Reading 7 (Sep 9)

"Collaborative Filtering" and Similarity Metrics

Goal

Learn more about an implementation that leverages similarity metrics within the context of "collaborative filtering." Witness an application of euclidean distance, pitfalls of relying on euclidean distance and how to use correlation to represent similarity.

Instructions

Your assignment is to read the following items and be prepared to discuss them in class.

• Read CI chapter 2, pages 7 - 28, Making Recommendations.

What is "collaborative filtering"? How can you use traditional similarity metrics to represent similarity?

Grading criteria

• 1 Participation Point: You must participate in our discussion on Monday, Sep 12.