Rumors Detection in Sina Weibo Based on Text and User Characteristics

M414 such as Twitter and Sina Weibo have brought convenience to people in recent years, but there are a large number of internet rumors bring serious harm to us. Because of the fuzziness and concealment of rumors, the automatic identification of microblog rumors is very difficult, which brings great challenges to microblog's supervision. In this paper, for the text information of rumors, we construct a new emotional classification method, and use this method to analyze the emotional orientation of comments, and then calculate the proportion of negative emotional comments. Combined with user information at the same time, we analyze retweet-list and comment-list, and then judge whether there are authoritative users who participated in the spread of rumors. Moreover, we extract features from the information of user and calculate the user's reputation value. Finally, we use the training classifier to detect microblog rumors. Experimental results illustrate the efficacy and efficiency of the proposed new features in microblog rumors detection.