基于线索效用理论的用户虚假信息感知与互动行为研究——面向深度伪造信息

Research on Users'Misinformation Perception and Interaction Based on Cue Utility Theory:Toward Deepfake Information

[研究目的]作为一种新形式的虚假信息,深度伪造信息有着更难以辨别真假的特点,影响着网络安全空间的发展。该文将从用户感知和行为角度来探索深度伪造类虚假信息的影响和对抗策略。[研究方法]根据线索效用理论,发掘并整合了三个外部线索:分享者信任度、分享者观点、附加提示,并设计了两个组间实验。[研究结论]310个实验样本的数据通过MANCOVA方法进行分析后验证了8个假设,分享者信任度、分享者观点、附加提示都会影响用户对深度伪造信息的感知真实性和互动行为(点赞、评论、转发),这表明用户判断深度伪造信息时会综合效用三个外部线索,同时附加提示的有效性也证明了干预手段在多线索集聚情景下的有效性,为提高用户识别深度伪造信息的能力,减少虚假信息的传播提供了可实践化的参考。

[Research purpose]As a new form of misinformation,deepfake information has the characteristic of being more difficult to distinguish true from false,which affects the development of cyber security space.In this paper,it will empirically estimate the impact and countermeasure of deepfake disinformation from the perspective of user perception and behavior.[Research method]Based on cue utility theory,we uncovered and integrated three external cues:sharer's trust,sharer's viewpoint and attaching tips,and designed two between-group experiments.[Research conclusion]The data from 310 experimental samples were analyzed by MANCOVA method to verify eight hypotheses.It was shown experimentally that sharer's trust,sharer's viewpoint,attaching tips all affect users'perceived authenticity and interaction(like,comment and retweet)of deepfake information,which shows that the users will combine effect of three external cues when they diagnosis the deepfake information.The effectiveness of the attaching tips also demonstrates the availability of the intervention in a multi-cue aggregation scenario,which provides a significant reference for improving users'ability to identify deepfake information and reducing the spread of misinformation on social media.