Collaborative Filtering

1
Collaborative Filtering

• Yang Song
• Computer Science Engineering
• The Penn State University

2
Definition

• Collaborative Filtering
• Step 1 Collaborative
• Step 2 Filtering

3
Definition (cont.)

• Recommend Similar Items to Users
• Ask questions
• Match like / dislike with other users
• Give proper answers to questions
• What kind of car should I buy?
• What web-pages would I find most interesting?
• What people in my company would be best assigned
  to a particular project team?

4
(No Transcript)
5
Introduction

• Formal Definition
• N of users
• M total of items in the database
• S N X M matrix of user ratings
• m rating scale (e.g., 0,1,,5)
• Sui is the rating given by user u to item i.

6
Introduction

• An Example
• 5 users (N) ratings of 10 movies (M)
• Ratings scale is 0-5 (m)

7
Introduction to Collaborative Filtering

• Objective
• Predict those Sui 0 in the matrix
• E.g., what is S11, S15, S19 in the previous
  example?

8
Issues

• Issues in CF
• Produce high quality recommendations
• Very Large Databases (VLDB)
• Millions of customers and products
• E.g., EachMovie dataset
• 1,623 movies (M)
• 5,000 users (N)
• 381,862 ratings (S)
• Sparse matrix
• Computational Cost / Speed
• Keep privacy