Twitter Users Crediblity Evaluation Based on Social Graph Impression

Abstract

In the era of social media, information became abundant, versatile and easily accessible. Such characteristics enable almost anyone to publish and share their opinion about a variety of subjects. Such freedom comes at a cost since information and source credibility are in jeopardy. Therefore, numerous solutions were proposed to evaluate source credibility. In this paper, we review previous efforts on twitter credibility evaluation. We also propose an approach that relies on the impression of user social network on his tweets to determine his credibility. Furthermore, we incorporate a feature that determines the credibility of the URLs shared by the user through his tweets. The evaluation of our approach has shown it outperforms the baseline in both precision and recall.