Web of Friends- Discovering a Social Network by Mining Data from Instant Messengers

This paper describes the first step in our research on egocentric social networks of users of instant messengers. Experiments with which we are proceeding are based on one of the popular Polish messengers-Tlen.pl. To collect information about interactions we have used a plug in for Tlen that cooperates with a dedicated server for storing data. Our goal is to find algorithms that will allow classifying the strength of a relationship between two users. Also, we want to answer the question, whether basing on a set of statistics from chats, without knowing the content of the discussion, we can correctly predict levels of acquaintance between peers.