Extraction technology of blog comments based on functional semantic units

Blog is an important kind of network information resources, extracting its comments information is essential for the researches of public opinion analysis and so on. In this paper we summarized the prevalent extraction algorithms of blog comments and described how to use page structure in information extraction. The indicator phrases such as "Home" have clear semantics and functional indication when people understand the web pages. The indicatior phrases are known as Functional Semantic Units (FSU). Base on the characteristic of FSU We propose a kind of comment information extracting model, and present a detailed model of thinking and implementation process. Such as the page structure linearized, functional semantic units are distinguished, main text are recognized and comments extraction algorithm etc. Finally, the experiments prove that the comment information extracting model is effective and better identification results.