Semi-Structured Information Extraction Applying Automatic Pattern Discovery

Abstract

Information extraction (IE) from semi-structured Web documents is a critical issue for information integration systems on the Internet. Previous work in wrapper in- duction aim to solve this problem by applying machine learning to automatically generate extractors. For ex- ample, WIEN, Stalker, Softmealy, etc. However, this approach still requires human intervention to provide training examples. Hence, the other track to informa- tion extraction tries to save human eort. For exam- ple, Embley et. al. and Chang et al. present dier- ent approaches to record boundary identication of a single Web pages without any training example. Emb- ley's work relies on the intra-page structure constructed by HTML tags (the parse tree), while Chang's work is motivated by repeated patterns formed by multiple aligned records. This paper expands Chang's work to IE and discuss the issues when applying pattern dis- covery for record identication, including the encoding schemes of HTML and ranking criteria of patterns to extract record boundary.