Semantic Wrappers for Semi-Structured Data Extraction

In this paper, we propose an approach to extract information from HTML pages and to add semantic (XML) tags to them. Wrapping is an essential technique used to automatically extract information from Web sources. This paper describes both, a general approach based on rules, which can be used to automatically generate wrappers, and an assistant generator wrapper called WebMantic. We also provide some experimental results to show that both the rule generation process and the preprocessing task are fast and reliable.