摘要
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5475条页面的分析数据进行科学的分析处理,每一页面的属性都有10个,根据这些属性建立基于支持向量机的视觉信息页面块分类模型,再利用LIBSVM工具对于所有页面的属性数据进行深入优化、分析、处理和解释,最后打得到的分类结果准确率达到97%,这样好的结果对网页的认识提供了非常有效的理论依据。
Gradually with the development of network technology, network of everyone needs to face the different website, and web page structure is complex, the understanding of the human senses for page with the human eye can see is not the same.In this paper, through the analysis of 5473 page data analysis process, the attribute of each page has 10, build piece of visual information page classification based on support vector machine (SVM) model, using LIBSVM tools for analyzing the complex pages of attribute data, processing, optimization and explanation, the results of the classification accuracy is as high as 97%, thus a deeper understanding of web page provides effective theoretical basis.
作者
董婷
Dong Ting(School of Information Engineering,Yulin University,Yulin719000,Chin)
出处
《现代科学仪器》
2018年第1期133-135,共3页
Modern Scientific Instruments
基金
教育厅项目(11JK0636)
关键词
支持向量机
核函数
页面快
训练变量
Support vector machine (SVM)
Kernel function
Page soon
Training variables