欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
当今主流网页分为单正文体网页和多正文体网页。这些网页的正文信息都具有多个正文特征。想要准确定位正文信息所在位置,可以从其所具有的多个特征和网页设计者的设计习惯着手。鉴于此,融合这些特征提出一种基于多特征融合的网页正文信息抽取方法。实验结果表明,该方法对单正文体网页和多正文体网页的正文抽取具有较高的准确率和通用性,很好地适应了风格多样的网页。
英文摘要：
Nowadays, the mainstream webpages are divided into single text body webpages and multiple text body webpages. These webpages text information all have the heterogeneous text features. In order to accurately position the location of text information, one can commence from the heterogeneous features they possess and the design habits of the web designers. Therefore, in this paper, we fuse these features and present a webpages text information extraction method which is based on heterogeneous features fusion. Experimental results show that the method has higher accuracy rate and universality for extracting the texts from single test body webpages and multiple text body webpages, fits well the webpages with a variety of styles.
同期刊论文项目
同项目期刊论文
期刊信息
位置：