Recommender systems apply intelligent access technologies to large information systems. These systems, especially collaborative filtering based ones, are achieving widespread success on the Web. In recent years, the amount of available information and the number of visitors to Web sites are increasing enormously. New recommender system technologies are needed that can quickly produce high quality recommendations, even for very large-scale information resources. In this paper we apply inductive learning algorithm to the recommendation process. Instead of computing user-user or item-item similarities, we construct a decision tree to represent user preference. Recommendations are performed by decision tree classification. To inspect the effectiveness of this technology, we set up a movie recommender system based on inductive learning and make online experiments for evaluation. Our results suggest that inductive-learning-based technology is promising for the solution of the very large-scale problems and high-quality recommendations can be expected.
[1]
Alberto Maria Segre,et al.
Programs for Machine Learning
,
1994
.
[2]
John Riedl,et al.
Item-based collaborative filtering recommendation algorithms
,
2001,
WWW '01.
[3]
Yoav Shoham,et al.
Fab: content-based, collaborative recommendation
,
1997,
CACM.
[4]
David Heckerman,et al.
Empirical Analysis of Predictive Algorithms for Collaborative Filtering
,
1998,
UAI.
[5]
Naoki Abe,et al.
Query Learning Strategies Using Boosting and Bagging
,
1998,
ICML.
[6]
Kirsten Swearingen,et al.
Beyond Algorithms: An HCI Perspective on Recommender Systems
,
2001
.
[7]
Eric Horvitz,et al.
Collaborative Filtering by Personality Diagnosis: A Hybrid Memory and Model-Based Approach
,
2000,
UAI.