A Content-Boosted Collaborative Filtering Approach for Movie Recommendation Based on Local and Global Similarity and Missing Data Prediction


Özbal G., Karaman H., Alpaslan F. N.

COMPUTER JOURNAL, vol.54, pp.1535-1546, 2011 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 54
  • Publication Date: 2011
  • Doi Number: 10.1093/comjnl/bxr001
  • Journal Name: COMPUTER JOURNAL
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.1535-1546
  • Keywords: recommender systems, user modelling, collaborative filtering, sparsity, Pearson correlation coefficient, Floyd-Warshall algorithm, SYSTEMS
  • Middle East Technical University Affiliated: Yes

Abstract

Most traditional recommender systems lack accuracy in the case where data used in the recommendation process is sparse. This study addresses the sparsity problem and aims to get rid of it by means of a content-boosted collaborative filtering approach applied to a web-based movie recommendation system. The main motivation is to investigate whether further success can be obtained by combining 'local and global user similarity' and 'effective missing data prediction' approaches, which were previously introduced and proved to be successful separately. The present work improves these approaches by taking the content information of the movies into account during the item similarity calculations. The comparison of the proposed approach with the original methods was carried out using mean absolute error, and more accurate predictions were achieved.