Personalized News Article Recommendation with Novelty Using Collaborative Filtering Based Rough Set Theory

摘要

Online news article reading has become very popular as the World Wide Web provides an access to variety of news articles from large volume of sources around the world. A key challenge of news portals is to provide articles to the users based on their interest. Personalized news recommendation systems provide news articles to the readers based on their interest rather than presenting articles in order of their occurrences. The effectiveness of news recommendation systems reduces due to lack of user ratings and automated novelty detection. A progressive summary helps a user to monitor changes in news items over a period of time. The automatic detection of novelty in personalized news recommendation system could improve a reader's search experience by providing news items that add more information's to already known information's to the users. This paper presents a rough set based collaborative filtering approach to predict a missing news category rating values of a user, and a new novelty detection approach to improve ranking of novel news items. The proposed approach maximizes the accuracy of the news article recommendation to the user according to their interest. Experimental results show the efficiency of the proposed approach.