Journal of Information Security Reserach ›› 2023, Vol. 9 ›› Issue (1): 57-.

Previous Articles     Next Articles

Social Network Rumor Suppression Scheme Based on Opinion Leaders

  

  • Online:2023-01-01 Published:2022-12-31

社交网络基于意见领袖的谣言抑制方案

陈雄逸;许力;张欣欣;尤玮婧;   

  • 通讯作者: 陈雄逸 硕士研究生.主要研究方向为移动社会网络、网络与信息安全. chenxiongyi0596@163.com
  • 作者简介:陈雄逸 硕士研究生.主要研究方向为移动社会网络、网络与信息安全. chenxiongyi0596@163.com 许力 博士,教授,博士生导师.主要研究方向为移动社会网络、大数据与信息化、无线通信与物联网、智能信息处理. xuli@fjnu.edu.cn 张欣欣 博士.主要研究方向为移动社会网络、网络与信息安全. 1175045320@qq.com 尤玮婧 博士,副教授,硕士生导师.主要研究方向为云存储数据安全. youweijing@fjnu.edu.cn

Abstract: With the rise of social networks, the security of information content caused by information transmission in the network has become more and more significant, among which the spread of rumor information is an important potential threat to the security of information content in the network. Most of thecurrent rumor suppression schemes in social networks consider the structural attributes of the network and adopt the information propagation model of a single entity to simulate the real communication scene. This study considers not only the structural attributes of the network, but also the social attributes of nodes, and adopts the Multientity Competitive Independent Cascade Model proposes a rumor suppression scheme based on opinion leaders. First, according to the structural properties of the network, the key users in the network are selected as the candidate set of opinion leaders through the dominating set. Secondly, according to the social attributes of users, a user influence calculation formula is proposed to identify the opinion leader groups in the network. Finally, the Multientity Competitive Independent Cascade Model is used to conduct simulation experiments on several real social network datasets of different sizes. The experimental results show that the proposed scheme can suppress the amount of rumor information in social networks better than other rumor suppression schemes in a reasonable time cost, and the final rumor suppression effect can be improved by about 5% with the increase of the amount of rumor information in the network.

Key words: opinion leader, multientity competitive independent cascade model (MCIC), social networks, suppression of rumors, user influence

摘要: 随着社交网络的兴起,网络中信息传播所引发的信息内容安全问题愈发突出,其中谣言信息的传播是网络中信息内容安全面临的重要潜在威胁.当前社交网络中的谣言抑制方案大多考虑网络的结构属性,并采用单一实体的信息传播模型模拟现实传播场景.不仅考虑了网络的结构属性,而且考虑了节点的社交属性,并采用多实体的竞争性独立级联模型,提出了一种基于意见领袖的谣言抑制方案:首先,根据网络的结构属性,通过支配集选取网络中的关键用户作为意见领袖候选集;其次,根据用户的社交属性,提出了用户影响力计算公式识别网络中的意见领袖群体;最后,采用多实体的竞争性独立级联模型在几个大小不同的真实的社交网络数据集上进行仿真实验.实验结果表明,该方案在合理的时间成本内能够比其他的谣言抑制方案更好地抑制社交网络中谣言信息的数量,并且随着网络中谣言信息数量的增加能够保持最终的谣言抑制效果有5%左右的提升.

关键词: 意见领袖, 多实体的竞争性独立级联模型, 社交网络, 谣言抑制, 用户影响力