Deepec: An Approach For Deep Web Content Extraction And Cataloguing

This paper presents DeepEC (Deep Web Extraction and Cataloguing Process), a new method for content extraction of Deep Web databases and its subsequent cataloguing. Our focus is on the extraction of hidden Web content presented in HTML pages generated from Web forms query submissions. While state-of-the-art information extraction and cataloguing methods address this issue separately, DeepEC is able to simultaneously perform the extraction and cataloguing of data without expert user intervention. This is accomplished with the support of a knowledge base that allows semantic inference for relevant records to be extracted and then catalogued. An experimental evaluation on some Deep Web domains shows that DeepEC achieves very good results. If compared to related work, DeepEC provides a unified process for Deep Web content extraction and cataloguing, being able to infer missing values for extracted records to be catalogued.