Semi-Automated Wrappers Using Rule Trees

In this paper we describe the concept of a semi-automated wrapper for extracting information from semi-structured pages, usually part of the e-commerce data intensive web sites. The process is based on creating extraction rules in a visual manner, using the DOM tree associated to a XHTML document, helping the user to make the right decisions. The extraction rules defined have a natural tree structure. Based on the model designed, the wrapper can then be used to navigate through the site and extract the relevant data.