A Sparse Image Recommendation Model Using Content and User Preference Information

With the incredibly growing amount of multimedia data uploaded and shared via the social media web sites, recommender systems have become an important necessity to ease users'burden on the information overload. In such a scenario, extensive amount of content information, such as tags, image content and user to item preferences are also available and extremely valuable for making effective recommendations. In this paper, we explore a novel topic model for image recommendation that jointly considers the problem of image content analysis with the users' preference on the basis of sparse representation. Our model is based on the classical probabilistic matrix factorization and can be easily extended to incorporate other useful information such as the social relationship. We evaluate our approach with a newly collected large scale social image data set from Flickr. The experimental results demonstrate that sparse topic modeling of the image content leads to more effective recommendations.