Optimization of LDA text microblogging recommendation algorithm based learning to rank
Yuxiang Xu · 2016
With the recent rise of web3.0 hot, personalized recommendation social networks become an important aspect of research.Social networks on behalf of the domestic microblogging abnormal hot, more and more domestic and foreign research applied over microblogging.Because the characteristics of micro-Bo short text, LDA topic model is more applicable to micro-blog user's interest analysis.Firstly, the use of the network topology, 10 to find a candidate set target users interested users, and then by LDA microblogging users of potential interest analysis to get a point of interest microblogging users, which are interested in the candidate set recommended target users .Experiments show that the method based on TF-IDF with respect to micro-blog content for the word to get user interest method more effective and efficient.