Influence analysis of posts in social networks by using quad-motifs

Abstract

Word of epidemic is a very popular term in modern world which social media has affect almost all people along all over the world. Modern World's social networks are very sophisticated networks and learning information is very crucial for analyze user behaviors. Finding user's effect on other users/people is very interesting point in complex networks. Almost all related works only focus on finding and analyzing user behaviors for creating user profiles. However, in real life, all posts actually have different effect so have different influence values. Our proposed method that based on fuse motif analysis (FMA), focus on finding most effective posts for Instagram Social Network which one of the most popular social networks. Proposed method firstly calculate all posts influence values on other people. In our method, it is take into account variety of factors include users' other posts popularity, emotional based sentimental analyze on comments and tag frequency. It has been proposed that to create a model to predict most influencing posts including all determined factors. Proposed method applied and analyzed on Instagram data which gathered by us and share our experimental results in the paper.