The Credibility Evaluation Of Microblogging Information Based On High Impact News Events

Twitter emerged as a popular platform both for conversation and substantive real-time events at an extraordinary rate, enabling its users to distribute, disseminate and obtain status messages. Depending on the diversity and fragmentation of information and converged fission mode of transmission, Twitter greatly optimizes the media effects of the news events, with false information or even rumors growing explosively, which increasingly importance has been attached to.Basing on previous studies on Twitter events over the information credibility, this paper comprehensively analyze and summarize three basic elements affecting the information credibility, as Twitter users, propagation and message content. Furthermore, we analyze various sub-elements and design corresponding assumption indexes. Additionally, we design questionnaire and accomplished hypothesis testing to obtain high impact factors based on assumption indexes, which includes quantification and assessment of parameter indexes, reliability and validity examination, parameter correction after parameter assessment. Ultimately, we construct credibility assessment model. This paper employs Combination Weighting Approach, which co-ordinates Factor Analysis based on Varimax orthogonal rotation, Correlation coefficient method and AHP to calculate the relative weight between each parameter index and verify the correlation parameters. Furthermore, we adopt Grey Evaluation Model which is based on AHP to calculate and assess the impact of each parameter index on the events information credibility on Twitter.As a result, the experimental results demonstrate that the three basic elements, as Twitter users, propagation and message content, have a great impact on the credibility assessment of events, as message content affecting the greatest. Furthermore, such impact factors as Twitter information which contains external links, semantic or non-semantic emotional notation, user fame, community characteristics and the depth and breadth of the dissemination tree significantly affect credibility assessment of events on Twitter.