A movie cold-start recommendation method optimized similarity measure
Authors:
- 24 November 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 19, 231-234
Abstract
With the explosion of Internet information, recommender system plays an increasingly important position in online video searching. Collaborative filtering technique the most popular recommendation algorithm is inefficient in cold-start scenario. In this paper, we focus on new movie cold start problem and aim to bridge the gap between movie labels and movie similarity. A useful approach is proposed to optimize the movie similarity measure by computing the similarities among directors and actors. We verify the usefulness of our method on MovieLens-1M data sets. The experiments show that our method has a significant improvement on movie cold start problem.Keywords
This publication has 8 references indexed in Scilit:
- ExcUseMe2015
- Wisdom of the better few2011
- Learning Attribute-to-Feature Mappings for Cold-Start Recommendations2010
- Matrix Factorization Techniques for Recommender SystemsComputer, 2009
- Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensionsIEEE Transactions on Knowledge and Data Engineering, 2005
- A collaborative filtering algorithm and evaluation metric that accurately model the user experience2004
- Item-based top-Nrecommendation algorithmsACM Transactions on Information Systems, 2004
- Item-based collaborative filtering recommendation algorithms2001