The Microblog Public Opinion Analysis Based on the SVM and the LDA Model Combining

In view of the LDA model, the superiority of long text clustering to Microblog about the use of the user and time to build a long text. According to implied rich semantic information in text and traditional text clustering makes because of the high dimension calculation results inaccurate faults, the SVM model is given and the LDA model combining the similarity, the last is the use of K - Means algorithm for clustering. Combining the experimental results show that the SVM and the LDA model significantly improves the clustering quality and stability.