Effective Personalized Recommendation in Collaborative Tagging Systems

摘要

Recently, collaborative tagging systems have attracted more and moreattention and have been widely applied in web systems. Tags provide highlyabstracted information about personal preferences and item content, and aretherefore potential to help in improving better personalized recommendations.In this paper, we propose a tag-based recommendation algorithm considering thepersonal vocabulary and evaluate it in a real-world dataset: Del.icio.us.Experimental results demonstrate that the usage of tag information cansignificantly improve the accuracy of personalized recommendations.