Analysis of Behavior Patterns to Identify Nicknames of a User in Online Community

摘要

An online community is a virtual group that is mediated through the Internet for users to share interests and hobbies. Unlike social network service (SNS), an online community is an anonymous service mainly based on nickname. Some users exploit this anonymity and conduct malicious activities. Actions should be taken to filter these users and limit their activities. One problem lies in that nicknames are frequently changed in online communities, and automatically filtering nicknames is difficult when they are constantly malicious. Another problem is data fragmentation in which the data of the same user exists under different nicknames due to the first problem. Therefore, to solve these issues, we propose a behavior pattern feature vector, which considers online community characteristics and identifies nicknames of the same user. Specifically, we propose a method to identify nicknames of the same user using actual data of an online community in Korea.