An dynamic mixed type collaborative recommendation algorithm base on RSS subscribing

In order to solve collaborative filtering recommendation system recommended decline in the quality for the sparse dataset, an dynamic mixed type collaborative recommendation algorithm base on RSS subscribing is presented, which users´ preference items vector based on information classification of their subscribing from RSS Feed is constructed and a users´ comprehensive interest model is built according to interest analysis based on users´ subscribing behavior, reading behavior including the information of users´ reading self-subscribing and recommended subscribing, then the dynamic recommending is done combining content and collaborative filtering . Experiment result shows that the algorithm is better than the traditional collaborative filtering ones in improving the recommendation dependability and accuracy.