Extracting Information from Web Content and Structure

Extracting Information from Web Content and Structure

Web is a vast data repository. By mining from this data efficiently, we can gain valuable knowledge. Unfortunately, in addition to useful content there are also many Web documents considered harmful (e.g. pornography, terrorism, illegal drugs). Web mining that includes three main areas ÔÇô content, structure, and usage mining ÔÇô may help us detect and eliminate these sites. In this paper, we concentrate on applications of Web content and Web structure mining. First, we introduce a system for detection of pornographic textual Web pages. We discuss its classification methods and depict its architecture. Second, we present analysis of relations among Czech academic computer science Web sites. We give an overview of ranking algorithms and determine importance of the sites we analyzed.

Keywords: Web mining, information retrieval, classification, ranking algorithms

Year: 2006

Download: download Full text [213 kB]