摘要
现有网络舆情监测分析系统大多采用人工建立模型、网页逐个匹配的方法识别网页类型,不仅费时费力,而且随着网页的变化和快速增长,效率不断下降,如何让机器快速准确识别出网页类型成为迫切需要解决的问题。针对现有网页分类算法无法自动识别网页所属网络舆情载体类型的问题,深入研究了网页的超链接特征、内容特征和结构特征,构建了面向网络舆情载体类型识别的特征集,提出了基于综合特征的网页类型自动识别算法,应用SVM分类器对数据集进行训练和建模。实验结果表明,该方法能够很好地对网络舆情载体类型进行分类。
Extensive human efforts are required to build templates in genre recognition of web pages in present monitoring and analyzing system for network public opinions.As the Web changes and continues to grow,this manual approach becomes less effective,and how to recognize the genre of Web pages accurately and rapidly has become a crucial problem.To solve this problem,a feature set for Web pages automatic recognition is proposed,which is based on research of the hyperlink features,content features and structure features of Web pages.Hence a Web pages automatic recognition method based on multiple features is proposed,with a SVM classifier for training and model building.Experimental results indicate that this new method can improve the accuracy of Web pages genre recognition and outperform traditional methods.
出处
《信息工程大学学报》
2011年第6期738-744,共7页
Journal of Information Engineering University
基金
国家863计划资助项目(2007AA01Z439)
国家社科基金资助项目(09&ZD014)
关键词
网页类型
特征提取
自动识别
Web genre
feature selection
automatic recognition
