A Personalized Recommendation Approach Based on Community Tag Cloud in Social Tagging System
Zhen Zhang · Journal of Intelligence · 2011
In social tagging systems,tags are results of users' annotations.The appearance of the tags changes the two way user-item matrices into three dimensional user-tag-item,which brings the recommender systems development chances into a new direction.Recommender systems have been utilizing tags for identifying similar resources and generate personalized recommendations.In this paper we propose a personalized recommendation approach based on community tag clouds.The proposed approach is realized in two aspects.Firstly,we cluster items into communities based on analyzing the tags annotated by users,at the same time,we can also cluster users into communities based on analyzing the tags they have utilized,and then we give every community a tag cloud.Secondly,we calculate the similarity between the item's community tag cloud and user's community tag cloud to make personalized recommendation.The results of empirical evaluation indicate the superiority of the approach in improving the precision and diversity of the recommendations.