xavier Amatriain – August 2014 – KDD
Item-based Collaborative Filtering
Xavier Amatriain – August 2014 – KDD
Item Based CF Algorithm
● Look into the items the target user has rated
● Compute how similar they are to the target item
○ Similarity only using past ratings from other
users!
● Select k most similar items.
● Compute Prediction by taking weighted average
on the target user’s ratings on the most similar
items.

Recommended for you

BeaconsAI engr 245 lean launchpad stanford 2019
BeaconsAI engr 245 lean launchpad stanford 2019BeaconsAI engr 245 lean launchpad stanford 2019
BeaconsAI engr 245 lean launchpad stanford 2019

Over the course of 9 weeks, the team went through multiple pivots to find product-market fit: 1. They started with facial recognition for smart door locks but learned their solution did not fully address parent concerns. 2. Their second pivot was to enterprise access control but customers said tailgating was a bigger problem. 3. Recognizing tailgating risks required surveillance, they pivoted again to surveillance software. 4. After failing to find problem-team fit, they did a complete restart and developed an AI assistant for streamlining similar work tasks. Customer feedback was positive and they identified a potential market of 500k ML professionals.

business modelcustomer developmente245
Modern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in MendeleyModern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in Mendeley

This document discusses modern perspectives on recommender systems and their applications at Mendeley. It covers the evolution of recommender problems from rating prediction to context-aware recommendations. It also discusses common recommender algorithms like collaborative filtering, content-based filtering and hybrid approaches. The document concludes by discussing how Mendeley uses recommenders for related research, researchers to follow, and other use cases.

mendeleyrecommendation systemsrecsys2014
Олександр Обєдніков “Рекомендательные системы”
Олександр Обєдніков “Рекомендательные системы”Олександр Обєдніков “Рекомендательные системы”
Олександр Обєдніков “Рекомендательные системы”

This document provides an overview of recommender systems. It begins with definitions and examples of recommender systems and their business value. It then discusses the problem formulation and history, including the Netflix Prize competition. Traditional collaborative filtering and latent factor models are explained. The document also covers content-based recommendations and novel approaches like learning to rank, sequence recommendation using deep learning, and social/trust-based systems. It concludes with a discussion of hybrid recommendation approaches.

dakiry
Xavier Amatriain – August 2014 – KDD
Item Similarity Computation
● Similarity: find users who have rated items and
apply a similarity function to their ratings.
● Cosine-based Similarity (difference in rating scale
between users is not taken into account)
● Adjusted Cosine Similarity (takes care of difference
in rating scale)
Xavier Amatriain – August 2014 – KDD
Challenges of memory-based CF
Xavier Amatriain – August 2014 – KDD
CF: Pros/Cons
● Requires minimal knowledge engineering efforts
● Users and products are symbols without any internal
structure or characteristics
● Produces good-enough results in most cases
Challenges:
● Sparsity – evaluation of large itemsets where user/item
interactions are under 1%.
● Scalability - Nearest neighbor require computation that
grows with both the number of users and the number of
items.
Xavier Amatriain – August 2014 – KDD
The Sparsity Problem
● Typically: large product sets, user
ratings for a small percentage of them
(e.g. Amazon: millions of books and a user may
have bought hundreds of books)
● If you represent the Netflix Prize
rating data in a User/Movie matrix you
get…
○ 500,000 x 17,000 = 8.5 B positions
○ Out of which only 100M are non-zero
● Number of users needs to be ~ 0.1 x
size of the catalog

Recommended for you

Fashiondatasc
FashiondatascFashiondatasc
Fashiondatasc

1. Fashion companies are leveraging data science and personalization to improve customers' shopping experience. Computer vision and deep learning allow them to use visual data from photos. 2. Recommendation systems combine traditional signals like ratings and clicks with visual signals from photos to improve accuracy. 3. Rigorous experimentation is important to test recommendation systems. Techniques like A/B testing and multi-armed bandits help optimize the personalized experience for each customer.

data sciencerecommender systemfashion
Lecture Notes on Recommender System Introduction
Lecture Notes on Recommender System IntroductionLecture Notes on Recommender System Introduction
Lecture Notes on Recommender System Introduction

This document provides an overview of recommender systems and the techniques used to build them. It discusses collaborative filtering, content-based filtering, knowledge-based recommendations, and hybrid approaches. For collaborative filtering, it describes user-based and item-based approaches, including measuring similarity, making predictions, and generating recommendations. It also discusses evaluation techniques and advanced topics like explanations.

recommendation system techunique and issue
recommendation system techunique and issuerecommendation system techunique and issue
recommendation system techunique and issue

This document discusses recommendation system techniques and issues. It covers common recommendation approaches like content-based filtering, collaborative filtering, and hybrid systems. It also addresses challenges like cold start problems, privacy issues, and data sparsity. Recommendation systems analyze user preferences to suggest new items, and are used by applications like ecommerce sites, streaming services, and social networks to provide personalized recommendations. While useful, they also present technical challenges for researchers.

data science
Xavier Amatriain – August 2014 – KDD
Model-based
Collaborative Filtering
Xavier Amatriain – August 2014 – KDD
Model Based CF Algorithms
● Memory based
○ Use the entire user-item database to generate a
prediction.
○ Usage of statistical techniques to find the neighbors – e.g.
nearest-neighbor.
● Model-based
○ First develop a model of user
○ Type of model:
■ Probabilistic (e.g. Bayesian Network)
■ Clustering
■ Rule-based approaches (e.g. Association Rules)
■ Classification
■ Regression
■ LDA
■ ...
Xavier Amatriain – August 2014 – KDD
Model-based CF:
What we learned from the
Netflix Prize
Xavier Amatriain – August 2014 – KDD
What we were interested in:
■ High quality recommendations
Proxy question:
■ Accuracy in predicted rating
■ Improve by 10% = $1million!

Recommended for you

Lean Product Development 101
Lean Product Development 101Lean Product Development 101
Lean Product Development 101

Mark Geene, CEO/Co-founder of Cloud Elements, presented "Lean Product Development" at Fort Collins Startup Week 2014. Check out the presentation for information on how to build a Lean startup. Based on principles from 'Lean Startup' by Eric Ries, 'Running Lean' by Ash Maurya and '500 Startups' by Dave McClure.

running leanlean startupagile
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems

Video: http://vimeo.com/16537278 An introduction to recommender systems. Given at Talis (www.talis.com) Oct. 28 2010

recsystalisrecommender systems
Natural Intelligence the human factor in AI
Natural Intelligence the human factor in AINatural Intelligence the human factor in AI
Natural Intelligence the human factor in AI

Presented at AI NEXTCon Seattle 1/17-20, 2018 http://aisea18.xnextcon.com join our free online AI group with 50,000+ tech engineers to learn and practice AI technology, including: latest AI news, tech articles/blogs, tech talks, tutorial videos, and hands-on workshop/codelabs, on machine learning, deep learning, data science, etc..

Xavier Amatriain – August 2014 – KDD
2007 Progress Prize
▪ Top 2 algorithms
▪ SVD - Prize RMSE: 0.8914
▪ RBM - Prize RMSE: 0.8990
▪ Linear blend Prize RMSE: 0.88
▪ Currently in use as part of Netflix’ rating prediction
component
▪ Limitations
▪ Designed for 100M ratings, we have 5B ratings
▪ Not adaptable as users add ratings
▪ Performance issues
Xavier Amatriain – August 2014 – KDD
SVD/MF
● X: m x n matrix (e.g., m users, n videos)
● U: m x r matrix (m users, r factors)
● S: r x r diagonal matrix (strength of each ‘factor’) (r: rank of the
matrix)
● V: r x n matrix (n videos, r factor)
Xavier Amatriain – August 2014 – KDD
Simon Funk’s SVD
● One of the most
interesting findings
during the Netflix
Prize came out of
a blog post
● Incremental,
iterative, and
approximate way
to compute the
SVD using
gradient descent
Xavier Amatriain – August 2014 – KDD
▪ User factor vectors and item-factors vector
▪ Baseline (bias) (user & item deviation from average)
▪ Predict rating as
▪ Asymetric SVD (Koren et. Al) asymmetric variation w. implicit
feedback
▪ Where
▪ are three item factor vectors
▪ Users are not parametrized, but rather represented by:
▪ R(u): items rated by user u
▪ N(u): items for which the user has given implicit preference (e.g. rated vs. not rated)
SVD for Rating Prediction

Recommended for you

Building a UX Research Program
Building a UX Research ProgramBuilding a UX Research Program
Building a UX Research Program

A presentation that ties together business strategy and UX design strategy into a program for actionable results from user research.

ux designux strategyuser research
Overview of recommender system
Overview of recommender systemOverview of recommender system
Overview of recommender system

The document provides an overview of recommender systems. It discusses the typical architecture of recommender systems and describes three main types: collaborative filtering systems, content-based systems, and knowledge-based systems. It also covers paradigms like collaborative filtering, content-based, knowledge-based, and hybrid recommender systems. The document then focuses on collaborative filtering techniques like user-based nearest neighbor collaborative filtering and item-based collaborative filtering. It also discusses latent factor models, matrix factorization approaches, and context-based recommender systems.

filteringrecommendercollaboration
Ariadne's Thread -- Exploring a world of networked information built from fre...
Ariadne's Thread -- Exploring a world of networked information built from fre...Ariadne's Thread -- Exploring a world of networked information built from fre...
Ariadne's Thread -- Exploring a world of networked information built from fre...

Most of the current interfaces to digital libraries are built on keyword-based search and list-based presentation. For users who do not have specific items to search for but would rather explore not-yet-familiar topics, it is not easy to figure out to what extend and on which aspects the returned records match the query. Users have to try different combinations of keywords to narrow down or broaden the search space in the hope of getting useful results in the end. In this talk, we will present a web interface that provides users an opportunity to interactively and visually explore the context of queries. In this interface, after entering a query, a contextual view about the query is visualised, where the most related journals, authors, subject headings, publishers, topical terms, etc. are positioned in 2D based on their relatedness to the query and among each other. By clicking any of these nodes, a new visualisation about the selected one is presented. With this click-through style, the users could get visual contexts about their selected entities (journal, author, topical terms, etc.) and shift their interests by choosing interested (types of) entities to investigate further. At any stop, a search in WorldCat.org with the currently focused entity (a topical word, a author or a journal) will return the most matched results (judged by the standard WorldCat search engine). We implemented this interface over WorldCat, the world largest bibliographic database. To guarantee the responsiveness of this interactive interface, we adopt a two-step approach: an off-line preparation phase with an on-line process. Off-line, we build the semantic representation of each entity where Random Projection is used to vigorously reduce dimensionality (from 6 million to 600). In the on-line interface terms from a query are compared to entities in the reduced semantic matrix where reciprocal relatedness is used to select genuine matches. The number of hits is further reduced to render a network layout easy to overview and navigate. In the end, we can investigate the relations between roughly 6 million topical terms, 5 million authors, 1 million subject headings 1000 Dewey decimal codes and 1.7 million publishers.

random projectionvisualisationoclc research
Xavier Amatriain – August 2014 – KDD
Restricted Boltzmann Machines
● Each unit is a state that can be active or not active
● Each input to a unit is associated to a weight
● The transfer function calculates a score for every unit
based on the weighted sum of inputs
● Score is passed to the activation function that calculates
the probability of the unit to be active
● Restrict the connectivity to make learning easier.
Only one layer of hidden units.
No connections between hidden units.
Hidden units are independent given visible states
Xavier Amatriain – August 2014 – KDD
RBM for Recommendations
● Each visible unit = an item
● Num. of hidden units a is parameter
● In training phase, for each user:
○ If user rated item, vi
is activated
○ Activation states of vi
= inputs to hj
○ Based on activation, hj
is computed
○ Activation state of hj
becomes input to vi
○ Activation state of vi
is recalculated
○ Difference between current and past
activation state for vi
used to update weights
wij
and thresholds
● In prediction phase:
○ For the items of the user the vi
are activated
○ Based on this the state of the hj
is computed
○ The activation of hj
is used as input to
recompute the state of vi
○ Activation probabilities are used to
recommend items
Xavier Amatriain – August 2014 – KDD
Putting all together
● Remember that current production
model includes an ensemble of both
SVD++ and RBMs
Xavier Amatriain – August 2014 – KDD
What about the final prize
ensembles?
● Our offline studies showed they were too
computationally intensive to scale
● Expected improvement not worth the engineering
effort
● Plus…. Focus had already shifted to other issues
that had more impact than rating prediction.

Recommended for you

Modern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in MendeleyModern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in Mendeley

Presentation given for one of Pearson's Data Research teams. It motivates the use of recommender systems, describes common approaches to building and evaluating them and gives examples of how they are used in Mendeley. Joint work with Kris Jack, Chief Data Scientist at Mendeley.

recommender systems
Recommender lecture
Recommender lectureRecommender lecture
Recommender lecture

Recommender systems allow online retailers to customize their sites to meet consumer tastes by aiding browsing and suggesting related items. Personalization is one of e-commerce's advantages over brick-and-mortar stores. Common techniques include item-to-item recommendations based on user ratings, user-to-user comparisons based on item preferences, and population-based suggestions of popular items. Challenges include obtaining user data, making novel recommendations, and addressing ethical issues.

Exploratory Analysis of User Data
Exploratory Analysis of User DataExploratory Analysis of User Data
Exploratory Analysis of User Data

Here is a Python code that automatically finds all research papers (and their authors) about a given set of keywords, where keywords is an input parameter: ```python import requests from bs4 import BeautifulSoup keywords = ["user data"] def find_papers(keywords): url = "https://dblp.org/search/publ/api?q=" + "+".join(keywords) r = requests.get(url) data = r.json() papers = data["result"]["hits"]["hit"] for paper in papers: title = paper["info"]["title"] authors = paper["info"]["author"] print(f"Title: {title}") print(

Xavier Amatriain – August 2014 – KDD
Clustering
Xavier Amatriain – August 2014 – KDD
Clustering
● Goal: cluster users and compute per-cluster
“typical” preferences
● Users receive recommendations computed at
the cluster level
Xavier Amatriain – August 2014 – KDD
Locality-sensitive Hashing (LSH)
● Method for grouping similar items in highly
dimensional spaces
● Find a hashing function s.t. similar items are
grouped in the same buckets
● Main application is Nearest-neighbors
○ Hashing function is found iteratively by
concatenating random hashing functions
○ Addresses one of NN main concerns:
performance
Xavier Amatriain – August 2014 – KDD
Other “interesting” clustering
techniques
● k-means and all its variations
● Affinity Propagation
● Spectral Clustering
● Non-parametric Bayesian Clustering (e.g.
HDPs)

Recommended for you

Endorphin. Making sense of social data
Endorphin. Making sense of social dataEndorphin. Making sense of social data
Endorphin. Making sense of social data

Endorphin is a social data analytics company that provides analytical summaries and ratings of individuals based on their digital footprints. It mines social network data from platforms like LinkedIn and Facebook to generate executive summaries, credibility ratings, and rankings of influence for customers, partners, and employees. This helps businesses evaluate potential partners and customers, and helps individuals learn more about their own connections and find experts in their fields. The document outlines Endorphin's business models, target markets, competitors, product roadmap, and investment opportunities as it works to launch its minimum viable product and scaling efforts.

startuppitchendorphin
Data/AI driven product development: from video streaming to telehealth
Data/AI driven product development: from video streaming to telehealthData/AI driven product development: from video streaming to telehealth
Data/AI driven product development: from video streaming to telehealth

Healthcare is different from any other application domain, or is it not? While it is true that there are specific aspects, such as high stakes decisions and a complex regulatory framework, that make healthcare somewhat different, it is also the case that many of the lessons learned from building data-driven products in other domains translate remarcably well into healthcare. This is particularly so because healthcare is also a user facing domain, where users can be both patients or healthcare professionals. Given that data has shown to improve user experience while ensuring quality and scalability, few would argue that healthcare cannot benefit from being much more data-driven than it has traditionally been. In this talk, I described how this experience building impactful data and AI solutions into user facing products for decades can be leveraged to revolutionize telehealth. At Curai, we combine approaches such as state-of-the-art large language models with expert systems in areas such as NLP, vision, and automated diagnosis to augment and scale doctors, and to improve user experience and healthcare outcomes. We will see some of those applications while analyzing the role of data and ML algorithms in making them possible.

recommender systemsmachine learninghealthtech
AI-driven product innovation: from Recommender Systems to COVID-19
AI-driven product innovation: from Recommender Systems to COVID-19AI-driven product innovation: from Recommender Systems to COVID-19
AI-driven product innovation: from Recommender Systems to COVID-19

AI/Machine Learning has become an integral part of many household tech products, from Netflix to our phones. In this talk I will draw from my experience driving AI teams at some of those companies to showcase how AI can positively impact products as different as Netflix and Curai, an online telehealth service.

machine learningdata scienceai
Xavier Amatriain – August 2014 – KDD
Association Rules
Xavier Amatriain – August 2014 – KDD
Association rules
● Past purchases are interpreted as transactions of “associated”
items
● If a visitor has some interest in Book 5, she will be
recommended to buy Book 3 as well
● Recommendations are constrained to some minimum levels
of confidence
● Fast to implement and execute (e.g. A Priori algorithm)
Xavier Amatriain – August 2014 – KDD
Classifiers
Xavier Amatriain – August 2014 – KDD
Classifiers
● Classifiers are general computational models trained
using positive and negative examples
● They may take in inputs:
○ Vector of item features (action / adventure, Bruce
Willis)
○ Preferences of customers (like action / adventure)
○ Relations among item
● E.g. Logistic Regression, Bayesian Networks,
Support Vector Machines, Decision Trees, etc...

Recommended for you

AI for COVID-19 - Q42020 update
AI for COVID-19 - Q42020 updateAI for COVID-19 - Q42020 update
AI for COVID-19 - Q42020 update

With half of the world’s population lacking access to healthcare services, and 30% of the adult population in the US having inadequate health insurance coverage to get even basic access to services, it should have been clear that a pandemic like COVID-19 would strain the global healthcare system way over its maximum capacity. In this context, many are trying to embrace and encourage the use of telehealth as a way to provide safe and convenient access to care. However, telehealth in itself can not scale to cover all our needs unless we improve scalability and efficiency through AI and automation. In this talk, we will describe how our work on combining latest AI advances with medical experts and online access has the potential to change the landscape in healthcare access and provide 24/7 quality healthcare. Combining areas such as NLP, vision, and automatic diagnosis we can augment and scale doctors. We will describe our work on combining expert systems with deep learning to build state-of-the-art medical diagnostic models that are also able to model the unknowns. We will also show our work on using language models for medical Q&A . More importantly, we will describe how those approaches have been used to address the urgent and immediate needs of the current pandemic.

aihealthcaremachine learning
AI for COVID-19: An online virtual care approach
AI for COVID-19: An online virtual care approachAI for COVID-19: An online virtual care approach
AI for COVID-19: An online virtual care approach

Slides for the talk I gave at the AI and COVID-19 virtual conference at Stanford. Video here: https://hai.stanford.edu/events/covid-19-and-ai-virtual-conference/video-archive

aitelemedicinehealthtech
Lessons learned from building practical deep learning systems
Lessons learned from building practical deep learning systemsLessons learned from building practical deep learning systems
Lessons learned from building practical deep learning systems

1. There are many lessons to be learned from building practical deep learning systems, including choosing the right evaluation metrics, being thoughtful about your data and potential biases, and understanding dependencies between data, models, and systems. 2. It is important to optimize only what matters and beware of biases in your data. Simple models are often better than complex ones, and feature engineering is crucial. 3. Both supervised and unsupervised learning are important, and ensembles often perform best. Your AI infrastructure needs to support both experimentation and production.

machine learningdeep learningbig data
Xavier Amatriain – August 2014 – KDD
Classifiers
● Classifiers can be used in CF and CB
Recommenders
● Pros:
○ Versatile
○ Can be combined with other methods to improve accuracy
of recommendations
● Cons:
○ Need a relevant training set
○ May overfit (Regularization)
● E.g. Logistic Regression, Bayesian Networks,
Support Vector Machines, Decision Trees, etc...
Xavier Amatriain – August 2014 – KDD
Limitations of
Collaborative Filtering
Xavier Amatriain – August 2014 – KDD
Limitations of Collaborative Filtering
● Cold Start: There needs to be enough other users
already in the system to find a match. New items
need to get enough ratings.
● Popularity Bias: Hard to recommend items to
someone with unique tastes.
○ Tends to recommend popular items (items from
the tail do not get so much data)
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Novel Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References

Recommended for you

AI for healthcare: Scaling Access and Quality of Care for Everyone
AI for healthcare: Scaling Access and Quality of Care for EveryoneAI for healthcare: Scaling Access and Quality of Care for Everyone
AI for healthcare: Scaling Access and Quality of Care for Everyone

Talk given with Anitha Kannan at MLConf 2019. Link to abstract here: https://mlconf.com/speakers/anitha-kannan/

healthcareaimachine learning
Towards online universal quality healthcare through AI
Towards online universal quality healthcare through AITowards online universal quality healthcare through AI
Towards online universal quality healthcare through AI

This document discusses the potential for using AI and automation to improve online universal healthcare. It notes that physicians currently do not have enough time to properly diagnose and treat patients given the large amount of information involved. The document proposes that AI, through techniques like knowledge extraction from medical literature and patient data, conversational systems, automated diagnosis and treatment recommendations, and processing of multimodal inputs, could help scale and improve healthcare access and quality by assisting physicians online. The goal would be to create an online healthcare system as good as top doctors that is universally accessible at low cost.

aihealthcaremachine learning
From one to zero: Going smaller as a growth strategy
From one to zero: Going smaller as a growth strategyFrom one to zero: Going smaller as a growth strategy
From one to zero: Going smaller as a growth strategy

This talk was designed for Engineering managers. Having been at companies of all sizes, I recommend managers who want to grow to go smaller. At the same time I reflect on what are the important things that remain constant regardless the size and context and which ones don't.

technologymanagementleadership
Xavier Amatriain – August 2014 – KDD
2.2 Content-based Recommenders
Xavier Amatriain – August 2014 – KDD
Content-Based Recommendation
● Recommendations based on content of items rather than on
other users’ opinions/interactions
● Goal: recommend items similar to those the user liked
● Common for recommending text-based products (web
pages, usenet news messages, )
● Items to recommend are “described” by their associated
features (e.g. keywords)
● User Model structured in a “similar” way as the content:
features/keywords more likely to occur in the preferred
documents (lazy approach)
● The user model can be a classifier based on whatever
technique (Neural Networks, Naïve Bayes...)
Xavier Amatriain – August 2014 – KDD
Pros/cons of CB Approach
Pros
● No need for data on other users: No cold-start or sparsity
● Able to recommend to users with unique tastes.
● Able to recommend new and unpopular items
● Can provide explanations by listing content-features
Cons
● Requires content that can be encoded as meaningful features
(difficult in some domains/catalogs)
● Users represented as learnable function of content features.
● Difficult to implement serendipity
● Easy to overfit (e.g. for a user with few data points)
Xavier Amatriain – August 2014 – KDD
A word of caution

Recommended for you

Learning to speak medicine
Learning to speak medicineLearning to speak medicine
Learning to speak medicine

The document discusses medical conversations and the potential for AI-enabled conversational systems to improve healthcare. It covers: 1) How doctors currently conduct medical diagnosis through a question-and-answer process to understand patients' symptoms and medical histories. 2) Challenges doctors face in making accurate diagnoses given time constraints and the growing complexity of medical data and knowledge. 3) Opportunities for AI to enhance healthcare through personalized recommendations, multimodal data integration, and proactive diagnosis via conversational systems and sensors. 4) Recent advances in natural language processing, machine learning and medical language representations that enable building such conversational medical assistants.

healthcarenlpmachine learning
ML to cure the world
ML to cure the worldML to cure the world
ML to cure the world

This document discusses how machine learning and artificial intelligence can be used to improve medical diagnosis and care. It summarizes that doctors currently have limited time to gather patient information, diagnose issues, and recommend treatments. ML and AI systems could help by automating processes, leveraging large datasets, and providing more personalized care and improved user experiences for both patients and doctors. The document outlines how such a system could integrate medical decision support, knowledge bases, natural language processing, multimodal input analysis, and personalization based on individual patient profiles and histories to help diagnose patients and inform care recommendations. Challenges to developing such systems include issues around data quality, user experience design, and ensuring recommendations are trustworthy and engage users effectively.

machine learninghealthcare
Recommender Systems In Industry
Recommender Systems In IndustryRecommender Systems In Industry
Recommender Systems In Industry

This document outlines key concepts in recommendation systems. It begins by defining the traditional recommender problem as predicting user ratings for items based on past behavior and relationships. It then discusses lessons learned from the Netflix Prize competition, including the effectiveness of singular value decomposition and the limitations of models designed only for rating prediction. The document outlines approaches beyond rating prediction, including ranking, similarity, social recommendations, and explore/exploit tradeoffs. It discusses optimizing recommendation pages and using higher-order models like tensor factorization. In summary, it provides an overview of traditional and modern approaches in recommendation systems.

recommender systemsmachine learning
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References
Xavier Amatriain – August 2014 – KDD
2.3 Hybrid Approaches
Xavier Amatriain – August 2014 – KDD
Comparison of methods (FAB
system)
• Content–based
recommendation with
Bayesian classifier
• Collaborative is
standard using
Pearson correlation
• Collaboration via
content uses the
content-based user
profiles
Averaged on 44 users
Precision computed in top 3 recommendations
Xavier Amatriain – August 2014 – KDD
Hybridization Methods
Hybridization Method Description
Weighted Outputs from several techniques (in the form of
scores or votes) are combined with different
degrees of importance to offer final
recommendations
Switching Depending on situation, the system changes from
one technique to another
Mixed Recommendations from several techniques are
presented at the same time
Feature combination Features from different recommendation sources
are combined as input to a single technique
Cascade The output from one technique is used as input of
another that refines the result
Feature augmentation The output from one technique is used as input
features to another
Meta-level The model learned by one recommender is used
as input to another

Recommended for you

Medical advice as a Recommender System
Medical advice as a Recommender SystemMedical advice as a Recommender System
Medical advice as a Recommender System

The document discusses medical practice as a recommender system. It outlines how medical decisions can be improved with recommender systems by making better use of data through algorithms and machine learning to provide more personalized recommendations. Current medical decision support systems are discussed, including knowledge-based approaches built from medical literature and data from electronic health records and ontologies. Machine learning techniques can be used to build diagnostic systems from data. The company Curai is working on combining AI/ML with good UX to build a medical tool for patients, leveraging techniques discussed in the document. Challenges include algorithmic approaches, data quality and bias, trustworthy UX, and legal issues. Medicine is an area that can greatly benefit from recommender system approaches

recommender systemsmedicinemachine learning
Past present and future of Recommender Systems: an Industry Perspective
Past present and future of Recommender Systems: an Industry PerspectivePast present and future of Recommender Systems: an Industry Perspective
Past present and future of Recommender Systems: an Industry Perspective

The document summarizes the past, present, and future of recommender systems from an industry perspective. [1] In the past, Netflix popularized recommender systems with their 2006 Netflix Prize competition. [2] Currently, recommender systems are used widely across many applications and industries. They have evolved to use implicit feedback and contextual information beyond just explicit ratings. Ranking items is also central to recommender systems. [3] Future directions include addressing indirect feedback challenges, incorporating the value or reward of recommendations, optimizing full pages rather than just individual recommendations, and personalizing not just what is recommended but how it is recommended to users.

netflixmachine learningbig data
Staying Shallow & Lean in a Deep Learning World
Staying Shallow & Lean in a Deep Learning WorldStaying Shallow & Lean in a Deep Learning World
Staying Shallow & Lean in a Deep Learning World

This document discusses how while deep learning has achieved success in areas like image recognition and natural language processing, it is not always the best or most accurate approach and should not be obsessively pursued to the exclusion of other machine learning techniques. Specifically, simpler models may perform equally well due to Occam's razor. Unsupervised learning and feature engineering are also important. Ensembles of different models can further improve results compared to relying on a single approach. The document cautions against an overemphasis on deep learning without considering factors like system complexity, costs, and the ability to distribute models.

machine learningdata mining
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References
Xavier Amatriain – August 2014 – KDD
3. Beyond traditional approaches to
Recommendation
Xavier Amatriain – August 2014 – KDD
3.1 Ranking
Xavier Amatriain – August 2014 – KDD
Ranking Key algorithm, sorts titles in most contexts
Ranking

Recommended for you

Machine Learning for Q&A Sites: The Quora Example
Machine Learning for Q&A Sites: The Quora ExampleMachine Learning for Q&A Sites: The Quora Example
Machine Learning for Q&A Sites: The Quora Example

Machine learning is used extensively at Q&A sites like Quora to improve user experience. Some applications include answer ranking to determine the best answers, feed ranking to present the most interesting stories, asking users to answer questions, recommending new topics and users, detecting duplicate questions, and moderating content. Quora uses a variety of machine learning models and does extensive experimentation and A/B testing to optimize different metrics.

quoramachine learningrecommender systems
BIG2016- Lessons Learned from building real-life user-focused Big Data systems
BIG2016- Lessons Learned from building real-life user-focused Big Data systemsBIG2016- Lessons Learned from building real-life user-focused Big Data systems
BIG2016- Lessons Learned from building real-life user-focused Big Data systems

1) More data is not always better than better models. Sometimes, better modeling techniques are needed rather than just collecting more data. 2) Ensembles of different models generally perform better than any single model and are commonly used in practice. Feature engineering to create new inputs for ensembles can improve their effectiveness. 3) Implicit signals from user behavior usually provide more useful information than explicit feedback, but both should be used to best represent users' long-term goals.

www2016big datamachine learning
Strata 2016 - Lessons Learned from building real-life Machine Learning Systems
Strata 2016 -  Lessons Learned from building real-life Machine Learning SystemsStrata 2016 -  Lessons Learned from building real-life Machine Learning Systems
Strata 2016 - Lessons Learned from building real-life Machine Learning Systems

Talk about practical machine learning lessons learned over the years. Given at the Hardcore Data Science track in Strata San Jose 2016

machine learningquorabig data
Xavier Amatriain – August 2014 – KDD
Ranking
● Most recommendations are presented in a sorted
list
● Recommendation can be understood as a ranking
problem
● Popularity is the obvious baseline
● Ratings prediction is a clear secondary data input
that allows for personalization
● Many other features can be added
Xavier Amatriain – August 2014 – KDD
Ranking by ratings
4.7 4.6 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5
Niche titles
High average ratings… by those who would watch it
Xavier Amatriain – August 2014 – KDD
RMSE
Popularity
PredictedRating
1
2
3
4
5
Linear Model:
frank
(u,v) = w1
p(v) + w2
r(u,v) + b
FinalRanking
Example: Two features, linear model

Recommended for you

Barcelona ML Meetup - Lessons Learned
Barcelona ML Meetup - Lessons LearnedBarcelona ML Meetup - Lessons Learned
Barcelona ML Meetup - Lessons Learned

The document discusses several lessons learned about machine learning including: 1) Sometimes more data is not needed and better models can be achieved through feature engineering and appropriate hyperparameters. 2) Models will learn what they are taught through the training data, so training data needs to be thoughtful and free of biases. 3) A combination of supervised and unsupervised learning along with ensemble methods often provides the best results rather than relying on any single approach.

systemsmachine learningrecommender systems
10 more lessons learned from building Machine Learning systems - MLConf
10 more lessons learned from building Machine Learning systems - MLConf10 more lessons learned from building Machine Learning systems - MLConf
10 more lessons learned from building Machine Learning systems - MLConf

1. Machine learning applications at Quora include answer ranking, feed ranking, topic recommendations, user recommendations, and more. A variety of models are used including logistic regression, gradient boosted decision trees, neural networks, and matrix factorization. 2. Implicit signals like watching and clicking tend to be more useful than explicit signals like ratings. However, both implicit and explicit signals combined can better represent long-term goals. 3. It is important to focus on feature engineering to create features that are reusable, transformable, interpretable, and reliable. The outputs of models may become inputs to other models, so care must be taken to avoid feedback loops and ensure proper data dependencies.

machine learningquorarecommender systems
10 more lessons learned from building Machine Learning systems
10 more lessons learned from building Machine Learning systems10 more lessons learned from building Machine Learning systems
10 more lessons learned from building Machine Learning systems

1. Machine learning applications at Quora include answer ranking, feed ranking, topic recommendations, user recommendations, and more. A variety of models are used including logistic regression, gradient boosted decision trees, neural networks, and matrix factorization. 2. Implicit signals like watching and clicking tend to be more useful than explicit signals like ratings. However, both implicit and explicit signals combined can better represent long-term goals. 3. The outputs of machine learning models will often become inputs to other models, so models need to be designed with this in mind to avoid issues like feedback loops.

machine learningbig datarecommender systems
Popularity
1
2
3
4
5
FinalRanking
PredictedRating
Example: Two features, linear model
Xavier Amatriain – August 2014 – KDD
Ranking
Ranking
Xavier Amatriain – August 2014 – KDD
Learning to rank
● Machine learning problem: goal is to construct
ranking model from training data
● Training data can be a partial order or binary
judgments (relevant/not relevant).
● Resulting order of the items typically induced
from a numerical score
● Learning to rank is a key element for
personalization
● You can treat the problem as a standard
supervised classification problem
Xavier Amatriain – August 2014 – KDD
Learning to rank - Metrics
● Quality of ranking measured using metrics as
○ Normalized Discounted Cumulative Gain
○ Mean Reciprocal Rank (MRR)
○ Fraction of Concordant Pairs (FCP)
○ Others…
● But, it is hard to optimize machine-learned models
directly on these measures (e.g. non-differentiable)
● Recent research on models that directly optimize
ranking measures

Recommended for you

What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024

This is a powerpoint that features Microsoft Teams Devices and everything that is new including updates to its software and devices for May 2024

microsoft teamsmicrosoft
Coordinate Systems in FME 101 - Webinar Slides
Coordinate Systems in FME 101 - Webinar SlidesCoordinate Systems in FME 101 - Webinar Slides
Coordinate Systems in FME 101 - Webinar Slides

If you’ve ever had to analyze a map or GPS data, chances are you’ve encountered and even worked with coordinate systems. As historical data continually updates through GPS, understanding coordinate systems is increasingly crucial. However, not everyone knows why they exist or how to effectively use them for data-driven insights. During this webinar, you’ll learn exactly what coordinate systems are and how you can use FME to maintain and transform your data’s coordinate systems in an easy-to-digest way, accurately representing the geographical space that it exists within. During this webinar, you will have the chance to: - Enhance Your Understanding: Gain a clear overview of what coordinate systems are and their value - Learn Practical Applications: Why we need datams and projections, plus units between coordinate systems - Maximize with FME: Understand how FME handles coordinate systems, including a brief summary of the 3 main reprojectors - Custom Coordinate Systems: Learn how to work with FME and coordinate systems beyond what is natively supported - Look Ahead: Gain insights into where FME is headed with coordinate systems in the future Don’t miss the opportunity to improve the value you receive from your coordinate system data, ultimately allowing you to streamline your data analysis and maximize your time. See you there!

Quality Patents: Patents That Stand the Test of Time
Quality Patents: Patents That Stand the Test of TimeQuality Patents: Patents That Stand the Test of Time
Quality Patents: Patents That Stand the Test of Time

Is your patent a vanity piece of paper for your office wall? Or is it a reliable, defendable, assertable, property right? The difference is often quality. Is your patent simply a transactional cost and a large pile of legal bills for your startup? Or is it a leverageable asset worthy of attracting precious investment dollars, worth its cost in multiples of valuation? The difference is often quality. Is your patent application only good enough to get through the examination process? Or has it been crafted to stand the tests of time and varied audiences if you later need to assert that document against an infringer, find yourself litigating with it in an Article 3 Court at the hands of a judge and jury, God forbid, end up having to defend its validity at the PTAB, or even needing to use it to block pirated imports at the International Trade Commission? The difference is often quality. Quality will be our focus for a good chunk of the remainder of this season. What goes into a quality patent, and where possible, how do you get it without breaking the bank? ** Episode Overview ** In this first episode of our quality series, Kristen Hansen and the panel discuss: ⦿ What do we mean when we say patent quality? ⦿ Why is patent quality important? ⦿ How to balance quality and budget ⦿ The importance of searching, continuations, and draftsperson domain expertise ⦿ Very practical tips, tricks, examples, and Kristen’s Musts for drafting quality applications https://www.aurorapatents.com/patently-strategic-podcast.html

patentspatent applicationpatent prosecution
Xavier Amatriain – August 2014 – KDD
Learning to rank - Approaches
1. Pointwise
■ Ranking function minimizes loss function defined on
individual relevance judgment
■ Ranking score based on regression or classification
■ Ordinal regression, Logistic regression, SVM, GBDT, …
2. Pairwise
■ Loss function is defined on pair-wise preferences
■ Goal: minimize number of inversions in ranking
■ Ranking problem is then transformed into the binary
classification problem
■ RankSVM, RankBoost, RankNet, FRank…
Xavier Amatriain – August 2014 – KDD
Learning to rank - Approaches
3. Listwise
■ Indirect Loss Function
− RankCosine: similarity between ranking list and ground truth
as loss function
− ListNet: KL-divergence as loss function by defining a
probability distribution
− Problem: optimization of listwise loss function may not optimize
IR metrics
■ Directly optimizing IR metric (difficult since they are
not differentiable)
− Genetic Programming or Simulated Annealing
− LambdaMart weights pairwise errors in RankNet by IR metric
− Gradient descent on smoothed version of objective function (e.
g. CLiMF or TFMAP)
− SVM-MAP relaxes MAP metric by adding to SVM constraints
− AdaRank uses boosting to optimize NDCG
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References
Xavier Amatriain – August 2014 – KDD
3.2 Similarity as Recommendation

Recommended for you

Cookies program to display the information though cookie creation
Cookies program to display the information though cookie creationCookies program to display the information though cookie creation
Cookies program to display the information though cookie creation

Java Servlet programs

Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024

This presentation explores the practical application of image description techniques. Familiar guidelines will be demonstrated in practice, and descriptions will be developed “live”! If you have learned a lot about the theory of image description techniques but want to feel more confident putting them into practice, this is the presentation for you. There will be useful, actionable information for everyone, whether you are working with authors, colleagues, alone, or leveraging AI as a collaborator. Link to presentation recording and transcript: https://bnctechforum.ca/sessions/details-of-description-part-ii-describing-images-in-practice/ Presented by BookNet Canada on June 25, 2024, with support from the Department of Canadian Heritage.

a11yaccessibilityalt text
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdfINDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf

These fighter aircraft have uses outside of traditional combat situations. They are essential in defending India's territorial integrity, averting dangers, and delivering aid to those in need during natural calamities. Additionally, the IAF improves its interoperability and fortifies international military alliances by working together and conducting joint exercises with other air forces.

air force fighter planebiggest submarinezambia port
Xavier Amatriain – August 2014 – KDD
▪ Displayed in many
different contexts
▪ In response to user
actions/context
(search, queue
add…)
▪ More like… rows
Similars
Xavier Amatriain – August 2014 – KDD
Graph-based similarities
0.8
0.2
0.3
0.
4
0.3
0.7
0.
3
Xavier Amatriain – August 2014 – KDD
Example of graph-based similarity: SimRank
▪ SimRank (Jeh & Widom, 02): “two objects are
similar if they are referenced by similar
objects.”
Xavier Amatriain – August 2014 – KDD
Similarity ensembles
• Similarity can refer to different dimensions
• Similar in metadata/tags
• Similar in user play behavior
• Similar in user rating behavior
• …
• Combine them using an ensemble
• Weights are learned using regression over existing
response
• Or… some MAB explore/exploit approach
• The final concept of “similarity” responds to what users vote
as similar

Recommended for you

UiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs ConferenceUiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs Conference

We are honored to launch and host this event for our UiPath Polish Community, with the help of our partners - Proservartner! We certainly hope we have managed to spike your interest in the subjects to be presented and the incredible networking opportunities at hand, too! Check out our proposed agenda below 👇👇 08:30 ☕ Welcome coffee (30') 09:00 Opening note/ Intro to UiPath Community (10') Cristina Vidu, Global Manager, Marketing Community @UiPath Dawid Kot, Digital Transformation Lead @Proservartner 09:10 Cloud migration - Proservartner & DOVISTA case study (30') Marcin Drozdowski, Automation CoE Manager @DOVISTA Pawel Kamiński, RPA developer @DOVISTA Mikolaj Zielinski, UiPath MVP, Senior Solutions Engineer @Proservartner 09:40 From bottlenecks to breakthroughs: Citizen Development in action (25') Pawel Poplawski, Director, Improvement and Automation @McCormick & Company Michał Cieślak, Senior Manager, Automation Programs @McCormick & Company 10:05 Next-level bots: API integration in UiPath Studio (30') Mikolaj Zielinski, UiPath MVP, Senior Solutions Engineer @Proservartner 10:35 ☕ Coffee Break (15') 10:50 Document Understanding with my RPA Companion (45') Ewa Gruszka, Enterprise Sales Specialist, AI & ML @UiPath 11:35 Power up your Robots: GenAI and GPT in REFramework (45') Krzysztof Karaszewski, Global RPA Product Manager 12:20 🍕 Lunch Break (1hr) 13:20 From Concept to Quality: UiPath Test Suite for AI-powered Knowledge Bots (30') Kamil Miśko, UiPath MVP, Senior RPA Developer @Zurich Insurance 13:50 Communications Mining - focus on AI capabilities (30') Thomasz Wierzbicki, Business Analyst @Office Samurai 14:20 Polish MVP panel: Insights on MVP award achievements and career profiling

#uipathcommunity#automation#automationdeveloper
Quantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLMQuantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLM

Quantum Communications Q&A with Gemini LLM. These are based on Shannon's Noisy channel Theorem and offers how the classical theory applies to the quantum world.

quantum communicationsshannon's channel theoremclassical theory
Best Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdfBest Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdf

As a popular open-source library for analytics engineering, dbt is often used in combination with Airflow. Orchestrating and executing dbt models as DAGs ensures an additional layer of control over tasks, observability, and provides a reliable, scalable environment to run dbt models. This webinar will cover a step-by-step guide to Cosmos, an open source package from Astronomer that helps you easily run your dbt Core projects as Airflow DAGs and Task Groups, all with just a few lines of code. We’ll walk through: - Standard ways of running dbt (and when to utilize other methods) - How Cosmos can be used to run and visualize your dbt projects in Airflow - Common challenges and how to address them, including performance, dependency conflicts, and more - How running dbt projects in Airflow helps with cost optimization Webinar given on 9 July 2024

apache airflowdbtdbt-core
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References
Xavier Amatriain – August 2014 – KDD
3.3 Deep Learning for
Recommendation
Xavier Amatriain – August 2014 – KDD
Deep Learning for Collaborative
Filtering
● Let’s look at how Spotify uses Recurrent Networks
for Playlist Prediction (http://erikbern.com/?p=589)
Xavier Amatriain – August 2014 – KDD
Deep Learning for Collaborative
Filtering
● In order to predict the next track or movie a user is going to
watch, we need to define a distribution
○ If we choose Softmax as it is common practice, we get:
● Problem: denominator (over all examples is very
expensive to compute)
● Solution: build a tree that implements a hierarchical
softmax
● More details on the blogpost

Recommended for you

Implementations of Fused Deposition Modeling in real world
Implementations of Fused Deposition Modeling  in real worldImplementations of Fused Deposition Modeling  in real world
Implementations of Fused Deposition Modeling in real world

The presentation showcases the diverse real-world applications of Fused Deposition Modeling (FDM) across multiple industries: 1. **Manufacturing**: FDM is utilized in manufacturing for rapid prototyping, creating custom tools and fixtures, and producing functional end-use parts. Companies leverage its cost-effectiveness and flexibility to streamline production processes. 2. **Medical**: In the medical field, FDM is used to create patient-specific anatomical models, surgical guides, and prosthetics. Its ability to produce precise and biocompatible parts supports advancements in personalized healthcare solutions. 3. **Education**: FDM plays a crucial role in education by enabling students to learn about design and engineering through hands-on 3D printing projects. It promotes innovation and practical skill development in STEM disciplines. 4. **Science**: Researchers use FDM to prototype equipment for scientific experiments, build custom laboratory tools, and create models for visualization and testing purposes. It facilitates rapid iteration and customization in scientific endeavors. 5. **Automotive**: Automotive manufacturers employ FDM for prototyping vehicle components, tooling for assembly lines, and customized parts. It speeds up the design validation process and enhances efficiency in automotive engineering. 6. **Consumer Electronics**: FDM is utilized in consumer electronics for designing and prototyping product enclosures, casings, and internal components. It enables rapid iteration and customization to meet evolving consumer demands. 7. **Robotics**: Robotics engineers leverage FDM to prototype robot parts, create lightweight and durable components, and customize robot designs for specific applications. It supports innovation and optimization in robotic systems. 8. **Aerospace**: In aerospace, FDM is used to manufacture lightweight parts, complex geometries, and prototypes of aircraft components. It contributes to cost reduction, faster production cycles, and weight savings in aerospace engineering. 9. **Architecture**: Architects utilize FDM for creating detailed architectural models, prototypes of building components, and intricate designs. It aids in visualizing concepts, testing structural integrity, and communicating design ideas effectively. Each industry example demonstrates how FDM enhances innovation, accelerates product development, and addresses specific challenges through advanced manufacturing capabilities.

fdmffffused deposition modeling
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdfBT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf

Presented at Gartner Data & Analytics, London Maty 2024. BT Group has used the Neo4j Graph Database to enable impressive digital transformation programs over the last 6 years. By re-imagining their operational support systems to adopt self-serve and data lead principles they have substantially reduced the number of applications and complexity of their operations. The result has been a substantial reduction in risk and costs while improving time to value, innovation, and process automation. Join this session to hear their story, the lessons they learned along the way and how their future innovation plans include the exploration of uses of EKG + Generative AI.

neo4jneo4j webinarsgraph database
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...

Have you noticed the OpenSSF Scorecard badges on the official Dart and Flutter repos? It's Google's way of showing that they care about security. Practices such as pinning dependencies, branch protection, required reviews, continuous integration tests etc. are measured to provide a score and accompanying badge. You can do the same for your projects, and this presentation will show you how, with an emphasis on the unique challenges that come up when working with Dart and Flutter. The session will provide a walkthrough of the steps involved in securing a first repository, and then what it takes to repeat that process across an organization with multiple repos. It will also look at the ongoing maintenance involved once scorecards have been implemented, and how aspects of that maintenance can be better automated to minimize toil.

dartflutteropenssf
Xavier Amatriain – August 2014 – KDD
Deep Learning for Content-based
Recommendations
● Another application of Deep Learning to recommendations also
from Spotify
○ http://benanne.github.io/2014/08/05/spotify-cnns.html also Deep content-based
music recommendation, Aäron van den Oord, Sander Dieleman and Benjamin
Schrauwen, NIPS 2013
● Application to coldstart new titles when very little CF information
is available
● Using mel-spectrograms from the audio signal as input
● Training the deep neural network to predict 40 latent factors
coming from Spotify’s CF solution
Xavier Amatriain – August 2014 – KDD
Deep Learning for Content-based
Recommendations
● Network architecture made of 4 convolutional layers + 4 fully
connected dense layers
● One dimensional convolutional layers using RELUs (Rectified Linear Units) with
activation max(0,x)
● Max-pooling operations between convolutional layers to downsample intermediate
representations in time, and add time invariance
● Global temporal pooling layer after last convolutional layer: pools across entire time axis,
computing statistics of the learned features across time:: mean, maximum and L2-norm
● Globally pooled features are fed into a series of fully-connected layers with 2048 RELUs
Xavier Amatriain – August 2014 – KDD
ANN Training over GPUS and AWS
● How did we implement our ANN solution at Netflix?
○ Level 1 distribution: machines over different AWS regions
○ Level 2 distribution: machines in AWS and same AWS region
■ Use coordination tools
● Spearmint or similar for parameter optimization
● Condor, StarCluster, Mesos… for distributed cluster coordination
○ Level 3 parallelization: highly optimized parallel CUDA code on GPUs
http://techblog.netflix.com/2014/02/distributed-neural-networks-with-gpus.html
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References

Recommended for you

Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...

This presentation explores the practical application of image description techniques. Familiar guidelines will be demonstrated in practice, and descriptions will be developed “live”! If you have learned a lot about the theory of image description techniques but want to feel more confident putting them into practice, this is the presentation for you. There will be useful, actionable information for everyone, whether you are working with authors, colleagues, alone, or leveraging AI as a collaborator. Link to presentation recording and slides: https://bnctechforum.ca/sessions/details-of-description-part-ii-describing-images-in-practice/ Presented by BookNet Canada on June 25, 2024, with support from the Department of Canadian Heritage.

a11yaccessibilityalt text
Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...
Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...
Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...

Today’s digitally connected world presents a wide range of security challenges for enterprises. Insider security threats are particularly noteworthy because they have the potential to cause significant harm. Unlike external threats, insider risks originate from within the company, making them more subtle and challenging to identify. This blog aims to provide a comprehensive understanding of insider security threats, including their types, examples, effects, and mitigation techniques.

insider securitycybersecurity threatsenterprise security
Observability For You and Me with OpenTelemetry
Observability For You and Me with OpenTelemetryObservability For You and Me with OpenTelemetry
Observability For You and Me with OpenTelemetry

Are you interested in dipping your toes in the cloud native observability waters, but as an engineer you are not sure where to get started with tracing problems through your microservices and application landscapes on Kubernetes? Then this is the session for you, where we take you on your first steps in an active open-source project that offers a buffet of languages, challenges, and opportunities for getting started with telemetry data. The project is called openTelemetry, but before diving into the specifics, we’ll start with de-mystifying key concepts and terms such as observability, telemetry, instrumentation, cardinality, percentile to lay a foundation. After understanding the nuts and bolts of observability and distributed traces, we’ll explore the openTelemetry community; its Special Interest Groups (SIGs), repositories, and how to become not only an end-user, but possibly a contributor.We will wrap up with an overview of the components in this project, such as the Collector, the OpenTelemetry protocol (OTLP), its APIs, and its SDKs. Attendees will leave with an understanding of key observability concepts, become grounded in distributed tracing terminology, be aware of the components of openTelemetry, and know how to take their first steps to an open-source contribution! Key Takeaways: Open source, vendor neutral instrumentation is an exciting new reality as the industry standardizes on openTelemetry for observability. OpenTelemetry is on a mission to enable effective observability by making high-quality, portable telemetry ubiquitous. The world of observability and monitoring today has a steep learning curve and in order to achieve ubiquity, the project would benefit from growing our contributor community.

cloudcloud native observabilitycloud native
Xavier Amatriain – August 2014 – KDD
3.4 Social Recommendations
Xavier Amatriain – August 2014 – KDD
Social and Trust-based
recommenders
● A social recommender system recommends items that are
“popular” in the social proximity of the user.
● Social proximity = trust (can also be topic-specific)
● Given two individuals - the source (node A) and sink (node C) -
derive how much the source should trust the sink.
● Algorithms
○ Advogato (Levien)
○ Appleseed (Ziegler and Lausen)
○ MoleTrust (Massa and Avesani)
○ TidalTrust (Golbeck)
Xavier Amatriain – August 2014 – KDD
Other ways to use Social
● Social connections can be used in
combination with other approaches
● In particular, “friendships” can be fed into
collaborative filtering methods in different
ways
− replace or modify user-user “similarity” by using
social network information
− use social connection as a part of the ML objective
function as regularizer
− ...
Xavier Amatriain – August 2014 – KDD
Demographic Methods
● Aim to categorize the user based on personal
attributes and make recommendation based
on demographic classes
● Demographic groups can come from
marketing research – hence experts decided
how to model the users
● Demographic techniques form people-to-
people correlations

Recommended for you

RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptxRPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx

Your comprehensive guide to RPA in healthcare for 2024. Explore the benefits, use cases, and emerging trends of robotic process automation. Understand the challenges and prepare for the future of healthcare automation

rpa in healthcarerpa in healthcare usarpa in healthcare industry
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - MydbopsScaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops

This presentation, delivered at the Postgres Bangalore (PGBLR) Meetup-2 on June 29th, 2024, dives deep into connection pooling for PostgreSQL databases. Aakash M, a PostgreSQL Tech Lead at Mydbops, explores the challenges of managing numerous connections and explains how connection pooling optimizes performance and resource utilization. Key Takeaways: * Understand why connection pooling is essential for high-traffic applications * Explore various connection poolers available for PostgreSQL, including pgbouncer * Learn the configuration options and functionalities of pgbouncer * Discover best practices for monitoring and troubleshooting connection pooling setups * Gain insights into real-world use cases and considerations for production environments This presentation is ideal for: * Database administrators (DBAs) * Developers working with PostgreSQL * DevOps engineers * Anyone interested in optimizing PostgreSQL performance Contact info@mydbops.com for PostgreSQL Managed, Consulting and Remote DBA Services

postgresqlpgsqldatabase
Password Rotation in 2024 is still Relevant
Password Rotation in 2024 is still RelevantPassword Rotation in 2024 is still Relevant
Password Rotation in 2024 is still Relevant

Password Rotation in 2024 is still Relevant

passwordmanagementrotation
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References
Xavier Amatriain – August 2014 – KDD
3.5. Page Optimization
Xavier Amatriain – August 2014 – KDD
Page Composition
Xavier Amatriain – August 2014 – KDD
10,000s
of
possible
rows
…
10-40
rows
Variable number of
possible videos per
row (up to
thousands)
1 personalized
page
per
device
Page Composition

Recommended for you

The Rise of Supernetwork Data Intensive Computing
The Rise of Supernetwork Data Intensive ComputingThe Rise of Supernetwork Data Intensive Computing
The Rise of Supernetwork Data Intensive Computing

Invited Remote Lecture to SC21 The International Conference for High Performance Computing, Networking, Storage, and Analysis St. Louis, Missouri November 18, 2021

distributed supercomputerdistributed machine learning
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf

Kief Morris rethinks the infrastructure code delivery lifecycle, advocating for a shift towards composable infrastructure systems. We should shift to designing around deployable components rather than code modules, use more useful levels of abstraction, and drive design and deployment from applications rather than bottom-up, monolithic architecture and delivery.

infrastructure as codeclouddevops
Xavier Amatriain – August 2014 – KDD
Page Composition
From “Modeling User Attention and
Interaction on the Web” 2014 - PhD Thesis by Dmitry Lagun (Emory U.)
Xavier Amatriain – August 2014 – KDD
More
likely to
see
Less likely
User Attention Modeling
Xavier Amatriain – August 2014 – KDD
User Attention Modeling
From “Modeling User Attention and
Interaction on the Web” 2014 - PhD Thesis by Dmitry Lagun (Emory U.)
Xavier Amatriain – August 2014 – KDD
vs.Accurate Diverse
vs.Discovery Continuation
vs.Depth Coverage
vs.Freshness Stability
vs.Recommendations Tasks
Page Composition
● To put things together we need to combine different elements
○ Navigational/Attention Model
○ Personalized Relevance Model
○ Diversity Model

Recommended for you

Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References
Xavier Amatriain – August 2014 – KDD
3.6 Tensor Factorization &
Factorization Machines
Xavier Amatriain – August 2014 – KDD
N-dimensional model
Xavier Amatriain – August 2014 – KDD
HOSVD: Higher Order Singular
Value Decomposition
Tensor Factorization

Recommended for you

Xavier Amatriain – August 2014 – KDD
Tensor Factorization
Where:
● We can use a simple squared error loss function:
● Or the absolute error loss
● The loss function over all users becomes
Xavier Amatriain – August 2014 – KDD
Factorization Machines
• Generalization of regularized matrix (and tensor)
factorization approaches combined with linear (or
logistic) regression
• Problem: Each new adaptation of matrix or tensor
factorization requires deriving new learning
algorithms
• Hard to adapt to new domains and add data sources
• Hard to advance the learning algorithms across approaches
• Hard to incorporate non-categorical variables
Xavier Amatriain – August 2014 – KDD
• Approach: Treat input as a real-valued feature vector
• Model both linear and pair-wise interaction of features (i.e.
polynomial regression)
• Traditional machine learning will overfit
• Factor pairwise interactions between features
• Reduced dimensionality of interactions promote generalization
• Different matrix factorizations become different feature
representations
• Tensors: Additional higher-order interactions
• Combines “generality of machine learning/regression
with quality of factorization models”
Factorization Machines
Xavier Amatriain – August 2014 – KDD
• Each feature gets a weight value and a factor vector
• O( ) parameters
• Model equation:
O( 2
)
O( )
Factorization Machines

Recommended for you

Xavier Amatriain – August 2014 – KDD
▪ Two categorical variables ( , ) encoded as real values:
▪ FM becomes identical to MF with biases:
Factorization Machines
Xavier Amatriain – August 2014 – KDD
▪ Makes it easy to add a time signal
▪ Equivalent equation:
From Rendle (2012) KDD Tutorial
Factorization Machines
Xavier Amatriain – August 2014 – KDD
• L2 regularized
• Regression: Optimize RMSE
• Classification: Optimize
logistic log-likelihood
• Ranking: Optimize scores
• Can be trained using:
• SGD
• Adaptive SGD
• ALS
• MCMC
Gradient:
Least squares SGD:
Factorization Machines (Rendle, 2010)
Xavier Amatriain – August 2014 – KDD
• Learning parameters:
• Number of factors
• Iterations
• Initialization scale
• Regularization (SGD, ALS) – Multiple
• Step size (SGD, A-SGD)
• MCMC removes the need to set those
hyperparameters
Factorization Machines (Rendle, 2010)

Recommended for you

Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References
Xavier Amatriain – August 2014 – KDD
3.7 MAB Explore/Exploit
Xavier Amatriain – August 2014 – KDD
• One of the key issues when building any kind of
personalization algorithm is how to trade off:
• Exploitation: Cashing in on what we know about
the user right now
• Exploration: Using the interaction as an
opportunity to learn more about the user
• We need to have informed and optimal strategies to
drive that tradeoff
• Solution: pick a reasonable set of candidates and
show users only “enough” to gather information
on them
Explore/Exploit
Xavier Amatriain – August 2014 – KDD
• Given possible strategies/candidates (slot machines) pick the
arm that has the maximum potential of being good (minimize
regret)
• Naive strategy:
• Explore with a small probability (e.g. 5%) -> choose an
arm at random
• Exploit with a high probability (1- ) (e.g. 95%) -> choose
the best-known arm so far
• Translation to recommender systems
• Choose an arm = choose an item/choose an algorithm
(MAB testing)
Multi-armed Bandits

Recommended for you

Xavier Amatriain – August 2014 – KDD
• Better strategies not only take into account the mean of the
posterior, but also the variance
• Upper Confidence Bound (UCB)
• Show item with maximum score
• Score = Posterior mean +
• Where can be computed differently depending on the
variant of UCB (e.g. to the number of trials or the
measured variance)
• Thompson Sampling
• Given a posterior distribution
• Sample on each iteration and choose the action that
maximizes the expected reward
Multi-armed Bandits
Xavier Amatriain – August 2014 – KDD
Multi-armed Bandits
Xavier Amatriain – August 2014 – KDD
Index
1. The Recommender Problem
2. “Traditional” Methods
2.1. Collaborative Filtering
2.2. Content-based Recommendations
2.3. Hybrid Approaches
3. Beyond Traditional Methods
3.1. Learning to Rank
3.2. Similarity
3.3. Deep Learning
3.4. Social Recommendations
3.5. Page Optimization
3.6. Tensor Factorization and Factorization Machines
3.7. MAB Explore/Exploit
4. References
Xavier Amatriain – August 2014 – KDD
4. References

Recommended for you

Xavier Amatriain – August 2014 – KDD
● "Recommender Systems Handbook." Ricci, Francesco, Lior Rokach,
Bracha Shapira, and Paul B. Kantor. (2010).
● “Recommender systems: an introduction”. Jannach, Dietmar, et al.
Cambridge University Press, 2010.
● “Toward the Next Generation of Recommender Systems: A Survey of the
State-of-the-Art and Possible Extensions”. G. Adomavicious and A.
Tuzhilin. 2005. IEEE Transactions on Knowledge and Data Engineering,
17 (6)
● “Item-based Collaborative Filtering Recommendation Algorithms”, B.
Sarwar et al. 2001. Proceedings of World Wide Web Conference.
● “Lessons from the Netflix Prize Challenge.”. R. M. Bell and Y. Koren.
SIGKDD Explor. Newsl., 9(2):75–79, December 2007.
● “Beyond algorithms: An HCI perspective on recommender systems”. K.
Swearingen and R. Sinha. In ACM SIGIR 2001 Workshop on
Recommender Systems
● “Recommender Systems in E-Commerce”. J. Ben Schafer et al. ACM
Conference on Electronic Commerce. 1999-
● “Introduction to Data Mining”, P. Tan et al. Addison Wesley. 2005
References
Xavier Amatriain – August 2014 – KDD
● “Evaluating collaborative filtering recommender systems”. J. L. Herlocker,
J. A. Konstan, L. G. Terveen, and J. T. Riedl. ACM Trans. Inf. Syst., 22(1):
5–53, 2004.
● “Trust in recommender systems”. J. O’Donovan and B. Smyth. In Proc. of
IUI ’05, 2005.
● “Content-based recommendation systems”. M. Pazzani and D. Billsus. In
The Adaptive Web, volume 4321. 2007.
● “Fast context-aware recommendations with factorization machines”. S.
Rendle, Z. Gantner, C. Freudenthaler, and L. Schmidt-Thieme. In Proc. of
the 34th ACM SIGIR, 2011.
● “Restricted Boltzmann machines for collaborative filtering”. R.
Salakhutdinov, A. Mnih, and G. E. Hinton.In Proc of ICML ’07, 2007
● “Learning to rank: From pairwise approach to listwise approach”. Z. Cao
and T. Liu. In In Proceedings of the 24th ICML, 2007.
● “Introduction to Data Mining”, P. Tan et al. Addison Wesley. 2005
References
Xavier Amatriain – August 2014 – KDD
● D. H. Stern, R. Herbrich, and T. Graepel. “Matchbox: large scale online
bayesian recommendations”. In Proc.of the 18th WWW, 2009.
● Koren Y and J. Sill. “OrdRec: an ordinal model for predicting personalized
item rating distributions”. In Rec-Sys ’11.
● Y. Koren. “Factorization meets the neighborhood: a multifaceted
collaborative filtering model”. In Proceedings of the 14th ACM SIGKDD,
2008.
● Yifan Hu, Y. Koren, and C. Volinsky. “Collaborative Filtering for Implicit
Feedback Datasets”. In Proc. Of the 2008 Eighth ICDM
● Y. Shi, A. Karatzoglou, L. Baltrunas, M. Larson, N. Oliver, and A. Hanjalic.
“CLiMF: learning to maximize reciprocal rank with collaborative less-is-
more filtering”. In Proc. of the sixth Recsys, 2012.
● Y. Shi, A. Karatzoglou, L. Baltrunas, M. Larson,A. Hanjalic, and N. Oliver.
“TFMAP: optimizing MAP for top-n context-aware recommendation”. In
Proc. Of the 35th SIGIR, 2012.
● C. Burges. 2010. From RankNet to LambdaRank to LambdaMART: An
Overview. MSFT Technical Report
References
Xavier Amatriain – August 2014 – KDD
● A. Karatzoglou, X. Amatriain, L. Baltrunas, and N. Oliver. “Multiverse
recommendation: n-dimensional tensor factorization for context-aware
collaborative filtering”. In Proc. of the fourth ACM Recsys, 2010.
● S. Rendle, Z. Gantner, C. Freudenthaler, and L. Schmidt-Thieme. “Fast
context-aware recommendations with factorization machines”. In Proc. of
the 34th ACM SIGIR, 2011.
● S.H. Yang, B. Long, A.J. Smola, H. Zha, and Z. Zheng. “Collaborative
competitive filtering: learning recommender using context of user choice. In
Proc. of the 34th ACM SIGIR, 2011.
● N. N. Liu, X. Meng, C. Liu, and Q. Yang. “Wisdom of the better few: cold
start recommendation via representative based rating elicitation”. In Proc.
of RecSys’11, 2011.
● M. Jamali and M. Ester. “Trustwalker: a random walk model for combining
trust-based and item-based recommendation”. In Proc. of KDD ’09, 2009.
References

Recommended for you

Xavier Amatriain – August 2014 – KDD
● J. Noel, S. Sanner, K. Tran, P. Christen, L. Xie, E. V. Bonilla, E.
Abbasnejad, and N. Della Penna. “New objective functions for social
collaborative filtering”. In Proc. of WWW ’12, pages 859–868, 2012.
● X. Yang, H. Steck, Y. Guo, and Y. Liu. “On top-k recommendation using
social networks”. In Proc. of RecSys’12, 2012.
● Dmitry Lagun. “Modeling User Attention and Interaction on the Web” 2014
PhD Thesis (Emory Un.)
● “Robust Models of Mouse Movement on Dynamic Web Search Results
Pages - F. Diaz et al. In Proc. of CIKM 2013
● Amr Ahmed et al. “Fair and balanced”. In Proc. WSDM 2013
● Deepak Agarwal. 2013. Recommending Items to Users: An Explore Exploit
Perspective. CIKM ‘13
References
Xavier Amatriain – August 2014 – KDD
● Recsys Wiki: http://recsyswiki.com/
● Recsys conference Webpage: http://recsys.acm.org/
● Recommender Systems Books Webpage: http://www.
recommenderbook.net/
● Mahout Project: http://mahout.apache.org/
● MyMediaLite Project: http://www.mymedialite.net/
Online resources
Xavier Amatriain – August 2014 – KDD
Thanks!
Questions?
Xavier Amatriain
xavier@netflix.com
@xamat

More Related Content

What's hot

Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
Francesco Casalegno
 
Past, Present & Future of Recommender Systems: An Industry Perspective
Past, Present & Future of Recommender Systems: An Industry PerspectivePast, Present & Future of Recommender Systems: An Industry Perspective
Past, Present & Future of Recommender Systems: An Industry Perspective
Justin Basilico
 
Recommender system
Recommender systemRecommender system
Recommender system
Nilotpal Pramanik
 
Interactive Recommender Systems with Netflix and Spotify
Interactive Recommender Systems with Netflix and SpotifyInteractive Recommender Systems with Netflix and Spotify
Interactive Recommender Systems with Netflix and Spotify
Chris Johnson
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
Lior Rokach
 
Recommender systems: Content-based and collaborative filtering
Recommender systems: Content-based and collaborative filteringRecommender systems: Content-based and collaborative filtering
Recommender systems: Content-based and collaborative filtering
Viet-Trung TRAN
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
Carlos Castillo (ChaTo)
 
Recent Trends in Personalization at Netflix
Recent Trends in Personalization at NetflixRecent Trends in Personalization at Netflix
Recent Trends in Personalization at Netflix
Justin Basilico
 
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Xavier Amatriain
 
Tutorial: Context In Recommender Systems
Tutorial: Context In Recommender SystemsTutorial: Context In Recommender Systems
Tutorial: Context In Recommender Systems
YONG ZHENG
 
Recommendation engines
Recommendation enginesRecommendation engines
Recommendation engines
Georgian Micsa
 
Sequential Decision Making in Recommendations
Sequential Decision Making in RecommendationsSequential Decision Making in Recommendations
Sequential Decision Making in Recommendations
Jaya Kawale
 
Recommendation System
Recommendation SystemRecommendation System
Recommendation System
Anamta Sayyed
 
Recent Trends in Personalization: A Netflix Perspective
Recent Trends in Personalization: A Netflix PerspectiveRecent Trends in Personalization: A Netflix Perspective
Recent Trends in Personalization: A Netflix Perspective
Justin Basilico
 
An introduction to Recommender Systems
An introduction to Recommender SystemsAn introduction to Recommender Systems
An introduction to Recommender Systems
David Zibriczky
 
AbemaTVにおける推薦システム
AbemaTVにおける推薦システムAbemaTVにおける推薦システム
AbemaTVにおける推薦システム
cyberagent
 
Recommender systems using collaborative filtering
Recommender systems using collaborative filteringRecommender systems using collaborative filtering
Recommender systems using collaborative filtering
D Yogendra Rao
 
Personalized Page Generation for Browsing Recommendations
Personalized Page Generation for Browsing RecommendationsPersonalized Page Generation for Browsing Recommendations
Personalized Page Generation for Browsing Recommendations
Justin Basilico
 
分散表現を用いたリアルタイム学習型セッションベース推薦システム
分散表現を用いたリアルタイム学習型セッションベース推薦システム分散表現を用いたリアルタイム学習型セッションベース推薦システム
分散表現を用いたリアルタイム学習型セッションベース推薦システム
Okamoto Laboratory, The University of Electro-Communications
 
Past, present, and future of Recommender Systems: an industry perspective
Past, present, and future of Recommender Systems: an industry perspectivePast, present, and future of Recommender Systems: an industry perspective
Past, present, and future of Recommender Systems: an industry perspective
Xavier Amatriain
 

What's hot (20)

Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
 
Past, Present & Future of Recommender Systems: An Industry Perspective
Past, Present & Future of Recommender Systems: An Industry PerspectivePast, Present & Future of Recommender Systems: An Industry Perspective
Past, Present & Future of Recommender Systems: An Industry Perspective
 
Recommender system
Recommender systemRecommender system
Recommender system
 
Interactive Recommender Systems with Netflix and Spotify
Interactive Recommender Systems with Netflix and SpotifyInteractive Recommender Systems with Netflix and Spotify
Interactive Recommender Systems with Netflix and Spotify
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
 
Recommender systems: Content-based and collaborative filtering
Recommender systems: Content-based and collaborative filteringRecommender systems: Content-based and collaborative filtering
Recommender systems: Content-based and collaborative filtering
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
 
Recent Trends in Personalization at Netflix
Recent Trends in Personalization at NetflixRecent Trends in Personalization at Netflix
Recent Trends in Personalization at Netflix
 
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
 
Tutorial: Context In Recommender Systems
Tutorial: Context In Recommender SystemsTutorial: Context In Recommender Systems
Tutorial: Context In Recommender Systems
 
Recommendation engines
Recommendation enginesRecommendation engines
Recommendation engines
 
Sequential Decision Making in Recommendations
Sequential Decision Making in RecommendationsSequential Decision Making in Recommendations
Sequential Decision Making in Recommendations
 
Recommendation System
Recommendation SystemRecommendation System
Recommendation System
 
Recent Trends in Personalization: A Netflix Perspective
Recent Trends in Personalization: A Netflix PerspectiveRecent Trends in Personalization: A Netflix Perspective
Recent Trends in Personalization: A Netflix Perspective
 
An introduction to Recommender Systems
An introduction to Recommender SystemsAn introduction to Recommender Systems
An introduction to Recommender Systems
 
AbemaTVにおける推薦システム
AbemaTVにおける推薦システムAbemaTVにおける推薦システム
AbemaTVにおける推薦システム
 
Recommender systems using collaborative filtering
Recommender systems using collaborative filteringRecommender systems using collaborative filtering
Recommender systems using collaborative filtering
 
Personalized Page Generation for Browsing Recommendations
Personalized Page Generation for Browsing RecommendationsPersonalized Page Generation for Browsing Recommendations
Personalized Page Generation for Browsing Recommendations
 
分散表現を用いたリアルタイム学習型セッションベース推薦システム
分散表現を用いたリアルタイム学習型セッションベース推薦システム分散表現を用いたリアルタイム学習型セッションベース推薦システム
分散表現を用いたリアルタイム学習型セッションベース推薦システム
 
Past, present, and future of Recommender Systems: an industry perspective
Past, present, and future of Recommender Systems: an industry perspectivePast, present, and future of Recommender Systems: an industry perspective
Past, present, and future of Recommender Systems: an industry perspective
 

Similar to Kdd 2014 Tutorial - the recommender problem revisited

Sistema de recomendações de Filmes do Netflix
Sistema de recomendações de Filmes do NetflixSistema de recomendações de Filmes do Netflix
Sistema de recomendações de Filmes do Netflix
Gabriel Peixe
 
Recommandation systems -
Recommandation systems - Recommandation systems -
Recommandation systems -
Yousef Fadila
 
Mini-training: Personalization & Recommendation Demystified
Mini-training: Personalization & Recommendation DemystifiedMini-training: Personalization & Recommendation Demystified
Mini-training: Personalization & Recommendation Demystified
Betclic Everest Group Tech Team
 
Intro to R and Data Mining 2012 09 27
Intro to R and Data Mining 2012 09 27Intro to R and Data Mining 2012 09 27
Intro to R and Data Mining 2012 09 27
Raj Kasarabada
 
BeaconsAI engr 245 lean launchpad stanford 2019
BeaconsAI engr 245 lean launchpad stanford 2019BeaconsAI engr 245 lean launchpad stanford 2019
BeaconsAI engr 245 lean launchpad stanford 2019
Stanford University
 
Modern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in MendeleyModern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in Mendeley
Kris Jack
 
Олександр Обєдніков “Рекомендательные системы”
Олександр Обєдніков “Рекомендательные системы”Олександр Обєдніков “Рекомендательные системы”
Олександр Обєдніков “Рекомендательные системы”
Dakiry
 
Fashiondatasc
FashiondatascFashiondatasc
Lecture Notes on Recommender System Introduction
Lecture Notes on Recommender System IntroductionLecture Notes on Recommender System Introduction
Lecture Notes on Recommender System Introduction
PerumalPitchandi
 
recommendation system techunique and issue
recommendation system techunique and issuerecommendation system techunique and issue
recommendation system techunique and issue
NutanBhor
 
Lean Product Development 101
Lean Product Development 101Lean Product Development 101
Lean Product Development 101
Cloud Elements
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
Alan Said
 
Natural Intelligence the human factor in AI
Natural Intelligence the human factor in AINatural Intelligence the human factor in AI
Natural Intelligence the human factor in AI
Bill Liu
 
Building a UX Research Program
Building a UX Research ProgramBuilding a UX Research Program
Building a UX Research Program
Kelley Howell
 
Overview of recommender system
Overview of recommender systemOverview of recommender system
Overview of recommender system
Stanley Wang
 
Ariadne's Thread -- Exploring a world of networked information built from fre...
Ariadne's Thread -- Exploring a world of networked information built from fre...Ariadne's Thread -- Exploring a world of networked information built from fre...
Ariadne's Thread -- Exploring a world of networked information built from fre...
Shenghui Wang
 
Modern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in MendeleyModern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in Mendeley
Maya Hristakeva
 
Recommender lecture
Recommender lectureRecommender lecture
Recommender lecture
Aravindharamanan S
 
Exploratory Analysis of User Data
Exploratory Analysis of User DataExploratory Analysis of User Data
Exploratory Analysis of User Data
Behrooz Omidvar-Tehrani
 
Endorphin. Making sense of social data
Endorphin. Making sense of social dataEndorphin. Making sense of social data
Endorphin. Making sense of social data
Artem Zavyalov
 

Similar to Kdd 2014 Tutorial - the recommender problem revisited (20)

Sistema de recomendações de Filmes do Netflix
Sistema de recomendações de Filmes do NetflixSistema de recomendações de Filmes do Netflix
Sistema de recomendações de Filmes do Netflix
 
Recommandation systems -
Recommandation systems - Recommandation systems -
Recommandation systems -
 
Mini-training: Personalization & Recommendation Demystified
Mini-training: Personalization & Recommendation DemystifiedMini-training: Personalization & Recommendation Demystified
Mini-training: Personalization & Recommendation Demystified
 
Intro to R and Data Mining 2012 09 27
Intro to R and Data Mining 2012 09 27Intro to R and Data Mining 2012 09 27
Intro to R and Data Mining 2012 09 27
 
BeaconsAI engr 245 lean launchpad stanford 2019
BeaconsAI engr 245 lean launchpad stanford 2019BeaconsAI engr 245 lean launchpad stanford 2019
BeaconsAI engr 245 lean launchpad stanford 2019
 
Modern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in MendeleyModern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in Mendeley
 
Олександр Обєдніков “Рекомендательные системы”
Олександр Обєдніков “Рекомендательные системы”Олександр Обєдніков “Рекомендательные системы”
Олександр Обєдніков “Рекомендательные системы”
 
Fashiondatasc
FashiondatascFashiondatasc
Fashiondatasc
 
Lecture Notes on Recommender System Introduction
Lecture Notes on Recommender System IntroductionLecture Notes on Recommender System Introduction
Lecture Notes on Recommender System Introduction
 
recommendation system techunique and issue
recommendation system techunique and issuerecommendation system techunique and issue
recommendation system techunique and issue
 
Lean Product Development 101
Lean Product Development 101Lean Product Development 101
Lean Product Development 101
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
 
Natural Intelligence the human factor in AI
Natural Intelligence the human factor in AINatural Intelligence the human factor in AI
Natural Intelligence the human factor in AI
 
Building a UX Research Program
Building a UX Research ProgramBuilding a UX Research Program
Building a UX Research Program
 
Overview of recommender system
Overview of recommender systemOverview of recommender system
Overview of recommender system
 
Ariadne's Thread -- Exploring a world of networked information built from fre...
Ariadne's Thread -- Exploring a world of networked information built from fre...Ariadne's Thread -- Exploring a world of networked information built from fre...
Ariadne's Thread -- Exploring a world of networked information built from fre...
 
Modern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in MendeleyModern Perspectives on Recommender Systems and their Applications in Mendeley
Modern Perspectives on Recommender Systems and their Applications in Mendeley
 
Recommender lecture
Recommender lectureRecommender lecture
Recommender lecture
 
Exploratory Analysis of User Data
Exploratory Analysis of User DataExploratory Analysis of User Data
Exploratory Analysis of User Data
 
Endorphin. Making sense of social data
Endorphin. Making sense of social dataEndorphin. Making sense of social data
Endorphin. Making sense of social data
 

More from Xavier Amatriain

Data/AI driven product development: from video streaming to telehealth
Data/AI driven product development: from video streaming to telehealthData/AI driven product development: from video streaming to telehealth
Data/AI driven product development: from video streaming to telehealth
Xavier Amatriain
 
AI-driven product innovation: from Recommender Systems to COVID-19
AI-driven product innovation: from Recommender Systems to COVID-19AI-driven product innovation: from Recommender Systems to COVID-19
AI-driven product innovation: from Recommender Systems to COVID-19
Xavier Amatriain
 
AI for COVID-19 - Q42020 update
AI for COVID-19 - Q42020 updateAI for COVID-19 - Q42020 update
AI for COVID-19 - Q42020 update
Xavier Amatriain
 
AI for COVID-19: An online virtual care approach
AI for COVID-19: An online virtual care approachAI for COVID-19: An online virtual care approach
AI for COVID-19: An online virtual care approach
Xavier Amatriain
 
Lessons learned from building practical deep learning systems
Lessons learned from building practical deep learning systemsLessons learned from building practical deep learning systems
Lessons learned from building practical deep learning systems
Xavier Amatriain
 
AI for healthcare: Scaling Access and Quality of Care for Everyone
AI for healthcare: Scaling Access and Quality of Care for EveryoneAI for healthcare: Scaling Access and Quality of Care for Everyone
AI for healthcare: Scaling Access and Quality of Care for Everyone
Xavier Amatriain
 
Towards online universal quality healthcare through AI
Towards online universal quality healthcare through AITowards online universal quality healthcare through AI
Towards online universal quality healthcare through AI
Xavier Amatriain
 
From one to zero: Going smaller as a growth strategy
From one to zero: Going smaller as a growth strategyFrom one to zero: Going smaller as a growth strategy
From one to zero: Going smaller as a growth strategy
Xavier Amatriain
 
Learning to speak medicine
Learning to speak medicineLearning to speak medicine
Learning to speak medicine
Xavier Amatriain
 
ML to cure the world
ML to cure the worldML to cure the world
ML to cure the world
Xavier Amatriain
 
Recommender Systems In Industry
Recommender Systems In IndustryRecommender Systems In Industry
Recommender Systems In Industry
Xavier Amatriain
 
Medical advice as a Recommender System
Medical advice as a Recommender SystemMedical advice as a Recommender System
Medical advice as a Recommender System
Xavier Amatriain
 
Past present and future of Recommender Systems: an Industry Perspective
Past present and future of Recommender Systems: an Industry PerspectivePast present and future of Recommender Systems: an Industry Perspective
Past present and future of Recommender Systems: an Industry Perspective
Xavier Amatriain
 
Staying Shallow & Lean in a Deep Learning World
Staying Shallow & Lean in a Deep Learning WorldStaying Shallow & Lean in a Deep Learning World
Staying Shallow & Lean in a Deep Learning World
Xavier Amatriain
 
Machine Learning for Q&A Sites: The Quora Example
Machine Learning for Q&A Sites: The Quora ExampleMachine Learning for Q&A Sites: The Quora Example
Machine Learning for Q&A Sites: The Quora Example
Xavier Amatriain
 
BIG2016- Lessons Learned from building real-life user-focused Big Data systems
BIG2016- Lessons Learned from building real-life user-focused Big Data systemsBIG2016- Lessons Learned from building real-life user-focused Big Data systems
BIG2016- Lessons Learned from building real-life user-focused Big Data systems
Xavier Amatriain
 
Strata 2016 - Lessons Learned from building real-life Machine Learning Systems
Strata 2016 -  Lessons Learned from building real-life Machine Learning SystemsStrata 2016 -  Lessons Learned from building real-life Machine Learning Systems
Strata 2016 - Lessons Learned from building real-life Machine Learning Systems
Xavier Amatriain
 
Barcelona ML Meetup - Lessons Learned
Barcelona ML Meetup - Lessons LearnedBarcelona ML Meetup - Lessons Learned
Barcelona ML Meetup - Lessons Learned
Xavier Amatriain
 
10 more lessons learned from building Machine Learning systems - MLConf
10 more lessons learned from building Machine Learning systems - MLConf10 more lessons learned from building Machine Learning systems - MLConf
10 more lessons learned from building Machine Learning systems - MLConf
Xavier Amatriain
 
10 more lessons learned from building Machine Learning systems
10 more lessons learned from building Machine Learning systems10 more lessons learned from building Machine Learning systems
10 more lessons learned from building Machine Learning systems
Xavier Amatriain
 

More from Xavier Amatriain (20)

Data/AI driven product development: from video streaming to telehealth
Data/AI driven product development: from video streaming to telehealthData/AI driven product development: from video streaming to telehealth
Data/AI driven product development: from video streaming to telehealth
 
AI-driven product innovation: from Recommender Systems to COVID-19
AI-driven product innovation: from Recommender Systems to COVID-19AI-driven product innovation: from Recommender Systems to COVID-19
AI-driven product innovation: from Recommender Systems to COVID-19
 
AI for COVID-19 - Q42020 update
AI for COVID-19 - Q42020 updateAI for COVID-19 - Q42020 update
AI for COVID-19 - Q42020 update
 
AI for COVID-19: An online virtual care approach
AI for COVID-19: An online virtual care approachAI for COVID-19: An online virtual care approach
AI for COVID-19: An online virtual care approach
 
Lessons learned from building practical deep learning systems
Lessons learned from building practical deep learning systemsLessons learned from building practical deep learning systems
Lessons learned from building practical deep learning systems
 
AI for healthcare: Scaling Access and Quality of Care for Everyone
AI for healthcare: Scaling Access and Quality of Care for EveryoneAI for healthcare: Scaling Access and Quality of Care for Everyone
AI for healthcare: Scaling Access and Quality of Care for Everyone
 
Towards online universal quality healthcare through AI
Towards online universal quality healthcare through AITowards online universal quality healthcare through AI
Towards online universal quality healthcare through AI
 
From one to zero: Going smaller as a growth strategy
From one to zero: Going smaller as a growth strategyFrom one to zero: Going smaller as a growth strategy
From one to zero: Going smaller as a growth strategy
 
Learning to speak medicine
Learning to speak medicineLearning to speak medicine
Learning to speak medicine
 
ML to cure the world
ML to cure the worldML to cure the world
ML to cure the world
 
Recommender Systems In Industry
Recommender Systems In IndustryRecommender Systems In Industry
Recommender Systems In Industry
 
Medical advice as a Recommender System
Medical advice as a Recommender SystemMedical advice as a Recommender System
Medical advice as a Recommender System
 
Past present and future of Recommender Systems: an Industry Perspective
Past present and future of Recommender Systems: an Industry PerspectivePast present and future of Recommender Systems: an Industry Perspective
Past present and future of Recommender Systems: an Industry Perspective
 
Staying Shallow & Lean in a Deep Learning World
Staying Shallow & Lean in a Deep Learning WorldStaying Shallow & Lean in a Deep Learning World
Staying Shallow & Lean in a Deep Learning World
 
Machine Learning for Q&A Sites: The Quora Example
Machine Learning for Q&A Sites: The Quora ExampleMachine Learning for Q&A Sites: The Quora Example
Machine Learning for Q&A Sites: The Quora Example
 
BIG2016- Lessons Learned from building real-life user-focused Big Data systems
BIG2016- Lessons Learned from building real-life user-focused Big Data systemsBIG2016- Lessons Learned from building real-life user-focused Big Data systems
BIG2016- Lessons Learned from building real-life user-focused Big Data systems
 
Strata 2016 - Lessons Learned from building real-life Machine Learning Systems
Strata 2016 -  Lessons Learned from building real-life Machine Learning SystemsStrata 2016 -  Lessons Learned from building real-life Machine Learning Systems
Strata 2016 - Lessons Learned from building real-life Machine Learning Systems
 
Barcelona ML Meetup - Lessons Learned
Barcelona ML Meetup - Lessons LearnedBarcelona ML Meetup - Lessons Learned
Barcelona ML Meetup - Lessons Learned
 
10 more lessons learned from building Machine Learning systems - MLConf
10 more lessons learned from building Machine Learning systems - MLConf10 more lessons learned from building Machine Learning systems - MLConf
10 more lessons learned from building Machine Learning systems - MLConf
 
10 more lessons learned from building Machine Learning systems
10 more lessons learned from building Machine Learning systems10 more lessons learned from building Machine Learning systems
10 more lessons learned from building Machine Learning systems
 

Recently uploaded

What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024
Stephanie Beckett
 
Coordinate Systems in FME 101 - Webinar Slides
Coordinate Systems in FME 101 - Webinar SlidesCoordinate Systems in FME 101 - Webinar Slides
Coordinate Systems in FME 101 - Webinar Slides
Safe Software
 
Quality Patents: Patents That Stand the Test of Time
Quality Patents: Patents That Stand the Test of TimeQuality Patents: Patents That Stand the Test of Time
Quality Patents: Patents That Stand the Test of Time
Aurora Consulting
 
Cookies program to display the information though cookie creation
Cookies program to display the information though cookie creationCookies program to display the information though cookie creation
Cookies program to display the information though cookie creation
shanthidl1
 
Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024
BookNet Canada
 
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdfINDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf
jackson110191
 
UiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs ConferenceUiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs Conference
UiPathCommunity
 
Quantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLMQuantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLM
Vijayananda Mohire
 
Best Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdfBest Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdf
Tatiana Al-Chueyr
 
Implementations of Fused Deposition Modeling in real world
Implementations of Fused Deposition Modeling  in real worldImplementations of Fused Deposition Modeling  in real world
Implementations of Fused Deposition Modeling in real world
Emerging Tech
 
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdfBT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
Neo4j
 
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Chris Swan
 
Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...
BookNet Canada
 
Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...
Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...
Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...
Bert Blevins
 
Observability For You and Me with OpenTelemetry
Observability For You and Me with OpenTelemetryObservability For You and Me with OpenTelemetry
Observability For You and Me with OpenTelemetry
Eric D. Schabell
 
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptxRPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
SynapseIndia
 
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - MydbopsScaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Mydbops
 
Password Rotation in 2024 is still Relevant
Password Rotation in 2024 is still RelevantPassword Rotation in 2024 is still Relevant
Password Rotation in 2024 is still Relevant
Bert Blevins
 
The Rise of Supernetwork Data Intensive Computing
The Rise of Supernetwork Data Intensive ComputingThe Rise of Supernetwork Data Intensive Computing
The Rise of Supernetwork Data Intensive Computing
Larry Smarr
 
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
Kief Morris
 

Recently uploaded (20)

What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024What’s New in Teams Calling, Meetings and Devices May 2024
What’s New in Teams Calling, Meetings and Devices May 2024
 
Coordinate Systems in FME 101 - Webinar Slides
Coordinate Systems in FME 101 - Webinar SlidesCoordinate Systems in FME 101 - Webinar Slides
Coordinate Systems in FME 101 - Webinar Slides
 
Quality Patents: Patents That Stand the Test of Time
Quality Patents: Patents That Stand the Test of TimeQuality Patents: Patents That Stand the Test of Time
Quality Patents: Patents That Stand the Test of Time
 
Cookies program to display the information though cookie creation
Cookies program to display the information though cookie creationCookies program to display the information though cookie creation
Cookies program to display the information though cookie creation
 
Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024Details of description part II: Describing images in practice - Tech Forum 2024
Details of description part II: Describing images in practice - Tech Forum 2024
 
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdfINDIAN AIR FORCE FIGHTER PLANES LIST.pdf
INDIAN AIR FORCE FIGHTER PLANES LIST.pdf
 
UiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs ConferenceUiPath Community Day Kraków: Devs4Devs Conference
UiPath Community Day Kraków: Devs4Devs Conference
 
Quantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLMQuantum Communications Q&A with Gemini LLM
Quantum Communications Q&A with Gemini LLM
 
Best Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdfBest Practices for Effectively Running dbt in Airflow.pdf
Best Practices for Effectively Running dbt in Airflow.pdf
 
Implementations of Fused Deposition Modeling in real world
Implementations of Fused Deposition Modeling  in real worldImplementations of Fused Deposition Modeling  in real world
Implementations of Fused Deposition Modeling in real world
 
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdfBT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
BT & Neo4j: Knowledge Graphs for Critical Enterprise Systems.pptx.pdf
 
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
Fluttercon 2024: Showing that you care about security - OpenSSF Scorecards fo...
 
Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...Transcript: Details of description part II: Describing images in practice - T...
Transcript: Details of description part II: Describing images in practice - T...
 
Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...
Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...
Understanding Insider Security Threats: Types, Examples, Effects, and Mitigat...
 
Observability For You and Me with OpenTelemetry
Observability For You and Me with OpenTelemetryObservability For You and Me with OpenTelemetry
Observability For You and Me with OpenTelemetry
 
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptxRPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
RPA In Healthcare Benefits, Use Case, Trend And Challenges 2024.pptx
 
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - MydbopsScaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
Scaling Connections in PostgreSQL Postgres Bangalore(PGBLR) Meetup-2 - Mydbops
 
Password Rotation in 2024 is still Relevant
Password Rotation in 2024 is still RelevantPassword Rotation in 2024 is still Relevant
Password Rotation in 2024 is still Relevant
 
The Rise of Supernetwork Data Intensive Computing
The Rise of Supernetwork Data Intensive ComputingThe Rise of Supernetwork Data Intensive Computing
The Rise of Supernetwork Data Intensive Computing
 
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
[Talk] Moving Beyond Spaghetti Infrastructure [AOTB] 2024-07-04.pdf
 

Kdd 2014 Tutorial - the recommender problem revisited

  • 1. Xavier Amatriain – August 2014 – KDD The Recommender Problem Revisited Xavier Amatriain Research/Engineering Director @ Netflix @xamat Bamshad Mobasher Professor @ DePaul University
  • 2. Xavier Amatriain – August 2014 – KDD Index PART I - Xavier Amatriain (2h) 1. The Recommender Problem 2. Traditional Recommendation Methods 3. Beyond Traditional Methods PART 2 - Bamshad Mobasher (1h) 1. Context-aware Recommendations
  • 3. Xavier Amatriain – August 2014 – KDD Recent/Upcoming Publications ● The Recommender Problem Revisited. KDD and Recsys 2014 Tutorial ● KDD: Big & Personal: data and models behind Netflix recommendations. 2013 ● SIGKDD Explorations: Mining large streams of user data for personalized recommendations. 2012 ● Recsys: Building industrial-scale real-world recommender systems. 2012 ● Recys - Walk the Talk: Analyzing the relation between implicit and explicit feedback for preference elicitation. 2011 ● SIGIR – Temporal behavior of CF. 2010 ● Web Intelligence – Expert-based CF for music. 2010 ● Recsys – Tensor Factorization. 2010 ● Mobile HCI – Tourist Recommendation. 2010 ● Recsys Handbook (book) – Data mining for recsys. 2010 & Recommender Systems in Industry. 2014 ● SIGIR – Wisdom of the Few. 2009 ● Recsys – Denoising by re-rating. 2009 ● CARS – Implicit context-aware recommendations. 2009 ● UMAP – I like it I like it not. 2009
  • 4. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. Traditional Recommendation Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 5. Xavier Amatriain – August 2014 – KDD 1. The Recommender Problem
  • 6. Xavier Amatriain – August 2014 – KDD The Age of Search has come to an end • ... long live the Age of Recommendation! • Chris Anderson in “The Long Tail” • “We are leaving the age of information and entering the age of recommendation” • CNN Money, “The race to create a 'smart' Google”: • “The Web, they say, is leaving the era of search and entering one of discovery. What's the difference? Search is what you do when you're looking for something. Discovery is when something wonderful that you didn't know existed, or didn't know how to ask for, finds you.”
  • 7. Xavier Amatriain – August 2014 – KDD Information overload “People read around 10 MB worth of material a day, hear 400 MB a day, and see 1 MB of information every second” - The Economist, November 2006 In 2015, consumption will raise to 74 GB a day - UCSD Study 2014
  • 8. Xavier Amatriain – August 2014 – KDD Everything is personalized
  • 9. Xavier Amatriain – August 2014 – KDD The “Recommender problem” ● Traditional definition: Estimate a utility function that automatically predicts how a user will like an item. ● Based on: ○ Past behavior ○ Relations to other users ○ Item similarity ○ Context ○ …
  • 10. Xavier Amatriain – August 2014 – KDD Recommendation as data mining The core of the Recommendation Engine can be assimilated to a general data mining problem (Amatriain et al. Data Mining Methods for Recommender Systems in Recommender Systems Handbook)
  • 11. Xavier Amatriain – August 2014 – KDD Machine Learning + all those other things ● User Interface ● System requirements (efficiency, scalability, privacy....) ● Serendipity ● Diversity ● Awareness ● Explanations ● …
  • 12. Xavier Amatriain – August 2014 – KDD Serendipity ● Unsought finding ● Don't recommend items the user already knows or would have found anyway. ● Expand the user's taste into neighboring areas by improving the obvious ● Collaborative filtering can offer controllable serendipity (e.g. controlling how many neighbors to use in the recommendation)
  • 13. Xavier Amatriain – August 2014 – KDD Explanation/Support for Recommendations Social Support
  • 14. Xavier Amatriain – August 2014 – KDD Personalization awareness Diversity DadAll SonDaughterDad&Mom MomAll Daughter MomAll? Diversity & Awareness
  • 15. Xavier Amatriain – August 2014 – KDD What works ● Depends on the domain and particular problem ● However, in the general case it has been demonstrated that the best isolated approach is CF. ○ Other approaches can be hybridized to improve results in specific cases (cold-start problem...) ● What matters: ○ Data preprocessing: outlier removal, denoising, removal of global effects (e.g. individual user's average) ○ “Smart” dimensionality reduction using MF/SVD ○ Combining methods through ensembles
  • 16. Xavier Amatriain – August 2014 – KDD Evolution of the Recommender Problem Rating Ranking Page Optimization 4.7 Context-aware Recommendations Context
  • 17. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 18. Xavier Amatriain – August 2014 – KDD 2. Traditional Approaches
  • 19. Xavier Amatriain – August 2014 – KDD 2.1. Collaborative Filtering
  • 20. Xavier Amatriain – August 2014 – KDD The CF Ingredients ● List of m Users and a list of n Items ● Each user has a list of items with associated opinion ○ Explicit opinion - a rating score ○ Sometime the rating is implicitly – purchase records or listen to tracks ● Active user for whom the CF prediction task is performed ● Metric for measuring similarity between users ● Method for selecting a subset of neighbors ● Method for predicting a rating for items not currently rated by the active user.
  • 21. Xavier Amatriain – August 2014 – KDD Personalized vs Non-Personalised CF ● CF recommendations are personalized: prediction based only on similar users ● Non-personalized collaborative-based recommendation: averagge the recommendations of ALL the users ● How would the two approaches compare?
  • 22. Xavier Amatriain – August 2014 – KDD Personalized vs. Not Personalized ● Netflix Prize: it is very simple to produce “reasonable” recommendations and extremely difficult to improve them to become “great” ● But there is a huge difference in business value between reasonable and great 0,1510,2230,0222811718164974424EachMovie 0,1790,2330,041100020939526040MovieLens 0,1520,2200,725351944910048483Jester MAE Pers MAE Non Pers density total ratings itemsusersData Set
  • 23. Xavier Amatriain – August 2014 – KDD User-based Collaborative Filtering
  • 24. Xavier Amatriain – August 2014 – KDD Collaborative Filtering The basic steps: 1. Identify set of ratings for the target/active user 2. Identify set of users most similar to the target/active user according to a similarity function (neighborhood formation) 3. Identify the products these similar users liked 4. Generate a prediction - rating that would be given by the target user to the product - for each one of these products 5. Based on this predicted rating recommend a set of top N products
  • 25. Xavier Amatriain – August 2014 – KDD UB Collaborative Filtering ● A collection of user ui , i=1, …n and a collection of products pj , j=1, …, m ● An n × m matrix of ratings vij , with vij = ? if user i did not rate product j ● Prediction for user i and product j is computed as • Similarity can be computed by Pearson correlation or or
  • 26. Xavier Amatriain – August 2014 – KDD User-based CF Example
  • 27. Xavier Amatriain – August 2014 – KDD User-based CF Example
  • 28. Xavier Amatriain – August 2014 – KDD User-based CF Example
  • 29. Xavier Amatriain – August 2014 – KDD User-based CF Example
  • 30. Xavier Amatriain – August 2014 – KDD User-based CF Example
  • 31. Xavier Amatriain – August 2014 – KDD Item-based Collaborative Filtering
  • 32. Xavier Amatriain – August 2014 – KDD Item Based CF Algorithm ● Look into the items the target user has rated ● Compute how similar they are to the target item ○ Similarity only using past ratings from other users! ● Select k most similar items. ● Compute Prediction by taking weighted average on the target user’s ratings on the most similar items.
  • 33. Xavier Amatriain – August 2014 – KDD Item Similarity Computation ● Similarity: find users who have rated items and apply a similarity function to their ratings. ● Cosine-based Similarity (difference in rating scale between users is not taken into account) ● Adjusted Cosine Similarity (takes care of difference in rating scale)
  • 34. Xavier Amatriain – August 2014 – KDD Challenges of memory-based CF
  • 35. Xavier Amatriain – August 2014 – KDD CF: Pros/Cons ● Requires minimal knowledge engineering efforts ● Users and products are symbols without any internal structure or characteristics ● Produces good-enough results in most cases Challenges: ● Sparsity – evaluation of large itemsets where user/item interactions are under 1%. ● Scalability - Nearest neighbor require computation that grows with both the number of users and the number of items.
  • 36. Xavier Amatriain – August 2014 – KDD The Sparsity Problem ● Typically: large product sets, user ratings for a small percentage of them (e.g. Amazon: millions of books and a user may have bought hundreds of books) ● If you represent the Netflix Prize rating data in a User/Movie matrix you get… ○ 500,000 x 17,000 = 8.5 B positions ○ Out of which only 100M are non-zero ● Number of users needs to be ~ 0.1 x size of the catalog
  • 37. Xavier Amatriain – August 2014 – KDD Model-based Collaborative Filtering
  • 38. Xavier Amatriain – August 2014 – KDD Model Based CF Algorithms ● Memory based ○ Use the entire user-item database to generate a prediction. ○ Usage of statistical techniques to find the neighbors – e.g. nearest-neighbor. ● Model-based ○ First develop a model of user ○ Type of model: ■ Probabilistic (e.g. Bayesian Network) ■ Clustering ■ Rule-based approaches (e.g. Association Rules) ■ Classification ■ Regression ■ LDA ■ ...
  • 39. Xavier Amatriain – August 2014 – KDD Model-based CF: What we learned from the Netflix Prize
  • 40. Xavier Amatriain – August 2014 – KDD What we were interested in: ■ High quality recommendations Proxy question: ■ Accuracy in predicted rating ■ Improve by 10% = $1million!
  • 41. Xavier Amatriain – August 2014 – KDD 2007 Progress Prize ▪ Top 2 algorithms ▪ SVD - Prize RMSE: 0.8914 ▪ RBM - Prize RMSE: 0.8990 ▪ Linear blend Prize RMSE: 0.88 ▪ Currently in use as part of Netflix’ rating prediction component ▪ Limitations ▪ Designed for 100M ratings, we have 5B ratings ▪ Not adaptable as users add ratings ▪ Performance issues
  • 42. Xavier Amatriain – August 2014 – KDD SVD/MF ● X: m x n matrix (e.g., m users, n videos) ● U: m x r matrix (m users, r factors) ● S: r x r diagonal matrix (strength of each ‘factor’) (r: rank of the matrix) ● V: r x n matrix (n videos, r factor)
  • 43. Xavier Amatriain – August 2014 – KDD Simon Funk’s SVD ● One of the most interesting findings during the Netflix Prize came out of a blog post ● Incremental, iterative, and approximate way to compute the SVD using gradient descent
  • 44. Xavier Amatriain – August 2014 – KDD ▪ User factor vectors and item-factors vector ▪ Baseline (bias) (user & item deviation from average) ▪ Predict rating as ▪ Asymetric SVD (Koren et. Al) asymmetric variation w. implicit feedback ▪ Where ▪ are three item factor vectors ▪ Users are not parametrized, but rather represented by: ▪ R(u): items rated by user u ▪ N(u): items for which the user has given implicit preference (e.g. rated vs. not rated) SVD for Rating Prediction
  • 45. Xavier Amatriain – August 2014 – KDD Restricted Boltzmann Machines ● Each unit is a state that can be active or not active ● Each input to a unit is associated to a weight ● The transfer function calculates a score for every unit based on the weighted sum of inputs ● Score is passed to the activation function that calculates the probability of the unit to be active ● Restrict the connectivity to make learning easier. Only one layer of hidden units. No connections between hidden units. Hidden units are independent given visible states
  • 46. Xavier Amatriain – August 2014 – KDD RBM for Recommendations ● Each visible unit = an item ● Num. of hidden units a is parameter ● In training phase, for each user: ○ If user rated item, vi is activated ○ Activation states of vi = inputs to hj ○ Based on activation, hj is computed ○ Activation state of hj becomes input to vi ○ Activation state of vi is recalculated ○ Difference between current and past activation state for vi used to update weights wij and thresholds ● In prediction phase: ○ For the items of the user the vi are activated ○ Based on this the state of the hj is computed ○ The activation of hj is used as input to recompute the state of vi ○ Activation probabilities are used to recommend items
  • 47. Xavier Amatriain – August 2014 – KDD Putting all together ● Remember that current production model includes an ensemble of both SVD++ and RBMs
  • 48. Xavier Amatriain – August 2014 – KDD What about the final prize ensembles? ● Our offline studies showed they were too computationally intensive to scale ● Expected improvement not worth the engineering effort ● Plus…. Focus had already shifted to other issues that had more impact than rating prediction.
  • 49. Xavier Amatriain – August 2014 – KDD Clustering
  • 50. Xavier Amatriain – August 2014 – KDD Clustering ● Goal: cluster users and compute per-cluster “typical” preferences ● Users receive recommendations computed at the cluster level
  • 51. Xavier Amatriain – August 2014 – KDD Locality-sensitive Hashing (LSH) ● Method for grouping similar items in highly dimensional spaces ● Find a hashing function s.t. similar items are grouped in the same buckets ● Main application is Nearest-neighbors ○ Hashing function is found iteratively by concatenating random hashing functions ○ Addresses one of NN main concerns: performance
  • 52. Xavier Amatriain – August 2014 – KDD Other “interesting” clustering techniques ● k-means and all its variations ● Affinity Propagation ● Spectral Clustering ● Non-parametric Bayesian Clustering (e.g. HDPs)
  • 53. Xavier Amatriain – August 2014 – KDD Association Rules
  • 54. Xavier Amatriain – August 2014 – KDD Association rules ● Past purchases are interpreted as transactions of “associated” items ● If a visitor has some interest in Book 5, she will be recommended to buy Book 3 as well ● Recommendations are constrained to some minimum levels of confidence ● Fast to implement and execute (e.g. A Priori algorithm)
  • 55. Xavier Amatriain – August 2014 – KDD Classifiers
  • 56. Xavier Amatriain – August 2014 – KDD Classifiers ● Classifiers are general computational models trained using positive and negative examples ● They may take in inputs: ○ Vector of item features (action / adventure, Bruce Willis) ○ Preferences of customers (like action / adventure) ○ Relations among item ● E.g. Logistic Regression, Bayesian Networks, Support Vector Machines, Decision Trees, etc...
  • 57. Xavier Amatriain – August 2014 – KDD Classifiers ● Classifiers can be used in CF and CB Recommenders ● Pros: ○ Versatile ○ Can be combined with other methods to improve accuracy of recommendations ● Cons: ○ Need a relevant training set ○ May overfit (Regularization) ● E.g. Logistic Regression, Bayesian Networks, Support Vector Machines, Decision Trees, etc...
  • 58. Xavier Amatriain – August 2014 – KDD Limitations of Collaborative Filtering
  • 59. Xavier Amatriain – August 2014 – KDD Limitations of Collaborative Filtering ● Cold Start: There needs to be enough other users already in the system to find a match. New items need to get enough ratings. ● Popularity Bias: Hard to recommend items to someone with unique tastes. ○ Tends to recommend popular items (items from the tail do not get so much data)
  • 60. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Novel Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 61. Xavier Amatriain – August 2014 – KDD 2.2 Content-based Recommenders
  • 62. Xavier Amatriain – August 2014 – KDD Content-Based Recommendation ● Recommendations based on content of items rather than on other users’ opinions/interactions ● Goal: recommend items similar to those the user liked ● Common for recommending text-based products (web pages, usenet news messages, ) ● Items to recommend are “described” by their associated features (e.g. keywords) ● User Model structured in a “similar” way as the content: features/keywords more likely to occur in the preferred documents (lazy approach) ● The user model can be a classifier based on whatever technique (Neural Networks, Naïve Bayes...)
  • 63. Xavier Amatriain – August 2014 – KDD Pros/cons of CB Approach Pros ● No need for data on other users: No cold-start or sparsity ● Able to recommend to users with unique tastes. ● Able to recommend new and unpopular items ● Can provide explanations by listing content-features Cons ● Requires content that can be encoded as meaningful features (difficult in some domains/catalogs) ● Users represented as learnable function of content features. ● Difficult to implement serendipity ● Easy to overfit (e.g. for a user with few data points)
  • 64. Xavier Amatriain – August 2014 – KDD A word of caution
  • 65. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 66. Xavier Amatriain – August 2014 – KDD 2.3 Hybrid Approaches
  • 67. Xavier Amatriain – August 2014 – KDD Comparison of methods (FAB system) • Content–based recommendation with Bayesian classifier • Collaborative is standard using Pearson correlation • Collaboration via content uses the content-based user profiles Averaged on 44 users Precision computed in top 3 recommendations
  • 68. Xavier Amatriain – August 2014 – KDD Hybridization Methods Hybridization Method Description Weighted Outputs from several techniques (in the form of scores or votes) are combined with different degrees of importance to offer final recommendations Switching Depending on situation, the system changes from one technique to another Mixed Recommendations from several techniques are presented at the same time Feature combination Features from different recommendation sources are combined as input to a single technique Cascade The output from one technique is used as input of another that refines the result Feature augmentation The output from one technique is used as input features to another Meta-level The model learned by one recommender is used as input to another
  • 69. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 70. Xavier Amatriain – August 2014 – KDD 3. Beyond traditional approaches to Recommendation
  • 71. Xavier Amatriain – August 2014 – KDD 3.1 Ranking
  • 72. Xavier Amatriain – August 2014 – KDD Ranking Key algorithm, sorts titles in most contexts Ranking
  • 73. Xavier Amatriain – August 2014 – KDD Ranking ● Most recommendations are presented in a sorted list ● Recommendation can be understood as a ranking problem ● Popularity is the obvious baseline ● Ratings prediction is a clear secondary data input that allows for personalization ● Many other features can be added
  • 74. Xavier Amatriain – August 2014 – KDD Ranking by ratings 4.7 4.6 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 Niche titles High average ratings… by those who would watch it
  • 75. Xavier Amatriain – August 2014 – KDD RMSE
  • 76. Popularity PredictedRating 1 2 3 4 5 Linear Model: frank (u,v) = w1 p(v) + w2 r(u,v) + b FinalRanking Example: Two features, linear model
  • 78. Xavier Amatriain – August 2014 – KDD Ranking Ranking
  • 79. Xavier Amatriain – August 2014 – KDD Learning to rank ● Machine learning problem: goal is to construct ranking model from training data ● Training data can be a partial order or binary judgments (relevant/not relevant). ● Resulting order of the items typically induced from a numerical score ● Learning to rank is a key element for personalization ● You can treat the problem as a standard supervised classification problem
  • 80. Xavier Amatriain – August 2014 – KDD Learning to rank - Metrics ● Quality of ranking measured using metrics as ○ Normalized Discounted Cumulative Gain ○ Mean Reciprocal Rank (MRR) ○ Fraction of Concordant Pairs (FCP) ○ Others… ● But, it is hard to optimize machine-learned models directly on these measures (e.g. non-differentiable) ● Recent research on models that directly optimize ranking measures
  • 81. Xavier Amatriain – August 2014 – KDD Learning to rank - Approaches 1. Pointwise ■ Ranking function minimizes loss function defined on individual relevance judgment ■ Ranking score based on regression or classification ■ Ordinal regression, Logistic regression, SVM, GBDT, … 2. Pairwise ■ Loss function is defined on pair-wise preferences ■ Goal: minimize number of inversions in ranking ■ Ranking problem is then transformed into the binary classification problem ■ RankSVM, RankBoost, RankNet, FRank…
  • 82. Xavier Amatriain – August 2014 – KDD Learning to rank - Approaches 3. Listwise ■ Indirect Loss Function − RankCosine: similarity between ranking list and ground truth as loss function − ListNet: KL-divergence as loss function by defining a probability distribution − Problem: optimization of listwise loss function may not optimize IR metrics ■ Directly optimizing IR metric (difficult since they are not differentiable) − Genetic Programming or Simulated Annealing − LambdaMart weights pairwise errors in RankNet by IR metric − Gradient descent on smoothed version of objective function (e. g. CLiMF or TFMAP) − SVM-MAP relaxes MAP metric by adding to SVM constraints − AdaRank uses boosting to optimize NDCG
  • 83. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 84. Xavier Amatriain – August 2014 – KDD 3.2 Similarity as Recommendation
  • 85. Xavier Amatriain – August 2014 – KDD ▪ Displayed in many different contexts ▪ In response to user actions/context (search, queue add…) ▪ More like… rows Similars
  • 86. Xavier Amatriain – August 2014 – KDD Graph-based similarities 0.8 0.2 0.3 0. 4 0.3 0.7 0. 3
  • 87. Xavier Amatriain – August 2014 – KDD Example of graph-based similarity: SimRank ▪ SimRank (Jeh & Widom, 02): “two objects are similar if they are referenced by similar objects.”
  • 88. Xavier Amatriain – August 2014 – KDD Similarity ensembles • Similarity can refer to different dimensions • Similar in metadata/tags • Similar in user play behavior • Similar in user rating behavior • … • Combine them using an ensemble • Weights are learned using regression over existing response • Or… some MAB explore/exploit approach • The final concept of “similarity” responds to what users vote as similar
  • 89. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 90. Xavier Amatriain – August 2014 – KDD 3.3 Deep Learning for Recommendation
  • 91. Xavier Amatriain – August 2014 – KDD Deep Learning for Collaborative Filtering ● Let’s look at how Spotify uses Recurrent Networks for Playlist Prediction (http://erikbern.com/?p=589)
  • 92. Xavier Amatriain – August 2014 – KDD Deep Learning for Collaborative Filtering ● In order to predict the next track or movie a user is going to watch, we need to define a distribution ○ If we choose Softmax as it is common practice, we get: ● Problem: denominator (over all examples is very expensive to compute) ● Solution: build a tree that implements a hierarchical softmax ● More details on the blogpost
  • 93. Xavier Amatriain – August 2014 – KDD Deep Learning for Content-based Recommendations ● Another application of Deep Learning to recommendations also from Spotify ○ http://benanne.github.io/2014/08/05/spotify-cnns.html also Deep content-based music recommendation, Aäron van den Oord, Sander Dieleman and Benjamin Schrauwen, NIPS 2013 ● Application to coldstart new titles when very little CF information is available ● Using mel-spectrograms from the audio signal as input ● Training the deep neural network to predict 40 latent factors coming from Spotify’s CF solution
  • 94. Xavier Amatriain – August 2014 – KDD Deep Learning for Content-based Recommendations ● Network architecture made of 4 convolutional layers + 4 fully connected dense layers ● One dimensional convolutional layers using RELUs (Rectified Linear Units) with activation max(0,x) ● Max-pooling operations between convolutional layers to downsample intermediate representations in time, and add time invariance ● Global temporal pooling layer after last convolutional layer: pools across entire time axis, computing statistics of the learned features across time:: mean, maximum and L2-norm ● Globally pooled features are fed into a series of fully-connected layers with 2048 RELUs
  • 95. Xavier Amatriain – August 2014 – KDD ANN Training over GPUS and AWS ● How did we implement our ANN solution at Netflix? ○ Level 1 distribution: machines over different AWS regions ○ Level 2 distribution: machines in AWS and same AWS region ■ Use coordination tools ● Spearmint or similar for parameter optimization ● Condor, StarCluster, Mesos… for distributed cluster coordination ○ Level 3 parallelization: highly optimized parallel CUDA code on GPUs http://techblog.netflix.com/2014/02/distributed-neural-networks-with-gpus.html
  • 96. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 97. Xavier Amatriain – August 2014 – KDD 3.4 Social Recommendations
  • 98. Xavier Amatriain – August 2014 – KDD Social and Trust-based recommenders ● A social recommender system recommends items that are “popular” in the social proximity of the user. ● Social proximity = trust (can also be topic-specific) ● Given two individuals - the source (node A) and sink (node C) - derive how much the source should trust the sink. ● Algorithms ○ Advogato (Levien) ○ Appleseed (Ziegler and Lausen) ○ MoleTrust (Massa and Avesani) ○ TidalTrust (Golbeck)
  • 99. Xavier Amatriain – August 2014 – KDD Other ways to use Social ● Social connections can be used in combination with other approaches ● In particular, “friendships” can be fed into collaborative filtering methods in different ways − replace or modify user-user “similarity” by using social network information − use social connection as a part of the ML objective function as regularizer − ...
  • 100. Xavier Amatriain – August 2014 – KDD Demographic Methods ● Aim to categorize the user based on personal attributes and make recommendation based on demographic classes ● Demographic groups can come from marketing research – hence experts decided how to model the users ● Demographic techniques form people-to- people correlations
  • 101. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 102. Xavier Amatriain – August 2014 – KDD 3.5. Page Optimization
  • 103. Xavier Amatriain – August 2014 – KDD Page Composition
  • 104. Xavier Amatriain – August 2014 – KDD 10,000s of possible rows … 10-40 rows Variable number of possible videos per row (up to thousands) 1 personalized page per device Page Composition
  • 105. Xavier Amatriain – August 2014 – KDD Page Composition From “Modeling User Attention and Interaction on the Web” 2014 - PhD Thesis by Dmitry Lagun (Emory U.)
  • 106. Xavier Amatriain – August 2014 – KDD More likely to see Less likely User Attention Modeling
  • 107. Xavier Amatriain – August 2014 – KDD User Attention Modeling From “Modeling User Attention and Interaction on the Web” 2014 - PhD Thesis by Dmitry Lagun (Emory U.)
  • 108. Xavier Amatriain – August 2014 – KDD vs.Accurate Diverse vs.Discovery Continuation vs.Depth Coverage vs.Freshness Stability vs.Recommendations Tasks Page Composition ● To put things together we need to combine different elements ○ Navigational/Attention Model ○ Personalized Relevance Model ○ Diversity Model
  • 109. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 110. Xavier Amatriain – August 2014 – KDD 3.6 Tensor Factorization & Factorization Machines
  • 111. Xavier Amatriain – August 2014 – KDD N-dimensional model
  • 112. Xavier Amatriain – August 2014 – KDD HOSVD: Higher Order Singular Value Decomposition Tensor Factorization
  • 113. Xavier Amatriain – August 2014 – KDD Tensor Factorization Where: ● We can use a simple squared error loss function: ● Or the absolute error loss ● The loss function over all users becomes
  • 114. Xavier Amatriain – August 2014 – KDD Factorization Machines • Generalization of regularized matrix (and tensor) factorization approaches combined with linear (or logistic) regression • Problem: Each new adaptation of matrix or tensor factorization requires deriving new learning algorithms • Hard to adapt to new domains and add data sources • Hard to advance the learning algorithms across approaches • Hard to incorporate non-categorical variables
  • 115. Xavier Amatriain – August 2014 – KDD • Approach: Treat input as a real-valued feature vector • Model both linear and pair-wise interaction of features (i.e. polynomial regression) • Traditional machine learning will overfit • Factor pairwise interactions between features • Reduced dimensionality of interactions promote generalization • Different matrix factorizations become different feature representations • Tensors: Additional higher-order interactions • Combines “generality of machine learning/regression with quality of factorization models” Factorization Machines
  • 116. Xavier Amatriain – August 2014 – KDD • Each feature gets a weight value and a factor vector • O( ) parameters • Model equation: O( 2 ) O( ) Factorization Machines
  • 117. Xavier Amatriain – August 2014 – KDD ▪ Two categorical variables ( , ) encoded as real values: ▪ FM becomes identical to MF with biases: Factorization Machines
  • 118. Xavier Amatriain – August 2014 – KDD ▪ Makes it easy to add a time signal ▪ Equivalent equation: From Rendle (2012) KDD Tutorial Factorization Machines
  • 119. Xavier Amatriain – August 2014 – KDD • L2 regularized • Regression: Optimize RMSE • Classification: Optimize logistic log-likelihood • Ranking: Optimize scores • Can be trained using: • SGD • Adaptive SGD • ALS • MCMC Gradient: Least squares SGD: Factorization Machines (Rendle, 2010)
  • 120. Xavier Amatriain – August 2014 – KDD • Learning parameters: • Number of factors • Iterations • Initialization scale • Regularization (SGD, ALS) – Multiple • Step size (SGD, A-SGD) • MCMC removes the need to set those hyperparameters Factorization Machines (Rendle, 2010)
  • 121. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 122. Xavier Amatriain – August 2014 – KDD 3.7 MAB Explore/Exploit
  • 123. Xavier Amatriain – August 2014 – KDD • One of the key issues when building any kind of personalization algorithm is how to trade off: • Exploitation: Cashing in on what we know about the user right now • Exploration: Using the interaction as an opportunity to learn more about the user • We need to have informed and optimal strategies to drive that tradeoff • Solution: pick a reasonable set of candidates and show users only “enough” to gather information on them Explore/Exploit
  • 124. Xavier Amatriain – August 2014 – KDD • Given possible strategies/candidates (slot machines) pick the arm that has the maximum potential of being good (minimize regret) • Naive strategy: • Explore with a small probability (e.g. 5%) -> choose an arm at random • Exploit with a high probability (1- ) (e.g. 95%) -> choose the best-known arm so far • Translation to recommender systems • Choose an arm = choose an item/choose an algorithm (MAB testing) Multi-armed Bandits
  • 125. Xavier Amatriain – August 2014 – KDD • Better strategies not only take into account the mean of the posterior, but also the variance • Upper Confidence Bound (UCB) • Show item with maximum score • Score = Posterior mean + • Where can be computed differently depending on the variant of UCB (e.g. to the number of trials or the measured variance) • Thompson Sampling • Given a posterior distribution • Sample on each iteration and choose the action that maximizes the expected reward Multi-armed Bandits
  • 126. Xavier Amatriain – August 2014 – KDD Multi-armed Bandits
  • 127. Xavier Amatriain – August 2014 – KDD Index 1. The Recommender Problem 2. “Traditional” Methods 2.1. Collaborative Filtering 2.2. Content-based Recommendations 2.3. Hybrid Approaches 3. Beyond Traditional Methods 3.1. Learning to Rank 3.2. Similarity 3.3. Deep Learning 3.4. Social Recommendations 3.5. Page Optimization 3.6. Tensor Factorization and Factorization Machines 3.7. MAB Explore/Exploit 4. References
  • 128. Xavier Amatriain – August 2014 – KDD 4. References
  • 129. Xavier Amatriain – August 2014 – KDD ● "Recommender Systems Handbook." Ricci, Francesco, Lior Rokach, Bracha Shapira, and Paul B. Kantor. (2010). ● “Recommender systems: an introduction”. Jannach, Dietmar, et al. Cambridge University Press, 2010. ● “Toward the Next Generation of Recommender Systems: A Survey of the State-of-the-Art and Possible Extensions”. G. Adomavicious and A. Tuzhilin. 2005. IEEE Transactions on Knowledge and Data Engineering, 17 (6) ● “Item-based Collaborative Filtering Recommendation Algorithms”, B. Sarwar et al. 2001. Proceedings of World Wide Web Conference. ● “Lessons from the Netflix Prize Challenge.”. R. M. Bell and Y. Koren. SIGKDD Explor. Newsl., 9(2):75–79, December 2007. ● “Beyond algorithms: An HCI perspective on recommender systems”. K. Swearingen and R. Sinha. In ACM SIGIR 2001 Workshop on Recommender Systems ● “Recommender Systems in E-Commerce”. J. Ben Schafer et al. ACM Conference on Electronic Commerce. 1999- ● “Introduction to Data Mining”, P. Tan et al. Addison Wesley. 2005 References
  • 130. Xavier Amatriain – August 2014 – KDD ● “Evaluating collaborative filtering recommender systems”. J. L. Herlocker, J. A. Konstan, L. G. Terveen, and J. T. Riedl. ACM Trans. Inf. Syst., 22(1): 5–53, 2004. ● “Trust in recommender systems”. J. O’Donovan and B. Smyth. In Proc. of IUI ’05, 2005. ● “Content-based recommendation systems”. M. Pazzani and D. Billsus. In The Adaptive Web, volume 4321. 2007. ● “Fast context-aware recommendations with factorization machines”. S. Rendle, Z. Gantner, C. Freudenthaler, and L. Schmidt-Thieme. In Proc. of the 34th ACM SIGIR, 2011. ● “Restricted Boltzmann machines for collaborative filtering”. R. Salakhutdinov, A. Mnih, and G. E. Hinton.In Proc of ICML ’07, 2007 ● “Learning to rank: From pairwise approach to listwise approach”. Z. Cao and T. Liu. In In Proceedings of the 24th ICML, 2007. ● “Introduction to Data Mining”, P. Tan et al. Addison Wesley. 2005 References
  • 131. Xavier Amatriain – August 2014 – KDD ● D. H. Stern, R. Herbrich, and T. Graepel. “Matchbox: large scale online bayesian recommendations”. In Proc.of the 18th WWW, 2009. ● Koren Y and J. Sill. “OrdRec: an ordinal model for predicting personalized item rating distributions”. In Rec-Sys ’11. ● Y. Koren. “Factorization meets the neighborhood: a multifaceted collaborative filtering model”. In Proceedings of the 14th ACM SIGKDD, 2008. ● Yifan Hu, Y. Koren, and C. Volinsky. “Collaborative Filtering for Implicit Feedback Datasets”. In Proc. Of the 2008 Eighth ICDM ● Y. Shi, A. Karatzoglou, L. Baltrunas, M. Larson, N. Oliver, and A. Hanjalic. “CLiMF: learning to maximize reciprocal rank with collaborative less-is- more filtering”. In Proc. of the sixth Recsys, 2012. ● Y. Shi, A. Karatzoglou, L. Baltrunas, M. Larson,A. Hanjalic, and N. Oliver. “TFMAP: optimizing MAP for top-n context-aware recommendation”. In Proc. Of the 35th SIGIR, 2012. ● C. Burges. 2010. From RankNet to LambdaRank to LambdaMART: An Overview. MSFT Technical Report References
  • 132. Xavier Amatriain – August 2014 – KDD ● A. Karatzoglou, X. Amatriain, L. Baltrunas, and N. Oliver. “Multiverse recommendation: n-dimensional tensor factorization for context-aware collaborative filtering”. In Proc. of the fourth ACM Recsys, 2010. ● S. Rendle, Z. Gantner, C. Freudenthaler, and L. Schmidt-Thieme. “Fast context-aware recommendations with factorization machines”. In Proc. of the 34th ACM SIGIR, 2011. ● S.H. Yang, B. Long, A.J. Smola, H. Zha, and Z. Zheng. “Collaborative competitive filtering: learning recommender using context of user choice. In Proc. of the 34th ACM SIGIR, 2011. ● N. N. Liu, X. Meng, C. Liu, and Q. Yang. “Wisdom of the better few: cold start recommendation via representative based rating elicitation”. In Proc. of RecSys’11, 2011. ● M. Jamali and M. Ester. “Trustwalker: a random walk model for combining trust-based and item-based recommendation”. In Proc. of KDD ’09, 2009. References
  • 133. Xavier Amatriain – August 2014 – KDD ● J. Noel, S. Sanner, K. Tran, P. Christen, L. Xie, E. V. Bonilla, E. Abbasnejad, and N. Della Penna. “New objective functions for social collaborative filtering”. In Proc. of WWW ’12, pages 859–868, 2012. ● X. Yang, H. Steck, Y. Guo, and Y. Liu. “On top-k recommendation using social networks”. In Proc. of RecSys’12, 2012. ● Dmitry Lagun. “Modeling User Attention and Interaction on the Web” 2014 PhD Thesis (Emory Un.) ● “Robust Models of Mouse Movement on Dynamic Web Search Results Pages - F. Diaz et al. In Proc. of CIKM 2013 ● Amr Ahmed et al. “Fair and balanced”. In Proc. WSDM 2013 ● Deepak Agarwal. 2013. Recommending Items to Users: An Explore Exploit Perspective. CIKM ‘13 References
  • 134. Xavier Amatriain – August 2014 – KDD ● Recsys Wiki: http://recsyswiki.com/ ● Recsys conference Webpage: http://recsys.acm.org/ ● Recommender Systems Books Webpage: http://www. recommenderbook.net/ ● Mahout Project: http://mahout.apache.org/ ● MyMediaLite Project: http://www.mymedialite.net/ Online resources
  • 135. Xavier Amatriain – August 2014 – KDD Thanks! Questions? Xavier Amatriain xavier@netflix.com @xamat