Advanced techniques for legal document processing and retrieval

Authors: Pietrosanti E.¹; Graziadio B.²

Source: Artificial Intelligence and Law, Volume 7, Number 4, 1999 , pp. 341-361(21)

Publisher: Springer

Abstract:

A large interest has been dedicated in recent years to the study of models for textual databases amenable to an effective integration of search and navigation functions. In the field of legal databases the need for sophisticated models is emphasised by the need to relate and combine in an effective way different types of texts, in order to solve legal problems.

In our research we have analysed several existing models, each providing specific benefits and exhibiting corresponding limitations, under both a functional and economical viewpoint.

Under a functional point of view, a distinctive feature of our model is the representation of relevant context information, aimed at improving the retrieval accuracy, in a framework in which the availability of multiple (structural, conceptual and functional) views over the legal texts emphasises the issues of the transparency of the model and of the incrementality of the search process. The model has been experimented on a significant excerpt of the Italian banking regulations and fiscal law, embodied in the NaviLex experimental system.

On the other hand, sophisticated models imply complex text encodings, which in turn entail high costs for the manual indexing/authoring task. This well-known problem, which hampers the development of large powerful systems, has been tackled with a set of specific linguistic tools, first experimented in the Esprit II project Nomos and subsequently developed in research and development projects carried out in the Finsiel Group. These tools – devoted to the automatic extraction from texts of the information structures considered in the retrieval model – use shallow techniques amenable to effective large-scale text processing in the legal domain, in order to overcome the state-of-the-art limitations of traditional `deep' NLP techniques.

This article presents an overview of our approach, providing a general description of the representation model and processing tools, and concentrating primarily on the representation features and search improvements related to the use of the functional context information.

Language: English

Document Type: Regular paper

Affiliations: 1: FINSIEL S.p.A., Via Carciano, 4 - 00131, Roma, Italy E-mail: e.pietrosanti@finsiel.it 2: Via del Ponte di Piscina Cupa, 43 - 00128, Roma, Italy E-mail: b.graziadio@finsiel.it

Publication date: 1999-01-01

• In this: publication
• By this: publisher
• In this Subject: Computer Science ,  Law
• By this author: Pietrosanti E. ;  Graziadio B.

Website © Publishing Technology. Article copyright remains with the publisher, society or author(s) as specified within the article.

• Terms and conditions
• Privacy policy
• Information for advertisers