Analysis of User Behavior on Private Chat System

Recently, more minors are using online communication services due to the popularity of mobile devices and smartphones. Even though minors can deepen their relationships by such communication, online communication sites introduce risks including cyber-bullying or the luring of sexual predators. Such communication risks involving minors are social matters that must be detected automatically and prohibited. In this paper, we clarify communication behaviors by focusing on private chat systems, which are sometimes used to sexually entrap minors. Our analysis found that most of users communicate with less then 5 users, and the communication ended within a half hour. Then we classified users into 15 clusters based on their communication behaviors. From the clustering results, we identified both active communication senders and active communication receivers. Such communication receivers must be monitored because they might have much more opportunities to be entrapped by sexual predators.