OLERA: On-Line Extraction Rule Analysis for Semi-structured Documents

The vast amount of online information available has led to renewed interest in information extraction (IE) systems that analyze input documents to produce a structured representation of selected information from the documents. Information extraction from semistructured documents has been studied extensively recently. Most researches focus on supervised learning approaches where targets must be labelled in the training set. Information extraction with unlabelled training set is hard and only works for multi-record documents. This paper introduces OLERA – OnLine Extraction Rule Analysis to the rapid generation of IE systems that can extract structured data from semistructured Web documents with unlabelled training set. In this novel framework, extraction rules can be trained not only from a multiple-record Web page but also from multiple single-record Web pages (called singular pages). Evaluation results show a high level of extraction performance for both singular pages and multi-record pages.