期刊名称：International Journal of Computer Science and Network
印刷版ISSN：2277-5420
出版年度：2017
卷号：6
期号：5
页码：625-634
出版社：IJCSN publisher
摘要：Recommender system is a tool that provides personalized service to help users to find their desired content. Context-aware
recommendations personalize the search for such desired content by considering the contextual information. In many recommendation
aspects, incorporating the context of users has been shown to improve the quality of recommendations. In this work, we propose a
model that improves the user experience of finding the content they desire by analyzing their contextual information. By conducting an
empirical analysis of a dataset from last.fm, we demonstrate the extraction of latent preferences for recommending items under a given
context and study how contextual information can be exploited to improve the prediction accuracy of recommender systems.
Additionally, we use an optimization function to further minimize the root mean square error (RMSE) measure of the resulted
prediction capability of the latent preference models. Finally, we proposed a latent collaborative preference model to predict the final
rating of users to items by combining the extracted latent preference models. The experimental results achieved in this work shows our
context-aware latent collaborative model can improve the prediction accuracy of recommender systems as compared to state of the art
non-context aware approaches.
