网络信息安全防范及Web数据挖掘系统设计及实现.doc

网络信息安全防范及Web数据挖掘系统设计及实现 　　摘 要： ?魍车男畔⑼诰蚍椒ㄍ诰蛎嬲?，扩展性差，无法有效挖掘出网络中的不安全信息。因此，设计并实现了网络信息安全防范与Web数据挖掘系统，其由Web文本采集模块、文本分类模块和类别判断模块构成。Web文本采集模块从网络Web网页中采集文本信息，并将信息反馈给文本分类模块。文本分类模块由训练模块、分类模块和分类器构成。训练模块采用完成分类的文本对文本分类模型进行训练，获取不同类别特征词间的关联性，塑造向量空间模型。分类模块对将要进行分类的Web文本进行分词处理，通过向量描述文本特征词。分类器运算待分类文本特征向量同各类中心向量间的相似度，确保Web文本被划分到具有最高相似度的文本类型中。类别判断模块辨识待分析的网络文本信息是否属于不安全信息类，并通过报警模块对不安全信息进行报警。软件部分给出了系统的功能结构以及文本分类模块的程序实现代码。实验结果表明，所设计系统具有较高的查全率、查准率和较高的检测性能 关键词： 网络信息； 安全防范； Web数据； 数据挖掘 中图分类号： TN711?34； TP309 文献标识码： A 文章编号： 1004?373X（2017）04?0061?05 Design and implementation of network information security protection and Web data mining system ZHAO Yuepin （Hebei Jiaotong Vocational and Technical college， Shijiazhuang 050091， China ） Abstract： The traditional information mining method has narrow mining face and poor scalability， so it cannot effectively dig out the unsafety information in the network. Therefore， the network information security protection and Web data mining system was designed and realized. It is composed of Web text acquisition module， text classification module and category judgment module. The Web text acquisition module is used to collect text information from the Internet Web pages， and feeds the information back to text classification module. The text classification module is made up of training module， classification module and classifier. The training module adopts the text completing classification to train text classification model to obtain the correlation among different category feature words and establish vector space model. The classification module is used to conduct the segmentation processing of words in Web text under classification and diescribe the text feature words through vector. The classifier is used to operate the similarity between the character vector of the text under classification and all kinds of central vector to ensure that the Web text is divided into the text type with the highest similarity. The category judgment module i