Towards a Method for Unsupervised Web Information Extraction

Abstract

The literature provides a variety of techniques to build the information extractors on which some data integration systems rely. In formation extraction techniques are usually based on extraction rules that require maintenance and adaptation if web sources change. We present our preliminary steps towards an unsupervised information ex traction technique that searches web documents for shared patterns and fragments them until finding the relevant information that should be ex tracted. Experimental results on 1230 real-web documents demonstrate that our system performs fast and achieves promising results.