Research for Information Extraction Based on Wrapper Model Algorithm

Abstract

Mainly on data-intensive Web site research experiment. In the web pages of the automatically generated wrapper method of research-based information extraction, the main job is to make the page tree matching algorithm, the sample tree and the tree wrapper DOM tree matching two pages compared to the first to discover the page selection mode, producing the primary template, and then selfcorrection of primary template found iterative model, and finally generate the page wrapper method. The wrapper generation process does not require human intervention to achieve a fully automated completion. Experiment with satisfactory results.