Tie Formation on Twitter: Homophily and Structure of Egocentric Networks

Both homophily and tie formation remain well studied problems in the social sciences literature since several decades. Prior empirical work on networks seems to indicate that in several situations, high degree of homophily was observed to be behind an individual´s desire to establish new ties. However in the context of today´s social media, not all individuals are the same, topologically. That is, we observe a variety of ego network structures based on individuals´ inlink-outlink relationships. In the context of tie formation in these networks, do different ego network topologies exhibit homophily along different attributes? How can we quantify and investigate these differences? To this end, we propose a variety of attributes along which homophily can be measured between individuals: including demographic attributes, activity-specific attributes and content-based attributes. We further categorize ego network structures as generators, mediators and receptors based on a measure called ego ratio. In our experiments on a large Twitter dataset comprising about 29.5M tweets, we observe, somewhat surprisingly, that for a few attributes, there was consistently high homophily (topical interest)/low homophily (gender, ethnicity) regardless of the ego network structure. Otherwise, mediators were observed to associate tie formation extensively with location, interactive ness and sentiment homophily, whereas generators were driven largely by information broadcasting behavior homophily. Implications of the findings in understanding the diverse range of motivations behind user participation in social media sites are discussed.