论文部分内容阅读
网络水军的存在极大地影响了网络上的信息质量,同时干扰了人们获取正常网络信息的渠道。根据采集到的用户信息,分析网络水军的行为模式,研究了网络水军的特征,并提取了多个水军特征指标,使用熵值法确定各指标权重,利用主题识别模型对用户特征进行降维,并结合多指标综合指数法建立了网络水军自动识别模型,使用数据挖掘的方法找到异常用户。实验结果表明,模型得到了82.4%的准确率和88.6%的召回率。
The existence of the network navy has greatly affected the quality of information on the network, and at the same time has interfered with the channels for people to obtain normal network information. Based on the collected user information, this paper analyzes the behavior pattern of the navy, studies the features of the navy, and extracts a number of navy features, uses the entropy method to determine the weight of each indicator, and uses the theme identification model to perform user characteristics Dimensionality reduction, combined with multi-index composite index method to establish a network water military automatic identification model, the use of data mining methods to find abnormal users. The experimental results show that the model got 82.4% accuracy rate and 88.6% recall rate.