基于改进Rocchio的网络信息过滤算法

Network Information Filtering Algorithm Based on the Improved Rocchio

由于网络上信息内容庞杂,传统基于内容的过滤方法难以建立准确的类别模板,导致信息过滤效果较差,提出基于改进Rocchio的网络信息过滤算法。文章首先利用数据包捕获技术与协议解析技术获取网络文本信息,其次基于改进Rocchio生成类别过滤模板,最后对比待过滤信息与类别模板的相似度,实现网络信息过滤。实验结果显示,设计的网络信息过滤算法的查全率为95.7%、查准率为96.8%,具有较好的过滤效果。

Due to the complexity of information content on the internet,traditional content-based filtering methods are difficult to establish accurate category templates,resulting in poor information filtering performance.Therefore,an improved Rocchio based network information filtering algorithm is proposed.The article first utilizes packet capture technology and protocol parsing technology to obtain network text information.Then,based on improved Rocchio,a category filtering template is generated.Finally,the similarity between the information to be filtered and the category template is compared to achieve network information filtering.The experimental results show that the designed network information filtering algorithm has a recall rate of 95.7%and a precision rate of 96.8%,which has a good filtering effect.