Game Bot Detection Dataset Based on User Behavioral Characteristics

This dataset contains long-term in-game behavioral logs collected from a major MMORPG service environment. The data spans 88 days of gameplay, covering 49,739 player characters, of which 7,702 were confirmed game bots identified by the game operator. The dataset includes raw and processed behavioral features derived from gameplay logs, such as player activities, social interactions, economic transactions, group play behavior, and network-based measures. These features were extracted from event logs generated by the game server whenever users performed actions such as hunting, harvesting, trading, chatting, or party play.The dataset integrates multiple behavioral categories, including player information (login patterns, play time, resource acquisition), action ratios, party behavior metrics, social interaction diversity, and graph-based measures computed from in-game interaction networks (party, friendship, trade, whisper, mail, shop). All personally identifiable information is anonymized according to the game provider\u2019s internal policies.