A Light Recommendation Algorithm of We-Media Articles Based on Content

Abstract

Since the explosive growth of we-medias today, personalized recommendation is playing an increasingly important role to help users to find their target articles in vast amounts of data. Deep learning, on the other hand, has shown good results in image processing, computer vision, natural language processing, and other fields. But it's a relative blank in the application of we-media articles recommendation. Combining the new features of we-media articles, this paper puts forward a recommendation algorithm of we-media articles based on topic model, Latent Dirichlet Allocation (LDA), and deep learning algorithm, Recurrent Neural Networks (RNNs). Experiments on the real datasets show that the combined method outperforms the traditional collaborative filtering recommendation and non-personalized recommendation method.