中文摘要：

在利益驱动下,社交网络中出现大量虚假账户,其发布的虚假消息可对正常用户产生误导。通过对社交网络中大量数据进行分析,发现虚假账户与正常账户在账户特性、行为特性上有较大差异。基于这些差异,结合Rough Set相关理论提出账户信任度的计算模型。所得信任度可用以区分虚假账户,并为正常用户的判断提供依据。实验显示,根据所得信任度对账户排序得到了较好效果,并能够有效区分虚假账户。

英文摘要：

Driven by interests, many fake accounts appear in online social network. These fake accounts release false news and mislead the normal users. Analysis on numerous accounts in online social network indicates that the fake accounts are different from the normal accounts in account characteristics and behavior characteristics. Based on Rough Set theory, a model is built up to evaluate trust value of the account. The trust value can be used to identify the false account, and provide decision basis for normal users. Experiments show that the, ranking of accounts according to the trust value could achieve good effect and effectively identify the fake account.