Today collaborative filtering is the most successful recommender system technology. However, in traditional collaborative filtering algorithms, users’ interest is considered to be static. That means, in these algorithms, ratings produced at different times are weighted equally, and changes in user purchase interest are not taken into consideration. For this reason, the system may recommend unsatisfactory items when users’ interest has changed. To solve this problem, the time factor has been brought into collaborative filtering. In new algorithms, we have divided users’ rating history into several periods, analyzed users’ interest distribution in these periods and quantize every user’s interest. At the same time, we find user’s recent interest by setting a time window. With these two technologies, we propose a collaborative algorithm time period partition named TPPCF. Experiments have shown that our new algorithm TPPCF substantially improves the precision of item-based collaborative filtering.
[1]
John Riedl,et al.
Item-based collaborative filtering recommendation algorithms
,
2001,
WWW '01.
[2]
Xing Chunxiao,et al.
A Collaborative Filtering Recommendation Algorithm Incorporated with User Interest Change
,
2007
.
[3]
Richi Nayak,et al.
Improving Recommendation Novelty Based on Topic Taxonomy
,
2007,
2007 IEEE/WIC/ACM International Conferences on Web Intelligence and Intelligent Agent Technology - Workshops.
[4]
Maria E. Orlowska,et al.
Recency-based collaborative filtering
,
2006,
ADC.
[5]
Greg Linden,et al.
Amazon . com Recommendations Item-to-Item Collaborative Filtering
,
2001
.
[6]
Zeng Chun,et al.
A Survey of Personalization Technology
,
2002
.
[7]
Xue Li,et al.
Time weight collaborative filtering
,
2005,
CIKM '05.