Abstract: Approaches of page segment reduce the unit of Web information extraction from page to block. This paper studies the main approaches of page segment and the basedlearning block importance model, and analyses the approach of Xpathbased Web information extraction. Combining the advantages of the two approaches, this paper proposes a new Web information extraction based on combining block importance model and Xpath, discusses its design process, and gives its formalized description and experimental result. The result shows that this approach is fit for extracting from the Web which has many records.

Key words: page segment, value of block importance, Xpath, Web information extraction