基于LIBSVM的视觉信息页面块分析模型

Analysis of Visual Information Page Based on LIBSVM

随着网络技术的发展,逐渐的每个人都需要面对网络,浏览不同的网页,而网页页面的组成结构复杂,人的感官对于页面的认识与人眼所见并不相同。本文通过对于5473条页面的数据先分析再处理,然后把每一页的属性设置为10个,最后在基于支持向量机的基础上建立分类模型,对复杂的属性数据,我们应用LIBSVM先对其进行分析然后再根据分析结果对其进行处理、紧接着对其优化和解释,分类结果最终的准确率是97%,从而给网页的更深层的认识提供了有效的理论依据。

With the development of network technology,everyone needs to face the different websites. Web page structure is complex,and the understanding of human senses on page is different from what human eye see. In this paper,through the analysis of 5473 pieces of page data,the attribute of each page is set as 10,and visual information page classification model based on support vector machine( SVM) model is established. LIBSVM is used for analysis,processing,optimization and explanation of complex attribute data,and the accuracy rate of the classification results is as high as 97%,which provides a effective theoretical basis for a deeper understanding of webpage.