- 1 -
Book Recommended
Systems
- 2 -
What is Recommended Systems
- 3 -
© Dietmar Jannach, Markus Zanker and Gerhard Friedrich
- 4 -
© Dietmar Jannach, Markus Zanker and Gerhard Friedrich
Recommender Systems
Application areas
- 5 -
© Dietmar Jannach, Markus Zanker and Gerhard Friedrich
In the Social Web
- 6 -
What are recommender systems for?
l Introduction
How do they work (Part I) ?
l Collaborative Filtering
How to measure their success?
l Evaluation techniques
How do they work (Part II) ?
l Content-based Filtering
l Knowledge-Based Recommendations
l Hybridization Strategies
Advanced topics
l Explanations
l Human decision making
Agenda
- 7 -
Introduction
- 8 -
Why using Recommender Systems?
Value for the customer
l Find things that are interesting
l Narrow down the set of choices
l Help me explore the space of options
l Discover new things
l Entertainment
l …
Value for the provider
l Additional and probably unique personalized service for
the customer
l Increase trust and customer loyalty
l Increase sales, click trough rates, conversion etc.
l Opportunities for promotion
l Obtain more knowledge about customers
l …
- 9 -
Real-world check
Myths from industry
l Amazon.com generates 30-40 percent of their sales through the recommendation
lists
l Netflix (DVD rental and movie streaming) generates 20-25 percent of their sales
through the recommendation lists
There must be some value in it
l See recommendation of groups, jobs or people on LinkedIn
l Friend recommendation and ad personalization on Facebook
l Song recommendation at last.fm
l News recommendation at Forbes.com (plus 37% CTR)
- 10 -
Recommender systems
Recommender systems reduce information overload by estimating relevance
- 11 -
Types of Recommender Systems?
l
Personal recommender systems
l
Collaborative recommender systems
l
Content-based recommender systems
l
Knowledge-based recommender systems
l
Hybrid recommender systems
- 12 -
Paradigms of recommender systems
- 13 -
Collaborative recommender systems
Collaborative: "Tell me what's popular
among my peers"
- 14 -
Content-based recommender systems
Content-based: "Show me more of the
same what I've liked"
- 15 -
Knowledge-based recommender systems
Knowledge-based: "Tell me what fits
based on my needs"
- 16 -
Hybrid recommender systems
Hybrid: combinations of various inputs
and/or composition of different
mechanism
- 17 -
Evaluation of Recommender Systems
- 18 -
What is a good recommendation?
Total sales numbers
Promotion of certain items
…
Click-through-rates
Interactivity on platform
…
Customer return rates
Customer satisfaction and loyalty
What are the measures in practice?
- 19 -
Group B : Assignments No. 7
l
Developing an book recommend-er ( a book that the
reader should read and is new) Expert system
l
Flow of application :
l
Top Rated and New added book will
recormanded first.
l
Functionality provide for add new book
l
Database: SQLite
l
Control Used :

Listview: To show list on screen

Spinner: To show drop down on
screen
- 20 -
Screens
l
Top rated books
Recommender
by system
- 21 -
Screens
l
Click on Book &
make
Recommendation
i.e give rating
- 22 -
Screens
l
Add New book.
- 23 -
Thank you for your attention!
http://recsys.acm.org
http://www.recommenderbook.net

Book Recommender System using python .pptx

  • 1.
    - 1 - BookRecommended Systems
  • 2.
    - 2 - Whatis Recommended Systems
  • 3.
    - 3 - ©Dietmar Jannach, Markus Zanker and Gerhard Friedrich
  • 4.
    - 4 - ©Dietmar Jannach, Markus Zanker and Gerhard Friedrich Recommender Systems Application areas
  • 5.
    - 5 - ©Dietmar Jannach, Markus Zanker and Gerhard Friedrich In the Social Web
  • 6.
    - 6 - Whatare recommender systems for? l Introduction How do they work (Part I) ? l Collaborative Filtering How to measure their success? l Evaluation techniques How do they work (Part II) ? l Content-based Filtering l Knowledge-Based Recommendations l Hybridization Strategies Advanced topics l Explanations l Human decision making Agenda
  • 7.
  • 8.
    - 8 - Whyusing Recommender Systems? Value for the customer l Find things that are interesting l Narrow down the set of choices l Help me explore the space of options l Discover new things l Entertainment l … Value for the provider l Additional and probably unique personalized service for the customer l Increase trust and customer loyalty l Increase sales, click trough rates, conversion etc. l Opportunities for promotion l Obtain more knowledge about customers l …
  • 9.
    - 9 - Real-worldcheck Myths from industry l Amazon.com generates 30-40 percent of their sales through the recommendation lists l Netflix (DVD rental and movie streaming) generates 20-25 percent of their sales through the recommendation lists There must be some value in it l See recommendation of groups, jobs or people on LinkedIn l Friend recommendation and ad personalization on Facebook l Song recommendation at last.fm l News recommendation at Forbes.com (plus 37% CTR)
  • 10.
    - 10 - Recommendersystems Recommender systems reduce information overload by estimating relevance
  • 11.
    - 11 - Typesof Recommender Systems? l Personal recommender systems l Collaborative recommender systems l Content-based recommender systems l Knowledge-based recommender systems l Hybrid recommender systems
  • 12.
    - 12 - Paradigmsof recommender systems
  • 13.
    - 13 - Collaborativerecommender systems Collaborative: "Tell me what's popular among my peers"
  • 14.
    - 14 - Content-basedrecommender systems Content-based: "Show me more of the same what I've liked"
  • 15.
    - 15 - Knowledge-basedrecommender systems Knowledge-based: "Tell me what fits based on my needs"
  • 16.
    - 16 - Hybridrecommender systems Hybrid: combinations of various inputs and/or composition of different mechanism
  • 17.
    - 17 - Evaluationof Recommender Systems
  • 18.
    - 18 - Whatis a good recommendation? Total sales numbers Promotion of certain items … Click-through-rates Interactivity on platform … Customer return rates Customer satisfaction and loyalty What are the measures in practice?
  • 19.
    - 19 - GroupB : Assignments No. 7 l Developing an book recommend-er ( a book that the reader should read and is new) Expert system l Flow of application : l Top Rated and New added book will recormanded first. l Functionality provide for add new book l Database: SQLite l Control Used :  Listview: To show list on screen  Spinner: To show drop down on screen
  • 20.
    - 20 - Screens l Toprated books Recommender by system
  • 21.
    - 21 - Screens l Clickon Book & make Recommendation i.e give rating
  • 22.
  • 23.
    - 23 - Thankyou for your attention! http://recsys.acm.org http://www.recommenderbook.net