Identification of influential users on Twitter: A novel weighted correlated influence measure for Covid-19.

Abstract

In the era of advanced mobile technology, freedom of expression over social media has become prevalent among online users. This generates a huge amount of communication that eventually forms a ground for extensive research and analysis. The social network analysis allows identifying the influential people in society over microblogging platforms. Twitter, being an evolving social media platform, has become increasingly vital for online dialogues, trends, and content virality. Applications of discovering influential users over Twitter are manifold. It includes viral marketing, brand analysis, news dissemination, health awareness spreading, propagating political movement, and opinion leaders for empowering governance. In our research, we have proposed a sustainable approach, namely Weighted Correlated Influence (WCI), which incorporates the relative impact of timeline-based and trend-specific features of online users. Our methodology considers merging the profile activity and underlying network topology to designate online users with an influence score, which represents the combined effect. To quantify the performance of our proposed method, the Twitter trend #CoronavirusPandemic is used. Also, the results are validated for another social media trend. The experimental outcomes depict enhanced performance of proposed WCI over existing methods that are based on precision, recall, and F1-measure for validation.