Friendship paradox and hashtag embedding in the instagram social network

Abstract

Instagram is a social networking platform which gained popularity even faster than most of the other modern online social networks. It is relatively newer and less explored than other social networks, such as Facebook and Twitter. Therefore, we have conducted a research based on a sample data set extracted through the Instagram weekend hashtag project, in order to unveil some of its characteristics. First, we reveal the various forms of friendship paradox present in Instagram, which are often observed in social networks. Then, we conduct a detailed hashtag analysis and provide a method for hashtag representation and recommendation using natural language processing.