The document summarizes research on recommender systems in the media industry. It discusses how FD Mediagroup uses recommender systems for their SMART Radio and SMART Journalism products. Key aspects of building a recommender system that FD focuses on include relevance, usefulness, and trust. Relevance is evaluated using metrics like NDCG, MAP, and R-Precision. Usefulness considers both algorithmic goals like diversity and business goals. Trust is evaluated based on whether users engage with the recommender system.
Slidedeck of my lecture at SIKS Course "Advances in Information Retrieval"
Read more here: https://graus.nu/blog/bias-in-recommendations-lecture-siks-course-on-advances-in-ir/
Are you looking for tools to help you run your business that won’t break the bank? This fast-paced session is for you! Learn about free and low-cost tools for productivity, marketing, communications and basic office functions. We’ll cover what the tools can do for you and where to get them. Don’t miss this opportunity to explore new ways to solve common problems with uncommon tools. Assumes a working knowledge of web browsers. Primarily for PC users, although Mac availability will be covered where possible.
How you and your gateway can benefit from the services of the Science Gateway...Katherine Lawrence
January 2017 webinar of the Science Gateways Community Institute. Recording and additional details available at http://sciencegateways.org/upcoming-events/webinars/#previous
This document provides an overview and agenda for a research data network event focusing on research data management. The event aims to discuss latest developments in RDM tools and services, share ideas and practices, and network. Presentations from the event will be shared online. Logistics like wifi access and note sharing are provided. The document also includes an agenda item on business case and costing for RDM that will discuss pain points around tracking costs, cost recovery, and evaluating benefits of RDM. It will share outputs from a Jisc project on developing business cases and costing models for RDM.
Learning in a digital world: trends & implications for learning professionalsscil CH
The document discusses trends and implications in learning in a digital world. It covers topics like digital transformation, microlearning, informal learning, video-based learning, games and gamification, personal learning environments, and emerging topics like immersive learning environments, AI/bots, analytics, and adaptive learning systems. Digital transformation is changing how companies create value and compete through new digital technologies. Microlearning focuses on short, focused learning experiences. Informal learning is a significant part of the 70:20:10 model for learning. Video is increasingly popular for learning due to its production ease and learner appeal. Games and gamification can motivate learners when properly designed. Emerging topics may further transform learning.
Beyond Green - Opportunity Analysis Projectsozamora
The document discusses the formation and initial research of the Beyond Green team. The team's goal was to create an online networking community for innovators, inventors, and environmentalists to collaborate on sustainable development projects. Through surveys, the team found a need for easier access to renewable energy data currently spread across various organizations. The team then refined their value proposition to focus on "Bringing customized data from aggregate renewable energy sources to market." An executive felt this new concept was viable if companies provided data to access other data and revenue was shared for providing data. The next steps are to further develop this concept.
Kristen Schafer
AIA Staff:
Markku Allison, AIA, Resource Architect
AIA Minnesota Staff:
Kate Black, Executive Director
Advisory Committee:
Paolo Tombesi, Chair of Construction, University of Melbourne
Jonathan Cohen, AIA, AIA-CC
Acknowledgements
We would like to thank the project teams who generously
shared their time and experiences for this study:
Cathedral Hill Hospital
MERCY Master Plan Facility Remodel
Lawrence & Schiller Remodel
SpawGlass Austin Regional Office
Edith Green Wendell Wyatt Federal Building Modernization
Autodesk Inc.
Sutter Health Fairfield
Slidedeck of my lecture at SIKS Course "Advances in Information Retrieval"
Read more here: https://graus.nu/blog/bias-in-recommendations-lecture-siks-course-on-advances-in-ir/
Are you looking for tools to help you run your business that won’t break the bank? This fast-paced session is for you! Learn about free and low-cost tools for productivity, marketing, communications and basic office functions. We’ll cover what the tools can do for you and where to get them. Don’t miss this opportunity to explore new ways to solve common problems with uncommon tools. Assumes a working knowledge of web browsers. Primarily for PC users, although Mac availability will be covered where possible.
How you and your gateway can benefit from the services of the Science Gateway...Katherine Lawrence
January 2017 webinar of the Science Gateways Community Institute. Recording and additional details available at http://sciencegateways.org/upcoming-events/webinars/#previous
This document provides an overview and agenda for a research data network event focusing on research data management. The event aims to discuss latest developments in RDM tools and services, share ideas and practices, and network. Presentations from the event will be shared online. Logistics like wifi access and note sharing are provided. The document also includes an agenda item on business case and costing for RDM that will discuss pain points around tracking costs, cost recovery, and evaluating benefits of RDM. It will share outputs from a Jisc project on developing business cases and costing models for RDM.
Learning in a digital world: trends & implications for learning professionalsscil CH
The document discusses trends and implications in learning in a digital world. It covers topics like digital transformation, microlearning, informal learning, video-based learning, games and gamification, personal learning environments, and emerging topics like immersive learning environments, AI/bots, analytics, and adaptive learning systems. Digital transformation is changing how companies create value and compete through new digital technologies. Microlearning focuses on short, focused learning experiences. Informal learning is a significant part of the 70:20:10 model for learning. Video is increasingly popular for learning due to its production ease and learner appeal. Games and gamification can motivate learners when properly designed. Emerging topics may further transform learning.
Beyond Green - Opportunity Analysis Projectsozamora
The document discusses the formation and initial research of the Beyond Green team. The team's goal was to create an online networking community for innovators, inventors, and environmentalists to collaborate on sustainable development projects. Through surveys, the team found a need for easier access to renewable energy data currently spread across various organizations. The team then refined their value proposition to focus on "Bringing customized data from aggregate renewable energy sources to market." An executive felt this new concept was viable if companies provided data to access other data and revenue was shared for providing data. The next steps are to further develop this concept.
Kristen Schafer
AIA Staff:
Markku Allison, AIA, Resource Architect
AIA Minnesota Staff:
Kate Black, Executive Director
Advisory Committee:
Paolo Tombesi, Chair of Construction, University of Melbourne
Jonathan Cohen, AIA, AIA-CC
Acknowledgements
We would like to thank the project teams who generously
shared their time and experiences for this study:
Cathedral Hill Hospital
MERCY Master Plan Facility Remodel
Lawrence & Schiller Remodel
SpawGlass Austin Regional Office
Edith Green Wendell Wyatt Federal Building Modernization
Autodesk Inc.
Sutter Health Fairfield
Philip Bourne presented on the NIH's Big Data to Knowledge (BD2K) initiative and the Associate Director for Data Science (ADDS) office. The goals of BD2K are to use data science to accelerate biomedical research and enhance health outcomes. BD2K supports various centers, projects, and training programs related to data discovery, standards, cloud computing, sustainability, and workforce development. The ADDS office oversees BD2K and aims to establish a sustainable data science ecosystem and well-trained workforce to enable major scientific discoveries through data-driven research.
Cloud Based Language Learning Market Size, Share, & Trends Estimation Report ...subishsam
In cloud-based language learning, resources are kept in the cloud and can be accessed from computers and mobile devices. A cloud-based language learning platform makes it easy for users to get information and lets them share data and work together well. Cloud-based solutions are becoming more popular as the need for e-learning technology grows. Cloud-based solutions are great for a virtual classroom because they make it easy and effective to access information. Most cloud computing providers make cloud servers with C-Sharp and JAVA. Due to the rise of cloud-as-a-service, the global language market is expected to change in a big way over the next few years. Because of this, many businesses have moved to the cloud and now offer language learning through the cloud to their customers.
TargetX held a summit to showcase their Insights product. Insights provides higher education institutions with dashboards and data visualizations using their institutional data. It addresses many of the top IT issues in higher education reported by EDUCAUSE, including being a data-enabled and student-centered institution. An early adopter program provided feedback, including that visualizations help uncover data inconsistencies and comparisons over time help institutions make data-driven decisions. Future plans for Insights include additional content on programs, communications, and benchmarking against peer institutions.
This document discusses strategies for successful cloud projects and recovering from failed cloud projects. It provides three case studies of organizations that struggled with cloud implementations and how they were able to recover. The key lessons discussed are carefully planning cloud projects, selecting the right partners, integrating systems, training users, and establishing governance processes.
Programmatic advertising in china by daxue consultingDaxue Consulting
Programmatic advertising emerged in China in 2012 and has brought great change to China's digital advertising market. It is embracing a huge room for growth since the industry is just in the beginning stage. What is the market size of programmatic advertising? How do different sectors in programmatic advertising work? What are the further trends? All you need to know about China's programmatic advertising are included in the comprehensive report provided by daxue consulting.
Cybersecurity education for the next generationIBM Security
In a world of increasing information security threats, academic initiatives focused on cybersecurity are proliferating – yet, there is still the danger of falling short in addressing the long-term threat.
To avoid becoming too focused on near-term issues, academic programs must be more collaborative across their own institutions, with industry, government and among the global academic community.
Only by working in concert can we meet today’s demand while educating the next generation to create a more secure future.
Learn more about IBM Security: http://ibm.com/security
The document discusses BCG's efforts to reimagine its global talent acquisition strategy through a new employer branding and marketing campaign called "Beyond". It outlines the extensive research and testing done over 9 months with 66,000+ participants across 56 talent markets and 12+ geographies to develop the campaign's positioning, tagline ("Beyond is where we begin"), creative concepts, and templates. It also details the training and enablement approach used to launch the campaign and ensure BCGers could effectively represent the new brand positioning at scale.
Using Cloud Hyperscale Vendors Cognitive Artificial Intelligence NoOps MLaaSBjörn Rodén
Global Cloud Hyperscale Vendors
- The top three (3) for MLaaS (AWS, Azure, GCP)
- What is NoOps, Cognitive AI, and MLaaS
- What is Differentiating Cognitive AI MLaaS NoOps Vendors
Using Cognitive AI MLaaS NoOps services
- Vision Image
- Vision Video
- Audio
- Language & Text
Approach for rapid Business Impact
This document discusses optimizing cloud and multi-cloud environments. It begins with introductions of the speakers, Dennis Nils Drogseth from Enterprise Management Associates and Will Degener from Scalable Software. The agenda then covers predominant challenges in optimizing cloud/multi-cloud such as cloud sprawl and tracking usage across environments. It discusses how effective discovery, dependency mapping, and understanding usage can help address these challenges. Lastly, it suggests that organizational priorities around metrics like costs and benefits will best optimize cloud.
A study from IBM SWG developerWorks and IBM Center for Applied Insights. To gain a current global snapshot of how organizations are using big data and analytics, cloud, mobile, and social technologies, the IBM Center for Applied Insights conducted a survey of over 1200 IT and business decision makers. The study presents adoption and investment data for these emerging areas and also highlights skills and security challenges. Find out how a Pacesetter group exhibits market-driven, analytical, and experimental traits that allow them to capitalize on these emerging tech areas.
RWDG Slides: Stay Non-Invasive in Your Data Governance ApproachDATAVERSITY
There are three distinct approaches to implement Data Governance. The Command-and-Control Approach, the Traditional (if you build it they will come) Approach and the Non-Invasive Data Governance Approach. Some organizations select a single approach for their program while others select to follow a hybrid method.
Bob Seiner will provide information about each approach and indicate how the Non-Invasive Approach can follow the path of least resistance with the greatest success. You may be surprised to learn that many of your present activities can be leveraged to address Stewardship, Metadata, and governed processes – all directed at staying as non-invasive as possible.
In this webinar, Bob will discuss:
- A Data Governance framework completed in a Non-Invasive way
- How the three approaches differ and when to use each
- Sticking to a single approach versus implementing a hybrid model
- How to sell Data Governance as something you are already doing
- Using the Non-Invasive Approach to win friends and influence people.
The document discusses Integrated Project Delivery (IPD), an approach to construction project management that aims to improve collaboration and reduce costs and delays compared to traditional methods. IPD integrates owners, designers, and builders early in the design process through project completion. It aligns risks and rewards, uses mutual agreements, and limits liability among parties. Building Information Modeling (BIM) software and Revit can further foster collaboration when used with IPD. This allows issues to be identified earlier, improving quality, empowering parties, and enabling better budgeting and schedule adherence. Overall, IPD creates more value by allowing earlier risk assessment and adjustments to decrease costs and negative outcomes.
This white paper discusses Integrated Project Delivery (IPD), a new project delivery approach that aims to integrate people, systems, structures, and practices into a collaborative process. IPD seeks to bring together owners, architects, engineers, contractors, and other experts very early in design to improve coordination and reduce waste. The paper distinguishes between "IPDish" approaches that use some IPD tools within traditional contracts, and "Pure IPD" which binds the project team contractually with shared risks and rewards. Key aspects of IPD discussed include early involvement of all expertise, team collaboration using tools like BIM from project start, and multi-party agreements that legally connect team members. The paper examines drivers for IPD adoption from
Case Study #3: Teachers College - Writing SampleChris Klem
Teachers College was using a one-size-fits-all marketing approach that led to confusion among prospective students. They partnered with Earthbound Media Group to develop a targeted campaign using digital printing. EMG conducted research that found students wanted relevant information quickly. They created a system where prospective students fill out an online form selecting their interests, which triggers the automated printing and mailing of a personalized brochure within 7-10 days. The new approach provides customized information to each student and has increased qualified leads and applications to Teachers College while reducing waste compared to their previous mass marketing methods.
Assessing the Personalisation of Australian Google News ResultsAxel Bruns
Paper by Axel Bruns, Abdul Karim Obeid, Daniel Angus, and James Meese, presented at the International Communication Association conference, Paris, 30 May 2022.
Data-centric design and the knowledge graphAlan Morrison
The #knowledgegraph--smart data that can describe your business and its domains--is now eating software. We won't be able to scale AI or other emerging tech without knowledge graphs, because those techs all require a transformed data foundation, large-scale integration, and shared data infrastructure.
Key to knowledge graphs are #semantics, #graphdatabase technology and a Tinker Toy-style approach to adding the missing verbs (which provide connections and context) back into your data. A knowledge graph foundation provides a means of contextualizing business domains, your content and other data, for #AI at scale.
This is from a talk I gave at the Data Centric Design for SMART DATA & CONTENT Enthusiasts meetup on July 31, 2019 at PwC Chicago. Thanks to Mary Yurkovic and Matt Turner for a very fun event!.
PRISM - A Composite Score Model by Bongs LainjoCesToronto
This document presents the Program Indicator Screening Matrix (PRISM), a framework for evaluating programs. PRISM uses a consensus-based approach involving multiple stakeholders to develop a composite score for indicators. Indicators are grouped by theme and screened against criteria like specificity, reliability, and affordability. Scores are discussed within and between groups to establish concordance and select the final indicators. The goals of PRISM are to improve programs by streamlining indicators and establishing causal links between results levels. It provides lessons for effective implementation, including the importance of consensus building, time management, and involving management.
GraphTalk Copenhagen - Killing Data Silos in the Life Sciences with Neo4jNeo4j
This document discusses how clinical research data is often siloed across different systems and standards, making it difficult to integrate and analyze. It proposes using Neo4j, a graph database, to link clinical research data and overcome these silos. Key benefits include being able to trace data for regulatory purposes, maintain a single source of truth, and reduce redundant copies across different systems and standards. The document provides examples of how Neo4j could be used for a study workbench, integrating electronic health records, and mining clinical definitions.
Developing a maturity model for organisational digital capabilityJisc
The document discusses Jisc's development of an organizational digital maturity model to help institutions assess their digital capabilities. It outlines six elements of the model: organizational digital culture; learning, teaching and assessment; research and innovation; ICT infrastructure; content and information; and communication. For each element, it provides principles and indicators of maturity at emerging, established, and enhanced levels. The document also describes Jisc's digital capability resources and services that relate to the model, including a discovery tool and dashboard reports. It concludes by soliciting feedback on the draft model from a community of practice.
Pragmatic ethical and fair AI for data scientistsDavid Graus
1. David Graus presented on pragmatic and fair AI for recruitment and news recommendations.
2. He discussed how algorithms can unintentionally learn and reflect human biases around gender and race. However, AI may also help address these biases, such as through representational ranking in recruitment to achieve demographic parity.
3. Graus also explored using editorial values like diversity, dynamism and serendipity to guide news recommendations, and found their system could increase dynamism without loss of accuracy through constrained intervention.
Philip Bourne presented on the NIH's Big Data to Knowledge (BD2K) initiative and the Associate Director for Data Science (ADDS) office. The goals of BD2K are to use data science to accelerate biomedical research and enhance health outcomes. BD2K supports various centers, projects, and training programs related to data discovery, standards, cloud computing, sustainability, and workforce development. The ADDS office oversees BD2K and aims to establish a sustainable data science ecosystem and well-trained workforce to enable major scientific discoveries through data-driven research.
Cloud Based Language Learning Market Size, Share, & Trends Estimation Report ...subishsam
In cloud-based language learning, resources are kept in the cloud and can be accessed from computers and mobile devices. A cloud-based language learning platform makes it easy for users to get information and lets them share data and work together well. Cloud-based solutions are becoming more popular as the need for e-learning technology grows. Cloud-based solutions are great for a virtual classroom because they make it easy and effective to access information. Most cloud computing providers make cloud servers with C-Sharp and JAVA. Due to the rise of cloud-as-a-service, the global language market is expected to change in a big way over the next few years. Because of this, many businesses have moved to the cloud and now offer language learning through the cloud to their customers.
TargetX held a summit to showcase their Insights product. Insights provides higher education institutions with dashboards and data visualizations using their institutional data. It addresses many of the top IT issues in higher education reported by EDUCAUSE, including being a data-enabled and student-centered institution. An early adopter program provided feedback, including that visualizations help uncover data inconsistencies and comparisons over time help institutions make data-driven decisions. Future plans for Insights include additional content on programs, communications, and benchmarking against peer institutions.
This document discusses strategies for successful cloud projects and recovering from failed cloud projects. It provides three case studies of organizations that struggled with cloud implementations and how they were able to recover. The key lessons discussed are carefully planning cloud projects, selecting the right partners, integrating systems, training users, and establishing governance processes.
Programmatic advertising in china by daxue consultingDaxue Consulting
Programmatic advertising emerged in China in 2012 and has brought great change to China's digital advertising market. It is embracing a huge room for growth since the industry is just in the beginning stage. What is the market size of programmatic advertising? How do different sectors in programmatic advertising work? What are the further trends? All you need to know about China's programmatic advertising are included in the comprehensive report provided by daxue consulting.
Cybersecurity education for the next generationIBM Security
In a world of increasing information security threats, academic initiatives focused on cybersecurity are proliferating – yet, there is still the danger of falling short in addressing the long-term threat.
To avoid becoming too focused on near-term issues, academic programs must be more collaborative across their own institutions, with industry, government and among the global academic community.
Only by working in concert can we meet today’s demand while educating the next generation to create a more secure future.
Learn more about IBM Security: http://ibm.com/security
The document discusses BCG's efforts to reimagine its global talent acquisition strategy through a new employer branding and marketing campaign called "Beyond". It outlines the extensive research and testing done over 9 months with 66,000+ participants across 56 talent markets and 12+ geographies to develop the campaign's positioning, tagline ("Beyond is where we begin"), creative concepts, and templates. It also details the training and enablement approach used to launch the campaign and ensure BCGers could effectively represent the new brand positioning at scale.
Using Cloud Hyperscale Vendors Cognitive Artificial Intelligence NoOps MLaaSBjörn Rodén
Global Cloud Hyperscale Vendors
- The top three (3) for MLaaS (AWS, Azure, GCP)
- What is NoOps, Cognitive AI, and MLaaS
- What is Differentiating Cognitive AI MLaaS NoOps Vendors
Using Cognitive AI MLaaS NoOps services
- Vision Image
- Vision Video
- Audio
- Language & Text
Approach for rapid Business Impact
This document discusses optimizing cloud and multi-cloud environments. It begins with introductions of the speakers, Dennis Nils Drogseth from Enterprise Management Associates and Will Degener from Scalable Software. The agenda then covers predominant challenges in optimizing cloud/multi-cloud such as cloud sprawl and tracking usage across environments. It discusses how effective discovery, dependency mapping, and understanding usage can help address these challenges. Lastly, it suggests that organizational priorities around metrics like costs and benefits will best optimize cloud.
A study from IBM SWG developerWorks and IBM Center for Applied Insights. To gain a current global snapshot of how organizations are using big data and analytics, cloud, mobile, and social technologies, the IBM Center for Applied Insights conducted a survey of over 1200 IT and business decision makers. The study presents adoption and investment data for these emerging areas and also highlights skills and security challenges. Find out how a Pacesetter group exhibits market-driven, analytical, and experimental traits that allow them to capitalize on these emerging tech areas.
RWDG Slides: Stay Non-Invasive in Your Data Governance ApproachDATAVERSITY
There are three distinct approaches to implement Data Governance. The Command-and-Control Approach, the Traditional (if you build it they will come) Approach and the Non-Invasive Data Governance Approach. Some organizations select a single approach for their program while others select to follow a hybrid method.
Bob Seiner will provide information about each approach and indicate how the Non-Invasive Approach can follow the path of least resistance with the greatest success. You may be surprised to learn that many of your present activities can be leveraged to address Stewardship, Metadata, and governed processes – all directed at staying as non-invasive as possible.
In this webinar, Bob will discuss:
- A Data Governance framework completed in a Non-Invasive way
- How the three approaches differ and when to use each
- Sticking to a single approach versus implementing a hybrid model
- How to sell Data Governance as something you are already doing
- Using the Non-Invasive Approach to win friends and influence people.
The document discusses Integrated Project Delivery (IPD), an approach to construction project management that aims to improve collaboration and reduce costs and delays compared to traditional methods. IPD integrates owners, designers, and builders early in the design process through project completion. It aligns risks and rewards, uses mutual agreements, and limits liability among parties. Building Information Modeling (BIM) software and Revit can further foster collaboration when used with IPD. This allows issues to be identified earlier, improving quality, empowering parties, and enabling better budgeting and schedule adherence. Overall, IPD creates more value by allowing earlier risk assessment and adjustments to decrease costs and negative outcomes.
This white paper discusses Integrated Project Delivery (IPD), a new project delivery approach that aims to integrate people, systems, structures, and practices into a collaborative process. IPD seeks to bring together owners, architects, engineers, contractors, and other experts very early in design to improve coordination and reduce waste. The paper distinguishes between "IPDish" approaches that use some IPD tools within traditional contracts, and "Pure IPD" which binds the project team contractually with shared risks and rewards. Key aspects of IPD discussed include early involvement of all expertise, team collaboration using tools like BIM from project start, and multi-party agreements that legally connect team members. The paper examines drivers for IPD adoption from
Case Study #3: Teachers College - Writing SampleChris Klem
Teachers College was using a one-size-fits-all marketing approach that led to confusion among prospective students. They partnered with Earthbound Media Group to develop a targeted campaign using digital printing. EMG conducted research that found students wanted relevant information quickly. They created a system where prospective students fill out an online form selecting their interests, which triggers the automated printing and mailing of a personalized brochure within 7-10 days. The new approach provides customized information to each student and has increased qualified leads and applications to Teachers College while reducing waste compared to their previous mass marketing methods.
Assessing the Personalisation of Australian Google News ResultsAxel Bruns
Paper by Axel Bruns, Abdul Karim Obeid, Daniel Angus, and James Meese, presented at the International Communication Association conference, Paris, 30 May 2022.
Data-centric design and the knowledge graphAlan Morrison
The #knowledgegraph--smart data that can describe your business and its domains--is now eating software. We won't be able to scale AI or other emerging tech without knowledge graphs, because those techs all require a transformed data foundation, large-scale integration, and shared data infrastructure.
Key to knowledge graphs are #semantics, #graphdatabase technology and a Tinker Toy-style approach to adding the missing verbs (which provide connections and context) back into your data. A knowledge graph foundation provides a means of contextualizing business domains, your content and other data, for #AI at scale.
This is from a talk I gave at the Data Centric Design for SMART DATA & CONTENT Enthusiasts meetup on July 31, 2019 at PwC Chicago. Thanks to Mary Yurkovic and Matt Turner for a very fun event!.
PRISM - A Composite Score Model by Bongs LainjoCesToronto
This document presents the Program Indicator Screening Matrix (PRISM), a framework for evaluating programs. PRISM uses a consensus-based approach involving multiple stakeholders to develop a composite score for indicators. Indicators are grouped by theme and screened against criteria like specificity, reliability, and affordability. Scores are discussed within and between groups to establish concordance and select the final indicators. The goals of PRISM are to improve programs by streamlining indicators and establishing causal links between results levels. It provides lessons for effective implementation, including the importance of consensus building, time management, and involving management.
GraphTalk Copenhagen - Killing Data Silos in the Life Sciences with Neo4jNeo4j
This document discusses how clinical research data is often siloed across different systems and standards, making it difficult to integrate and analyze. It proposes using Neo4j, a graph database, to link clinical research data and overcome these silos. Key benefits include being able to trace data for regulatory purposes, maintain a single source of truth, and reduce redundant copies across different systems and standards. The document provides examples of how Neo4j could be used for a study workbench, integrating electronic health records, and mining clinical definitions.
Developing a maturity model for organisational digital capabilityJisc
The document discusses Jisc's development of an organizational digital maturity model to help institutions assess their digital capabilities. It outlines six elements of the model: organizational digital culture; learning, teaching and assessment; research and innovation; ICT infrastructure; content and information; and communication. For each element, it provides principles and indicators of maturity at emerging, established, and enhanced levels. The document also describes Jisc's digital capability resources and services that relate to the model, including a discovery tool and dashboard reports. It concludes by soliciting feedback on the draft model from a community of practice.
Similar to RecSys in the Media Industry: Relevance, Recency, Popularity, and Diversity. (20)
Pragmatic ethical and fair AI for data scientistsDavid Graus
1. David Graus presented on pragmatic and fair AI for recruitment and news recommendations.
2. He discussed how algorithms can unintentionally learn and reflect human biases around gender and race. However, AI may also help address these biases, such as through representational ranking in recruitment to achieve demographic parity.
3. Graus also explored using editorial values like diversity, dynamism and serendipity to guide news recommendations, and found their system could increase dynamism without loss of accuracy through constrained intervention.
Zoeken, vinden, en aanbevelen: personalisatie vs. privacyDavid Graus
Lezing op de VOGIN-IP-lezing op 28 maart 2018 bij de Openbare Bibliotheek Amsterdam.
DISCLAIMER: dit praatje is een mooi stukje ouderwetse (menselijke) manipulatie: expert komt met een 5-tal aanbevelingen :-).
"Tegenwoordig kijkt men steeds vaker met argusogen naar technologiebedrijven die op grote schaal gebruikersgedrag verzamelen. In dit praatje zet ik uiteen waarom het inzetten van gebruikersgedrag van belang is, en hoe het wordt gebruikt om informatie effectief te kunnen ontsluiten en doorzoekbaar maken, of het nu gaat om een zoekmachine als Google, die zich een weg moet banen door een web van miljarden pagina’s, of een service als Spotify, die haar gebruikers graag de juiste muziek blijft aanbieden."
Layman's Talk: Entities of Interest --- Discovery in Digital TracesDavid Graus
The document outlines a program that includes a committee grilling a speaker at 10:00, the committee retreating afterwards, a ceremony at 10:15, and a reception downstairs from 11:00 to 12:30.
Slides of the talk I gave at PyData Amsterdam.
Abstract:
"The FD Mediagroep collects, analyses and filters valuable and relevant information, 24/7, for an influential group of professionals, business executives and high net worth individuals. Company.info (part of FDMG) provides complete, reliable, up-to-date company information and business news about no less than 2.7 million companies and other legal entities in the Netherlands. For Company.info we continuously monitor and crawl hundreds of (online) news sources, resulting in a large archive of (Dutch) business-related news, spanning hundreds of thousands of articles. These articles are automatically enriched, by linking the profiles of companies that are mentioned in the articles, using a custom in-house entity linking framework built in Python. In this talk, I will briefly explain the entity linking task, I will detail the implementation of our custom entity linking framework, and our pipeline for crawling and enriching news articles."
De Macht van Data --- Hoe algoritmen ons leven vormgevenDavid Graus
Slides of the introductory talk I gave at an event at De Balie: "De macht van data" on June 18th, 2017.
For a video recording of the talk see: http://graus.co/blog/mini-college-algoritmen/
Talk I gave at the Data Science Northeast Netherlands Meetup, where I detail the custom in-house entity linking framework, sentiment analysis, and entity salience scoring model we developed for Company.info, in addition to showing some example applications of our corpus of news articles linked to organization profiles.
Dynamic Collective Entity Representations for Entity RankingDavid Graus
This document proposes using collective intelligence to dynamically enrich entity representations from multiple sources like knowledge bases, anchors, tags, and tweets. It presents an adaptive ranking model that learns optimal weights for ranking features like field similarity and importance over time. An experiment on query logs shows expanding entities with different sources improves ranking and retraining the ranker with new content further enhances performance.
Dynamic Collective Entity Representations for Entity RankingDavid Graus
This document proposes using dynamic collective entity representations to improve entity ranking. It describes enriching static entity representations from knowledge bases with descriptions from dynamic sources like tweets, queries, and tags. An adaptive ranking model individually weights each description source and retrains over time using clicks. Experimental results show expanding representations and retraining the ranker improves ranking performance compared to a non-adaptive model, with different sources providing varying benefits depending on their dynamic nature and entity coverage.
David Graus presents his research on using semantic search techniques to improve information retrieval for digital forensic evidence from emails and other electronic documents. He discusses using social network analysis of communication patterns and language models of email content to predict likely recipients of emails. By combining these approaches, he is able to more accurately rank potential recipients than using either technique alone. Future work includes incorporating organizational structure and decay of communication patterns over time.
David Graus - Entity Linking (at SEA), Search Engines Amsterdam, Fri June 27thDavid Graus
David Graus from the University of Amsterdam gave a presentation on entity linking at the Search Engines Amsterdam conference on June 27th. He began by defining entity linking as linking mentions of entities in text to their corresponding entities in a knowledge base. He then gave an example of entity linking and discussed ranking entity candidates based on their prior probabilities like link probability and commonness. Finally, he described using both local and global features in supervised learning models to improve entity linking accuracy.
This document discusses understanding email traffic patterns through recipient recommendation. It explores using social network analysis and language models of email content to predict likely recipients of a given email. Specifically, it examines using measures of node importance in the network, strength of connections between nodes, and similarity between language models of communication profiles to rank and select recipient nodes. The findings indicate that combining social network analysis and language modeling performs better than either approach individually, and that language model similarity is most important for interpersonal communication, while network metrics are more informative for highly active users. Recipient recommendation could help with applications like anomaly detection in e-discovery.
Generating Pseudo-ground Truth for Detecting New Concepts in Social StreamsDavid Graus
The manual curation of knowledge bases is a bottleneck in fast paced domains where new concepts constantly emerge. Identification of nascent concepts is important for improving early entity linking, content interpretation, and recommendation of new content in real-time applications. We present an unsupervised method for generating pseudo-ground truth for training a named entity recognizer to specifically identify entities that will become concepts in a knowledge base in the setting of social streams. We show that our method is able to deal with missing labels, justifying the use of pseudo-ground truth generation in this task. Finally, we show how our method significantly outperforms a lexical-matching baseline, by leveraging strategies for sampling pseudo-ground truth based on entity confidence scores and textual quality of input documents.
yourHistory - entity linking for a personalized timeline of historic eventsDavid Graus
The document describes an entity linking approach to generate a personalized timeline of historic events for a user. It involves 4 main parts: (1) fetching candidate historic events from DBpedia, (2) generating a user profile based on information extracted from the user's Facebook profile, (3) matching the candidate events to the user's interests in their profile, and (4) scoring and ranking the events to produce the final personalized timeline. The approach uses entity linking techniques to associate mentions of entities in the user's profile with the corresponding entries in a knowledge base, in order to identify the user's interests.
This document discusses research on applying text mining and information retrieval techniques for fact finding in regulatory investigations from electronic documents. The researchers are developing methods for semantic search in e-discovery to iteratively retrieve relevant evidence from emails, forums, and other sources by integrating structural context and extracting knowledge from unstructured text. Their current work includes using Twitter mining as a form of conversational search and entity linking to semantically enrich documents.
Semantic Annotation of the Cyttron DatabaseDavid Graus
Final Presentation for my MSc Graduation Project.
Abstract:
"Semantic annotation uses human knowledge formalized in ontologies to enrich texts, by providing structured and machine-understandable information of its content. This paper proposes an approach for automatically annotating texts of the Cyttron Scientific Image Database, using the NCI Thesaurus ontology. Several frequency-based keyword extraction algorithms were implemented and evaluated, aiming to extract important concepts and exclude less relevant ones. Furthermore, topic classification algorithms were applied to identify important concepts which do not occur in the text. The algorithms were evaluated by comparison to annotations provided by experts. Semantic networks were generated from these annotations and an ontology-based similarity metric was applied to perform the comparison. Finally the networks were visualized to provide further insights into the differences of the semantic structure generated by humans, and the algorithms."
More information: http://graus.nu/category/thesis
PyData London 2024: Mistakes were made (Dr. Rebecca Bilbro)Rebecca Bilbro
To honor ten years of PyData London, join Dr. Rebecca Bilbro as she takes us back in time to reflect on a little over ten years working as a data scientist. One of the many renegade PhDs who joined the fledgling field of data science of the 2010's, Rebecca will share lessons learned the hard way, often from watching data science projects go sideways and learning to fix broken things. Through the lens of these canon events, she'll identify some of the anti-patterns and red flags she's learned to steer around.
06-20-2024-AI Camp Meetup-Unstructured Data and Vector DatabasesTimothy Spann
Tech Talk: Unstructured Data and Vector Databases
Speaker: Tim Spann (Zilliz)
Abstract: In this session, I will discuss the unstructured data and the world of vector databases, we will see how they different from traditional databases. In which cases you need one and in which you probably don’t. I will also go over Similarity Search, where do you get vectors from and an example of a Vector Database Architecture. Wrapping up with an overview of Milvus.
Introduction
Unstructured data, vector databases, traditional databases, similarity search
Vectors
Where, What, How, Why Vectors? We’ll cover a Vector Database Architecture
Introducing Milvus
What drives Milvus' Emergence as the most widely adopted vector database
Hi Unstructured Data Friends!
I hope this video had all the unstructured data processing, AI and Vector Database demo you needed for now. If not, there’s a ton more linked below.
My source code is available here
https://github.com/tspannhw/
Let me know in the comments if you liked what you saw, how I can improve and what should I show next? Thanks, hope to see you soon at a Meetup in Princeton, Philadelphia, New York City or here in the Youtube Matrix.
Get Milvused!
https://milvus.io/
Read my Newsletter every week!
https://github.com/tspannhw/FLiPStackWeekly/blob/main/141-10June2024.md
For more cool Unstructured Data, AI and Vector Database videos check out the Milvus vector database videos here
https://www.youtube.com/@MilvusVectorDatabase/videos
Unstructured Data Meetups -
https://www.meetup.com/unstructured-data-meetup-new-york/
https://lu.ma/calendar/manage/cal-VNT79trvj0jS8S7
https://www.meetup.com/pro/unstructureddata/
https://zilliz.com/community/unstructured-data-meetup
https://zilliz.com/event
Twitter/X: https://x.com/milvusio https://x.com/paasdev
LinkedIn: https://www.linkedin.com/company/zilliz/ https://www.linkedin.com/in/timothyspann/
GitHub: https://github.com/milvus-io/milvus https://github.com/tspannhw
Invitation to join Discord: https://discord.com/invite/FjCMmaJng6
Blogs: https://milvusio.medium.com/ https://www.opensourcevectordb.cloud/ https://medium.com/@tspann
https://www.meetup.com/unstructured-data-meetup-new-york/events/301383476/?slug=unstructured-data-meetup-new-york&eventId=301383476
https://www.aicamp.ai/event/eventdetails/W2024062014
We are pleased to share with you the latest VCOSA statistical report on the cotton and yarn industry for the month of May 2024.
Starting from January 2024, the full weekly and monthly reports will only be available for free to VCOSA members. To access the complete weekly report with figures, charts, and detailed analysis of the cotton fiber market in the past week, interested parties are kindly requested to contact VCOSA to subscribe to the newsletter.
RecSys in the Media Industry: Relevance, Recency, Popularity, and Diversity.
1. Recommender Systems in the Media Industry:
Relevance, Recency, Popularity, and Diversity
! David Graus
✉ david.graus@fdmediagroep.nl
🐦 @dvdgrs
2. David Graus / Recsys Summer School / 11/09/2019
whoami !
• 🎓 Academia
• BA Media Studies @ UvA (2008)
• MSc Media Technology @ Universiteit Leiden (2012)
• PhD Information Retrieval @ UvA (2017)
• 🏢 Industry
• Editor radio/online public broadcaster NTR (between BA & MSc)
• Research Intern @ Microsoft Research, US
• Data Scientist @ FD Mediagroep
• Lead Data Scientist @ FD SMART Journalism / BNR SMART Radio
2
3. David Graus / Recsys Summer School / 11/09/2019
In what is to follow…
• An introduction of FD Mediagroep
• Personalization & RecSys at FD Mediagroep:
• SMART Radio
• SMART Journalism
• Lessons learnt building/productizing a RecSys;
• Relevance
• Usefulness
• Trust
3
15. David Graus / Recsys Summer School / 11/09/2019
SMART Radio
• (Transcribe)
• Segment
• Tag
• Serve
15
16. David Graus / Recsys Summer School / 11/09/2019
Transcribe
16
17. David Graus / Recsys Summer School / 11/09/2019
Segment
• Based on metadata,
text, and audio.
17
18. David Graus / Recsys Summer School / 11/09/2019
Tag
• Simple multilabel text
classifier
• Trained on transcripts of
segments + associated tags
from website
18
19. David Graus / Recsys Summer School / 11/09/2019
Serve
• iOS/Android
app
19
31. David Graus / Recsys Summer School / 11/09/2019
Goal
• Help users to discover content [1]
31Meyer, F. Recommender systems in industrial contexts (2012)
34. David Graus / Recsys Summer School / 11/09/2019
Building a RecSys for FD
• Definition of Done:
Our RecSys needs to provide relevant and useful results,
and be trustworthy.
• Relevance: IR evaluation metrics
• Usefulness: What we want our RecSys to achieve
• Trust: Whether we see that people like/use our RecSys
34
35. David Graus / Recsys Summer School / 11/09/2019
1. Relevance
• Using different (ranking) metrics, e.g.;
• NDCG
• MAP
• RPrec
• Evaluating models offline
• Comparing offline vs. online performance
35
36. David Graus / Recsys Summer School / 11/09/2019
Relevance: NDCG
• Item’s score: Gain
• Sum scores: Cumulative Gain
• Penalize scores at lower ranks: Discounted Cumulative Gain
• (e.g., divide CG by log of item position)
• Normalized by “ideal” ranking: NDCG
36http://fastml.com/evaluating-recommender-systems/
37. David Graus / Recsys Summer School / 11/09/2019
MAP
• Precision: TP / TP+FP
• Average Precision: Precision at each rank
• Mean Average Precision: average AP across
queries
37
https://nlp.stanford.edu/IR-book/html/htmledition/evaluation-of-ranked-retrieval-results-1.html
38. David Graus / Recsys Summer School / 11/09/2019
R-Precision
• Precision@k
• where k = R (= number of relevant documents)
38https://nlp.stanford.edu/IR-book/html/htmledition/evaluation-of-ranked-retrieval-results-1.html
39. David Graus / Recsys Summer School / 11/09/2019 https://github.com/cvangysel/pytrec_eval
But…
• It’s all we can do…
39Garcin et al., Offline and Online Evaluation of News Recommender Systems at swissinfo.ch (RecSys 2014)
40. David Graus / Recsys Summer School / 11/09/2019
2. Usefulness
• 🤖 Part algorithmic
• What can we build?
• 💰 Part business/values
• What do we want our RecSys to do?
40
42. David Graus / Recsys Summer School / 11/09/2019
Popularity
• “It is generally not useful to recommend very popular items as they
are generally already known by the user” [1]
• Presentation bias
• Skewed clicks
• Issue in historic data and online evaluation
42[1] Meyer, F. Recommender systems in industrial contexts (2012)
43. David Graus / Recsys Summer School / 11/09/2019
Coverage
• Unleash the long tail?
43
Most read:
Same 6 articles for everyone
Personalized:
178 articles
44. David Graus / Recsys Summer School / 11/09/2019
💰 Business: What do we want/expect?
• ⚠ WIP Study where journalists, data scientists, product,
developers were interviewed to identify shared
perceptions, expectations, attitudes towards
algorithms.
44
45. David Graus / Recsys Summer School / 11/09/2019
Business: perception of algorithms
• Distinguish between audience & and journalism-related values.
• Results TBD, but think of aspects such as:
• broadness vs. personalizedness (audience)
• usability (audience)
• objectivity/neutrality (journalism)
• speed/immediacy (journalism)
• etc.
45
48. David Graus / Recsys Summer School / 11/09/2019
Short story
• Tailor-made RecSys: we are in control
• RecSys as “yet another ranking”
• The slightly longer story…
48
50. David Graus / Recsys Summer School / 11/09/2019
1. Should we worry?
[We] focus on empirical evidence of the spread of personalised
news services and its likely effects on political polarisation and
political information.
50[Zuiderveen Borgesius et al., 2016]
51. David Graus / Recsys Summer School / 11/09/2019
1. Should we worry?
• It’s difficult to bubble yourself
• Both offline:
• “Those who use a lot of partisan information also use an above-average
amount of mainstream news.”
• “[M]ost people by far still get their news via traditional sources, most
notably public-service television.”
• And online:
• “People who choose personalisation are more likely to use an above-average
amount of general-interest news as well.”
• “A recent study suggests that the influence of [the Facebook] algorithm is
lower than the influence of the user’s choices.”
51[Zuiderveen Borgesius et al., 2016]
52. David Graus / Recsys Summer School / 11/09/2019
1. ⚠ Take homes
• “[T]here is no empirical evidence that warrants any strong
worries about filter bubbles.”
• “One lesson we should have learned from the past is that panic
does not lead to sane policies. More evidence is needed on the
process and effects of personalisation, so we can shift the basis of
policy discussions from fear to insight.”
52[Zuiderveen Borgesius et al., 2016]
53. David Graus / Recsys Summer School / 11/09/2019
2. Some empirical evidence…
• On average, 11.7% of results show differences due to
personalization on Google.
• Varies widely by search query and by result ranking.
• Only found measurable personalization as a result of searching
with a logged in account and the IP address of the searching user.
53
54. David Graus / Recsys Summer School / 11/09/2019
2. Method
1. 👤
1. Get 200 volunteers with Google accounts
2. Have them issue the same set of queries
3. Compare results
2. 🤖
1. Construct Google bot accounts
• Vary aspects such as location, demographics, click behavior, browsing + search
history, etc.
2. Have them issue the same set of queries
3. Compare results
54[Hannák et al., 2013]
55. David Graus / Recsys Summer School / 11/09/2019
2. Findings 👤
• On average, 11.7% of results show differences due to
personalization on Google.
• Top ranks tend to be less personalized than bottom ranks.
55[Hannák et al., 2013]
56. David Graus / Recsys Summer School / 11/09/2019
2. Findings 👤
• ✅ A great deal of personalization based on location
(especially for company names, where users received different store locations).
• ❌ The least personalized results tend to be factual and health related queries.
56[Hannák et al., 2013]
57. David Graus / Recsys Summer School / 11/09/2019
2. Findings 🤖
✅ Logged in vs. “cleared cookies” account
✅ Geolocation
❌ Gender
❌ Age
❌ Search history
❌ Click history
❌ Browsing history
57[Hannák et al., 2013]
58. David Graus / Recsys Summer School / 11/09/2019
3. Bursting bubbles
58
59. David Graus / Recsys Summer School / 11/09/2019
3. Aim
Increase exposure to varied political opinions
with a goal of improving civil discourse
59[Yom-Tov et al. 2014]
60. David Graus / Recsys Summer School / 11/09/2019
3. Method
• Classify searchers into political leaning (using geo data)
60[Yom-Tov et al. 2014]
61. David Graus / Recsys Summer School / 11/09/2019
3. Method
• Infer political leaning of news sources from user behavior.
• Identify polarized search queries (with strong political leanings —
in both directions).
61[Yom-Tov et al. 2014]
62. David Graus / Recsys Summer School / 11/09/2019
3. Method
• Treatment group: Insert red results for blue users, and blue
results for red users
• Control group: Do not adjust results
62[Yom-Tov et al. 2014]
63. David Graus / Recsys Summer School / 11/09/2019
3. Methode
1. Short term: Compare clicks/behavior between control &
treatment.
2. Long term: Measure during two weeks, per user;
1. Polarization: Difference of user’s leaning-score compared to
average leaning across all sources.
2. Engagement: Average number of queries + average read
articles.
63
64. David Graus / Recsys Summer School / 11/09/2019
3. Findings 1
• Less clicks on inserted opposing sources.
• But:
“Results pages of the opposing viewpoint which had a similarity
higher than the average tended to be clicked 38% more than those
below the average.”
64[Yom-Tov et al. 2014]
65. David Graus / Recsys Summer School / 11/09/2019
3. Findings 2
• Polarization:
• Treatment: Average leaning ‘moves’ ~25% to centre
• Control: Negligible difference (~1%)
• Engagement:
• Treatment: Number of queries: +9% / articles read: +4%
• Control: Small reduction in both (~2.5%)
65[Yom-Tov et al. 2014]
66. David Graus / Recsys Summer School / 11/09/2019
4. How do algorithmic recommendations 🤖 compare
to human 👤?
66
67. David Graus / Recsys Summer School / 11/09/2019
4. Method
• 🤖 Generate article recommendations for news articles using
different (off the shelf) recommender systems (CF & CB).
• 👤 Compare to hand-picked article recommendations.
• Measure “diversity” of recommended articles:
• At content level
• At tag level
• At category level
• At sentiment/subjectivity level
67[Möller et al. 2018]
68. David Graus / Recsys Summer School / 11/09/2019
4. Findings
“Conventional recommendation algorithms at least preserve the
topic/sentiment diversity of the article supply.”
68[Möller et al. 2018]
69. David Graus / Recsys Summer School / 11/09/2019
4. ⚠ Take home
• Diversity is preserved with conventional recommender systems.
69[Möller et al. 2018]
71. David Graus / Recsys Summer School / 11/09/2019
⚠ Besides
• It is trivial to model and incorporate aspects such as diversity (as
you’ll (hopefully) learn during our hands-on 🙌…)
71
74. David Graus / Recsys Summer School / 11/09/2019
Why fair and transparent?
• User studies have shown:
• Our users want personalized content
• Our users care for transparency
• FD
• Verifiability is one of the core values for FD Mediagroup
• Transparency enables verifiability
74
75. David Graus / Recsys Summer School / 11/09/2019
The Context
75
User-data
Rec engine
Inferred
data
Recomme
ndations
76. David Graus / Recsys Summer School / 11/09/2019
The Context
76
79. David Graus / Recsys Summer School / 11/09/2019
Discourse
• Who is the target audience of the explanations?
• What is the goal?
• What purpose do the explanations serve?
79
81. David Graus / Recsys Summer School / 11/09/2019
Framework
81
82. David Graus / Recsys Summer School / 11/09/2019
Framework
82
I want to
be an expert
I want to
stay informed
I want to
broaden my
horizon
I want to
discover the
unexplored
Values
Broadness, diversity, autonomy, objectivity,
match with the user needs, controllability
89. David Graus / Recsys Summer School / 11/09/2019
User study
89
System
Evaluation
Your goal is to Broaden your Horizons.
There may be topics you do not normally read about,
but you may actually find interesting. Exploring this
helps to build a broad perspective on the issues that
matter to you.
Your goal is to Discover the Unexplored.
There may be topics that you haven’t explored
before that may actually become new interests.
Exploring new topics can promote creativity
and objectivity.
90. David Graus / Recsys Summer School / 11/09/2019
User study
90
Aim:
Study whether being offered an explanation of
reading behavior and a particular goal, would
influence the user’s intended reading
behaviorSystem
Evaluation
91. David Graus / Recsys Summer School / 11/09/2019
User study
91
System
Evaluation
Objective
Pick a persona from four
data-driven profiles
Random assign
goal-order
Explain the goals: Broaden Horizons,
Discover the unexplored
Goal A
Show
visualization
Questionnaire
Persona 1
Goal B
Persona 4
92. David Graus / Recsys Summer School / 11/09/2019
Reading Behavior Explanation
92
Similarity
Familiarity
93. David Graus / Recsys Summer School / 11/09/2019
Hypotheses
93
System
Evaluation
2. Broaden horizon:
Users choose topics that have high similarity and
high familiarity compared to the non-selected topics.
3. Discover the unexplored:
Users choose topics that have low similarity and
low familiarity compared to the non-selected topics.
1. Goal Framework:
Broaden Horizon chosen topics are more similar and
more familiar than Discover the unexplored
95. David Graus / Recsys Summer School / 11/09/2019
Results
95
System
Evaluation
2. Broaden horizon:
Users choose topics that have high similarity and
high familiarity compared to the non-selected topics.
3. Discover the unexplored:
Users choose topics that have low similarity and
low familiarity compared to the non-selected topics.
1. Goal Framework:
Broaden Horizon chosen topics are more similar and
more familiar than Discover the unexplored
96. David Graus / Recsys Summer School / 11/09/2019
Results
96
System
Evaluation
2. Broaden horizon:
Users choose topics that have high similarity and
high familiarity compared to the non-selected topics.
3. Discover the unexplored:
Users choose topics that have low similarity and
low familiarity compared to the non-selected topics.
1. Goal Framework:
Broaden Horizon chosen topics are more similar and
more familiar than Discover the unexplored
97. David Graus / Recsys Summer School / 11/09/2019
Results
97
System
Evaluation
2. Broaden horizon:
Users choose topics that have high similarity and
high familiarity compared to the non-selected topics.
3. Discover the unexplored:
Users choose topics that have low similarity and
low familiarity compared to the non-selected topics.
1. Goal Framework:
Broaden Horizon chosen topics are more similar and
more familiar than Discover the unexplored
98. David Graus / Recsys Summer School / 11/09/2019
Results
98
System
Evaluation
2. Broaden horizon:
Users choose topics that have high similarity and
high familiarity compared to the non-selected topics.
3. Discover the unexplored:
Users choose topics that have low similarity and
low familiarity compared to the non-selected topics.
1. Goal Framework:
Broaden Horizon chosen topics are more similar and
more familiar than Discover the unexplored
99. David Graus / Recsys Summer School / 11/09/2019
Results
99
System
Evaluation
2. Broaden horizon:
Users choose topics that have high similarity and
high familiarity compared to the non-selected topics.
3. Discover the unexplored:
Users choose topics that have low similarity and
low familiarity compared to the non-selected topics.
1. Goal Framework:
Broaden Horizon chosen topics are more similar and
more familiar than Discover the unexplored
100. David Graus / Recsys Summer School / 11/09/2019
Results
100
System
Evaluation
2. Broaden horizon:
Users choose topics that have high similarity and
high familiarity compared to the non-selected topics.
3. Discover the unexplored:
Users choose topics that have low similarity and
low familiarity compared to the non-selected topics.
1. Goal Framework:
Broaden Horizon chosen topics are more similar and
more familiar than Discover the unexplored
101. David Graus / Recsys Summer School / 11/09/2019
Results
101
System
Evaluation
2. Broaden horizon:
Users choose topics that have high similarity and
high familiarity compared to the non-selected topics.
3. Discover the unexplored:
Users choose topics that have low similarity and
low familiarity compared to the non-selected topics.
1. Goal Framework:
Broaden Horizon chosen topics are more similar and
more familiar than Discover the unexplored
Partial
support
102. David Graus / Recsys Summer School / 11/09/2019
What have we done
• Novel, scalable and generalizable framework of user-profile
explanations
• Exploration of reader data; domain-knowledge and
data-driven ontology
• Interface mockup
• User-study to evaluate the value-driven explanations
• “Explainability” as a product
102
103. David Graus / Recsys Summer School / 11/09/2019
Future
• Further designing
implementation of our
framework
• Formalize the
epistemic goals
together with editors
• Extend user studies/
focus groups to all
value-driven goals
!103
104. David Graus / Recsys Summer School / 11/09/2019
In summary
• Shown you where/how we personalize @ FD Mediagroep
• Shown you how we evaluate and can determine usefulness of
RecSys using IR metrics
• Told you that these metrics are not all there is to it
• Told you it is not so difficult to incorporate diversity, discount for
popularity, etc.
• What we’ll work on after the break
104