Bootstrapping Information Extraction from Semi-structured Web Pages

We consider the problem of extracting structured records from semi-structured web pages with no human supervision required for each target web site. Previous work on this problem has either required s