Query-oriented topical influential users detection for top-k trending topics in twitter

Online Social Networks (OSNs) have become inevitable for any new methodology both for viral promoting applications and instructing the creation of inciting information and data. As a result, finding influential users in OSNs is one of the most studied research problems. Existing research works paid less attention to the temporal factors associated with the activities performed by the social users. Our motivation is to find influential users who show their most powerful interests towards a given query on various subjects (topics) at different time intervals by featuring more on users’ most recent activities as well as their associations with different users. To address this problem, we propose a temporal activity-biased weight model that gives higher weight to users’ recent activities and develops an algorithm to list the most effective influential users. In addition, our proposed model also considers the impacts of topical similarities both from direct and indirect neighbors of the users. Experimental results on two real datasets demonstrate that our proposed framework yields better outcomes than the baseline method.