Semantic homophily in online communication: evidence from Twitter

摘要: People are observed to assortatively connect on a set of traits. This phenomenon, termed assortative mixing or sometimes homophily, can be quantified through assortativity coefficient in social networks. Uncovering the exact causes strong found networks has been research challenge. Among main suggested from sociology tendency similar individuals (often itself referred as homophily) and influence among already connected individuals. An important question researchers practice tackled, we present here: understanding mechanisms interplay between these tendencies underlying network structure. Namely, addition mentioned coefficient, there several other static temporal properties substructures that linked homophily herein investigate those. Concretely, tackle computer-mediated \textit{communication network} (based Twitter mentions) particular type inferred semantic features communication content term \textit{semantic homophily}. Our work, best our knowledge, is first offer an in-depth analysis them. We quantify diverse levels identify aspects drivers show insights its evolution finally, intricate with Twitter. By analyzing increase what shape how they human communication.