Modeling Information Credibility in Time-Specific Online Social Media based on Structural and Attribute Properties

Abstract

Nowadays, millions of people actively participate in online social networks (OSNs) everyday to express and share their ideas, opinions, interests, feelings about various events, governmental policies, current economics, societal debates etc. Thus, these numerous user generated contents diffuse very rapidly in OSNs through different peers in various formats such as tweets, images, videos etc. Therefore, it is very crucial to extract credible information from these mass volume of social user-generated contents. Existing research works on measuring trustworthiness of any content in OSNs are typically deliberated on contents generated by the social users. However, such methodologies overlook the prospective temporality of users’ interests as well as paid less attention to the structure of the social network. In this work, we propose an approach to measure the level of credibility of a piece of information in OSNs based on the social users’ temporal topical interests and the structural properties of the underlying social network. The effectiveness of the proposed methodology is justified using three benchmark datasets.