I know what the BOTs did yesterday: Full action sequence analysis using Naïve Bayesian algorithm

Abstract

A game BOT is a major threat in the online game industry. There have been many efforts to distinguish game BOT users from normal users. Several studies have proposed BOT detection models based on the analysis of users' in-game action sequence data. These studies indicated that the analysis of users' in-game actions is effective to detect BOTs. However, they do not use sufficiently large data sets to train and test their algorithms. In this paper, we have proposed a BOT detection model that uses users' in-game action sequence data obtained with the aid of big data analysis environments. We did empirical analysis of the dataset of "Blade and Soul", the third largest MMORPG in Korea. The result shows that a large amount of sequence data leads to high accuracy.

Improved Mitigation of Cyber Threats in IIoT for Smart Cities: A New-Era Approach and Scheme

“…Trading network Examining a game character's possession event log and transaction event log to derive it as a feature [29,30] Gameplay style Investigating gameplay styles such as player information, player action, and combat ability [31,32] Social network Analysis of social network characteristics between players such as part play logs and chat logs [33][34][35] Sequence analysis Characterized by assuming that the player's actions, such as action sequences and battle sequences, are one sequence [36][37][38] Self-similarity Analyzed based on the assumption that the bots have self-similarity, and the action frequency and action type are used as features [13,39] Character movement Identifying a character's movement pattern and use movement speed, distance, and location [40,41] Character behavior Observering the character's behavior and using it as a feature by applying various statistics [17,[42][43][44][45] We analyzed the logs generated in a smart city's sensor layer using the hardware-inthe-loop (HIL)-based augmented ICS (HAI 1.0) dataset, which is an infrastructure dataset, as a case study. Furthermore, we investigated the applicability of the proposed approach to such sensor logs and determined how to best apply it.…”

“…In Kang et al [35], the party player logs related to social interaction were used; however, the interaction data between sensors cannot be known. In the case of a sequence analysis, the action and battle sequences were analyzed to detect any bots [36][37][38]. • Chen et al [41] • • Yu et al [42] • • Han et al [44] • • Chen et al [45] • • • Park et al [17] •…”

You are a Game Bot!: Uncovering Game Bots in MMORPGs via Self-similarity in the Wild

“…In [20], sequence mining techniques were used to detect bots and it was found that there is a difference in sequence patterns between bots and human players. We also examine that same characteristics in this work.…”

A study of RMT buyer detection for the collapse of GFG in MMORPG

“…현금거래시장은 자 금세탁 등의 문제로 이용되기도 하였다 [9]. GFG의 네트워크 구조는 현실세계의 마약밀매, 테러조직, 범죄 집단 구조와 유사한 형태를 보이는데 [4,17] -Idle time analysis [7] -Trajectory-based analysis [11,20] -Social interaction analysis [12,13,14] -Action frequency analysis [3] -Action sequence pattern analysis [25,26] GFG detection -Trade network analysis [4,22] -Role based analysis [18] Table. 3.와 같이 설정하였다. Table.…”