A personalized multimedia retrieval frame based on user interest

This paper describes a personalized multimedia retrieval frame based on user interest through the multi-model cross-mutual knowledge base. The most important feature of the frame is to integrate the multi- model data into a seamless retrieval system. It calculates both the semantic and content level similarity between media object and the query condition. And, using user interest model to push interesting information and sort the inquiry results, so that users can easily and quickly find interesting objects.