Egocentric Analysis and Visualization of Instant Messaging Activity

In this paper, we develop novel analysis and visualization schemes to reveal to the user the communication context of her instant messaging network. Cues for temporal communication context are critical during human communication, however they are not currently revealed in popular instant messaging tools. Our work addresses privacy issues by focusing on a single individual’s communication with her contacts. There are two key contributions – (a) analysis of social and semantic aspects of communication. We detect features such as frequency, dominance, density, longevity and determine the dominant semantic concepts. (b) a ring based network visualization – to compactly represent structural and semantic aspects of communication. Through a long term use of our system, we received detailed and rich observations on how it was used. It was well liked by all participants.