Project : Collaborative Filtering
Motivation
Please refer to wiki/collaborative filtering.
For more detail, you can check this paper:[1].
In briefly, the user-based filtering (ub-filtering)[3] and item-based filtering(ib-filtering)[4] are the widely used techniques for comparing the performance of preference prediction.
Besides, the MovieLen[2] is a widely used dataset is the recommendation society. However, the comparators cannot easy find ub/ib-filtering to redo the experiemnt on the Movielen dataset. Hence, we implement and release ub/ib-filtering for this inconvenience.
Besides, the MovieLen[2] is a widely used dataset is the recommendation society. However, the comparators cannot easy find ub/ib-filtering to redo the experiemnt on the Movielen dataset. Hence, we implement and release ub/ib-filtering for this inconvenience.
Input and Output
Based on the data downloaded from MovieLen,
Reference:
[1] G. Adomavicius and A. Tuzhilin, Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions, TKDE, 2005.
[2] MovieLens Dataset, http://www.grouplens.org/node/73
[3] J. S. Breese and D. Heckerman and C. Kadie, Empirical Analysis of Predictive Algorithms for Collaborative Filtering, UAI,98. [4] B. Sarwar, and G. Karypis and J. Konstan and J. Reidl, Item-based collaborative filtering recommendation algorithms, WWW'01.
[1] G. Adomavicius and A. Tuzhilin, Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions, TKDE, 2005.
[2] MovieLens Dataset, http://www.grouplens.org/node/73
[3] J. S. Breese and D. Heckerman and C. Kadie, Empirical Analysis of Predictive Algorithms for Collaborative Filtering, UAI,98. [4] B. Sarwar, and G. Karypis and J. Konstan and J. Reidl, Item-based collaborative filtering recommendation algorithms, WWW'01.