基于属性聚类的传销网站账户去重方法研究

在网络传销案件取证过程中,需要分析传销活动的实际参与人数,但是由于网络身份的虚拟性,经常存在一人使用多个账户的情况,因此需要对传销网站中的账户进行去重分析。目前常用的方法是将姓名、身份证号等信息一致的账户做去重处理,但是该方法无法对使用虚假身份注册的账户进行去重。文章提出了一种基于属性聚类的账户去重方法,通过对与使用者身份关联的属性进行聚类分析,从而实现对传销网站中的账户进行去重。

In the process of collecting evidence in network marketing cases, it is necessary to analyze the actual number of participants in these pyramid marketing activities, but due to the virtual nature of network identity, there is often a situation in which one person uses multiple accounts. Therefore, it is necessary to reanalyze the accounts in the network marketing website. At present, the commonly used method is to reprocess the account with the same information such as name and ID number, but this method cannot reprocess the account registered with false identity. In this paper, an account de-duplication method based on attribute clustering is proposed. Through the cluster analysis of the attributes associated with the user's identity, the account in the pyramid selling website can be deduplicated.