Book Recommendation: Improving Collaborative Filtering with Content Information

Collaborative filtering is a product recommendation technique for making automated product suggestions to a user based on the preference information from similar users. Traditional recommendation algorithms drive personalized recommendations using the data from user purchases and ratings. For e-book retailers, besides user purchases and ratings, product features such as book content and metadata also provide valuable information that can be used 'to improve the recommendation precision and recall. The goal of this project is to improve the recommendation algorithm in the e-book domain by incorporating additional user. and product features. Research topics include the application of collaborative filtering, topic modeling and collaborative topic regression as well as experimenting with novel approaches.

Faculty Supervisor:

Dr. Richard Zemel

Student:

Qingwei Ge

Partner:

Kobo Inc.

Discipline:

Computer science

Sector:

Information and communications technologies

University:

University of Toronto