多事件情景下机器人用户行为模式研究  

The Behavioral Patterns of Social Bots under Multi-event Context

作　　者：黄音莅 沈宜 贾宇 刘峤[2] HUANG Yinli;SHEN Yi;JIA Yu;LIU Qiao(Guangdong Cyberspace Cognitive Domain Engineering Technology Research Center,Shenzhen CyberAray Network Technology Co.,Ltd.,Shenzhen Guangdong 518000,China;University of Electronic Science and Technology of China,Chengdu Sichuan 611731,China)

机构地区：[1]深圳市网联安瑞网络科技有限公司广东省网络空间认知域工程技术研究中心,广东深圳518000 [2]电子科技大学,四川成都611731

出　　处：《通信技术》2024年第8期861-872,共12页Communications Technology

基　　金：国家自然科学基金企业创新发展联合基金项目(U22B2061)。

摘　　要：目前,利用机器人对社交媒体舆论进行操纵的现象变得愈发普遍。为应对社交媒体操纵对公众舆论的巨大影响,国内外学者主要对机器人用户识别、机器人用户行为及影响开展了研究。然而,鲜有研究关注不同事件中机器人用户的行为模式,并且不能确认从某个事件中分析出的行为模式是否适用于另一个事件。因此,通过采集社交平台X的数据,利用样本差异检验方法,对机器人用户在多事件场景中的异常特征和行为模式进行了交叉分析,发现非英语语境事件中机器人用户的身份设定异常较为显著,并且机器人用户会因为任务分工的差异而具备不同的行为模式。研究结论为机器人用户识别方法的改进和机器人用户行动目的推断提供了启示。At present,the phenomenon of manipulating public opinions on social media using social bots has become increasingly common.In response to the huge impact of social media manipulation on public opinions,scholars at home and abroad have mainly carried out research on social bot detection,social bot behavior and its influence.However,few studies have paid attention to the behavioral patterns of social bots in different events,and it is not possible to confirm whether the behavioral patterns gathered from one event are applicable to another.Therefore,by collecting data from social platform X and using the sample difference test method,this paper cross-analyzes the abnormal characteristics and behavioral patterns of social bots under multi-event context,and finds that the abnormal identity settings of social bots in nonEnglish context events are more significant,and that social bots will exhibit different behavioral patterns due to the differences in task assignments.The results provide inspiration for the improvement of social bots detection methods and the inference of social bots motivations.

关 键 词：社交机器人 社交机器人行为 社交机器人识别 多事件分析 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]