The Reserch And Implementation About Weibo User Interset Detection And User Intent Recognition

Micro-blog as a representative of the social networking platform in daily life is becoming more and more active . The massive user generated micro-blog content in real time brings challenges to users. Digging personal hobbies and interests to provide users with a better and more accurate use contents. Identify the behavioral intentions of micro-blog users to provide more precise guidance. These two issues are the major problems in the research of micro-blog. This paper focus on these two problems, and proposes some targeted solutions.The main problem of micro-blog user interest discovery is the short text feature of micro-blog text. Traditional text processing methods have some limitations in the lack of context features and micro-blog context dependence. In this paper, we improved the method on the basis of the traditional LDA topic model feature. the text feature in text-topic distribution is introduced into the text feature space, which further expands the characteristics of micro-blog. In order to further improve the performance of the classification algorithm , morevore, we can identify the polysemous words in micro-blog and eliminate the influence of polysemy. The validity of the method is verified by experiments.Micro-blog user intention recognition is a new research field, this study extracts the potential intention of micro-blog firstly, according to the intention of micro-blog features, recognizing the intented micro-blog which contains intent indicated word. Then we classify the intention of micro-blog , according to the graph propagation model. The validity of the method is verified by experiments.