A Behavior Analysis-Based Game Bot Detection Approach Considering Various Play Styles

摘要

An approach for game bot detection in MMORPGs is proposed based on theanalysis of game playing behavior. Since MMORPGs are large scale games, userscan play in various ways. This variety in playing behavior makes it hard todetect game bots based on play behaviors. In order to cope with this problem,the proposed approach observes game playing behaviors of users and groups themby their behavioral similarities. Then, it develops a local bot detection modelfor each player group. Since the locally optimized models can more accuratelydetect game bots within each player group, the combination of those modelsbrings about overall improvement. For a practical purpose of reducing theworkloads of the game servers in service, the game data is collected at a lowresolution in time. Behavioral features are selected and developed toaccurately detect game bots with the low resolution data, considering commonaspects of MMORPG playing. Through the experiment with the real data from agame currently in service, it is shown that the proposed local model approachyields more accurate results.