一种基于ELM的在线游戏外挂检测方法  

A Method of Online Game Bot Detection Based on Extreme Learning Machine

作　　者：黄旭[1,2] 曾孟佳[1] 范婧 高少文[2] 胡文军 Huang Xu;Zeng Mengjia;Fan Jing;Gao Shaowen;Hu.Wenjun(School of Information Engineering, Huzhou University, Huzhou 313000;College of Control Science and Engineering, Zhejiang University, Hangzhou 310027)

机构地区：[1]湖州师范学院信息工程学院,湖州313000 [2]浙江大学控制科学与工程学院,杭州310027

出　　处：《微型电脑应用》2018年第4期1-4,12,共5页Microcomputer Applications

基　　金：国家自然科学基金项目(61202290;61370173;61772198);中国博士后科学基金项目(2015M581933)资助

摘　　要：游戏外挂是一种替代人参与游戏的第三方软件或程序,在大规模多用户在线角色扮演游戏中,一些用户利用这一类程序快速提升用户等级、快速积累虚拟财富。游戏外挂的使用,扰乱了正常的游戏秩序,降低了大量玩家的良好体验。如何准确快速检测游戏外挂是各大游戏公司亟待解决的问题。将极限学习机(Extreme Learning Machine,ELM)方法用于游戏外挂检测,分析游戏标签数据并形成用户行为序列矩阵,利用ELM对游戏外挂进行分析和筛查。并在网易游戏数据集上进行了测试。同时将该方法的实验结果与传统的SVM和KNN算法进行了对比。实验结果表明,在分类精度和效率方面,其算法具有一定的优势。Game bot is a third party software or program that substitutes for people to participate in games.Some players of massively multiuser online role-playing games（MMORPG）manipulate game bots to accumulate property quickly in the game world,for getting a high-level experience quickly without spending too much time and energy.It has a great impact on the game experience of human players,and lead to an unfair phenomenon in games.We analyze and screen players in online games to quickly capture game bots,and let game operators do subsequent processing.First,we analyze game log data and arrange user behavior sequences to form a matrix with user information.Second,extreme learning machine（ELM）is used for classifing and screening.Some traditional classification methods,i.e.SVM and KNN,are used on the same data to verify the algorithm effect.Empirical study demonstrates that the proposed method is competitive with these traditional methods in accuracy and efficiency.

关 键 词：游戏外挂 大规模多用户在线角色扮演游戏 极限学习机 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]