Lect 13 DM

DISCLAIMER
In preparation of these slides, materials have been taken from

different online sources in the shape of books, websites, research
papers and presentations etc. However, the author does not have
any intention to take any benefit of these in her/his own name. This
lecture (audio, video, slides etc) is prepared and delivered only for
educational purposes and is not intended to infringe upon the
copyrighted material. Sources have been acknowledged where
applicable. The views expressed are presenter’s alone and do not
necessarily represent actual author(s) or the institution.
Recommender Systems
Recommender Systems
 Recommender Systems (RS) help to match users with items

 ease information overload
 sales assistance (guidance, advisory…)
 Recommender Systems can be seen as a function

 Given:
- User model (e.g. ratings, preferences, demographics,
situational context)
- Items (with or without description of item characteristics)
 Find:
- Relevance score. Used for determining Top-K items.
 Concrete system design depends

 on availability of exploitable data
 implicit and explicit user feedback
 domain characteristics
Application Domains of Recommender Systems
 Which digital camera should I buy?

 Which movie should I rent?
 Which web sites will I find interesting?
 Which book should I buy for my next vacation?
 Which is the best investment for my retirement money?
Universität Mannheim – Bizer: Web Usage Mining – FFS2015 (Version: 19.2.2015) – Slide 24
Paradigms of Recommender Systems
Collaborative: "Tell me what's popular

among my peers"
User–Item Rating Matrix

Item1 Item2
Alice 5 ?
User1 3 1
User2 4 3
Content-based: "Show me more of the

same what I've liked"
Hybrid: Combinations of various inputs

and/or composition of different
mechanisms
When does a Recommender do a good Job?
1. User’s Perspective
 Recommend me items that I like and did not know about.
 Accident of finding something good
while not specifically searching for it.
2. Merchant’s Perspective
 Increase sales of high-revenue items.
Collaborative Filtering
 The most prominent approach to generate recommendations

 used by large e-commerce sites
 applicable in many domains (book, movies, DVDs, ..)
 Approach
 use the "wisdom of the crowd" to recommend items
 Basic Assumptions
 Users give ratings to catalog items (implicitly or explicitly)
 Customers who had similar tastes in the past, will have similar
tastes in the future
 Input: Matrix of given user–item ratings

 Output types
1. (Numerical) prediction indicating to what degree the current user will
like or dislike a certain item
2. Top-K list of recommended items
User-Based Nearest-Neighbor Collaborative Filtering
User-Based Nearest-Neighbor Collaborative Filtering
 Some first questions

1. How do we measure user similarity?
2. How many neighbors should we consider?
3. How do we generate a prediction from the neighbors' ratings?
Item1 Item2 Item3 Item4 Item5

Alice 5 3 4 4 ?
User1 3 1 2 3 3
User2 4 3 4 3 5
User3 3 3 1 5 4
User4 1 5 5 2 1
Measuring User Similarity
Example: Pearson Correlation
Making Predictions
Memory-based and Model-based Approaches
 User-based CF is said to be "memory-based"

 the rating matrix is directly used to find neighbors / make predictions
 does not scale for most real-world scenarios as large e-commerce
sites have tens of millions of customers and 10,000s of items
 Model-based approaches
 based on an offline pre-processing or "model-learning" phase
 at run-time, only the learned model is used to make predictions
 models are updated / re-trained periodically
 A large variety of techniques is used
1. Item-based Collaborative Filtering
2. Using Association Rules for Recommendations
3. Probabilistic Methods
4. Matrix Factorization Techniques
Item-based Collaborative Filtering
Calculating Item-to-Item Similarity
Making Predictions
End of Lecture
Lecture Slides Taken From:

Tutorial: Recommender System
International Joint Conference on Artificial Intelligence
Dietmar Jannach, Gerhard Friedrich
Association Rules
https://www.javatpoint.com/apriori-algorithm-in-machine-learning

Lect 13 DM

Uploaded by

Copyright:

Available Formats

Lect 13 DM

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lect 13 DM

Uploaded by

Copyright:

Available Formats

DISCLAIMER

In preparation of these slides, materials have been taken from

 Recommender Systems (RS) help to match users with items

 Recommender Systems can be seen as a function

 Concrete system design depends

 Which digital camera should I buy?

Collaborative: "Tell me what's popular

User–Item Rating Matrix

Content-based: "Show me more of the

Hybrid: Combinations of various inputs

 The most prominent approach to generate recommendations

 Input: Matrix of given user–item ratings

 Some first questions

Item1 Item2 Item3 Item4 Item5

 User-based CF is said to be "memory-based"

Lecture Slides Taken From:

You might also like