Tag recommendation system for multimedia retrieval in mobile Environment

Nowadays, due to the explosive growth of mobile devices, social networks and multimedia contents sharing systems have become increasingly popular. However, the characteristics of mobile devices and tags make it difficult to perform an accurate tag-based multimedia retrieval. In this paper, we propose a novel tag recommendation system for multimedia retrieval based on the user trust model which takes advantage of tag clusters of Flickr and visual similarity of multimedia. The proposed system makes it possible to provide the reliable and relevant tags to user in mobile environment.