Inferring who-is-who in the Twitter social network

In this paper, we design and evaluate a novel who-is-who service for inferring attributes that characterize individual Twitter users. Our methodology exploits the Lists feature, which allows a user to group other users who tend to tweet on a topic that is of interest to her, and follow their collective tweets. Our key insight is that the List meta-data (names and descriptions) provides valuable semantic cues about who the users included in the Lists are, including their topics of expertise and how they are perceived by the public. Thus, we can infer a user's expertise by analyzing the meta-data of crowdsourced Lists that contain the user. We show that our methodology can accurately and comprehensively infer attributes of millions of Twitter users, including a vast majority of Twitter's influential users (based on ranking metrics like number of followers). Our work provides a foundation for building better search and recommendation services on Twitter.