XML as a means to support information extraction from legal documents

摘要

XML seems to he the preferred standard for legal documents. Several DTDs have been created specifically for these documents by public institutions and private legal publishers, and work continues with the adaptation of the additional legislation for XML representation. We present an approach to automate information extraction processes from legal documents, that pays special attention to document structure and relationships. It combines information extraction techniques to extract document structure, to detect references (relationships), and tackles consolidation (document updating) of normative documents as a knowledge extraction process. In this paper we emphasise the role of XML during the implementation of these techniques. Its appropriateness to manipulate structure, to model semantic features, and to share information are decisive in making this phase feasible.