An Efficient Ranking Model for Multiuser Social Media applications using Customized Page Rank Method

Abstract

It has been found that the web is being used as a tool by radical or extremist groups and users to carry out a variety of acts with hidden agendas and promote their ideas in every part of the world. Some web forums are mainly used for open discussion on serious issues influenced by radical ideas. Influential users dominate and influence newly connected innocent users through their radical ideas. This paper introduces a principle for identifying radically influential users in web forums. Posts commented with a list of threats are captured by the user by a criterion based on the degree of match. This web forum basically identifies influential users. Eleven different collocation metrics are formulated to identify the association between users, and they are finally embedded in a customized page rank algorithm to create a numbered list of basic influential users. Connection theory is more effective in dealing with such ranking problem than measures based on tactical and temporal equality.