基于写作风格的中文Web信息作者同一认定方法

摘要

为了阻止网上非法信息现象发生,提出了一种同一认定Web信息作者的方法,通过分析中文Web信息作者的写作风格,提取能表达Web信息作者写作特点的三种特征,包括词汇特征、结构特征和格式特征,利用支持向2机分类学习算法,同一认定Web信息的作者,为计算机取证提供证据.在Blog、电子邮件数据集上实验的分类识别正确率超过8000,表明所提出的方法是有效的,用于计算机取证是切实可行的.%To prevent illegal information in the Internet from happening, one Web information identical cognizance method was provided. By analyzing Chinese Web information author's writing style, lexical features, structural features, format features which could express Web information author's writing habit were extracted. Support vector machine algorithm was used to cognize Web information's author identically. The purpose of the method was to investigate evidence to computer forensic. The accuracy exceeded 80 percent by experimenting on Blog and E-mail datasets. The three features combination had a better result than single feature. The experimental results proved that the method was effective and feasible to apply for computer forensic.