Abstract

Movie recommender systems attempt to find the movies which might be of interest for their users. However, as new movies are released on a daily basis, and new users join movie recommendation services, the problem of recommending becomes increasingly harder. In this paper, we present a simple way of using a priori data about movies in order to improve the accuracy of collaborative filtering recommender systems. The approach decreases the sparsity of the rating matrix by inferring personal ratings on tags assigned to movies. These new tag ratings are then used to find the most appropriate movies to recommend. Experiments performed on data from the movie recommendation community Moviepilot show a positive effect on the quality of recommended items.

@inproceedings{Said:2011:PTF:2039320.2039328,
 author = {Said, Alan and Kille, Benjamin and De Luca, Ernesto W. and Albayrak, Sahin},
 title = {Personalizing tags: a folksonomy-like approach for recommending movies},
 booktitle = {Proceedings of the 2nd International Workshop on Information Heterogeneity and Fusion in Recommender Systems},
 series = {HetRec '11},
 year = {2011},
 isbn = {978-1-4503-1027-7},
 location = {Chicago, Illinois},
 pages = {53--56},
 numpages = {4},
 url = {http://doi.acm.org/10.1145/2039320.2039328},
 doi = {10.1145/2039320.2039328},
 acmid = {2039328},
 publisher = {ACM},
 address = {New York, NY, USA},
 keywords = {folksonomy, movie recommendation, recommender system},
}