2005 AutoSemExtrInLawDocs
- (Biagioli et al., 2005) ⇒ C. Biagioli, E. Francesconi, Andrea Passerini, S. Montemagni, and C. Soria. (2005). “Automatic Semantics Extraction In Law Documents.” In: Proceedings Of The Tenth International Conference On Artificial Intelligence and Law. (ICAIL 2005)
Subject Headings: Legal Document.
Notes
Cited By
Quotes
Abstract
Normative texts can be viewed as composed by formal partitions (articles, paragraphs, etc.) or by semantic units containing fragments of a regulation (provisions). Provisions can be described according to a metadata scheme which consists of provision types and their arguments. This semantic annotation of a normative text can make the retrieval of norms easier. The detection and description of the provisions according to the established metadata scheme is an analytic intellectual activity aiming at classifying portions of a normative text into provision types and to extract their arguments. Automatic facilities supporting this intellectual activity are desirable. Particularly, in this paper, two modules able to qualify fragments of a normative text in terms of provision types and to extract their arguments are presented.
1. Introduction
The legal system usually suffers from scarce transparency which is caused by a non-systematic organization of the legal order. Law, in fact, is a normative and documentary unit of reference, hence the inability to obtain an analytical/systematic vision of a legal order itself, allowing to query a legal information system according to the content of each norm. This necessarily creates obstacles to the knowledge and upkeep of the legal order: from the uncertainty of the impact of new laws on the legal order in terms of coherency preservation, to the difficulties in norm accessing by both citizens and legal experts. For these reasons a more analytical unit of reference has been indentified in order to take a more organic view of the legal system [5, 7]. According to this point of view a normative text may be seen as a vehicle that contains and transports rules, or provisions, and the legal order as a set of rules rather than of laws. Recently, in Italy, the “Norme in Rete” (NIR) project (“Legislation on the Net”) has adopted such a perspective within a context of strategies aiming at creating a unique access point of normative documents on the Web with search and retrieval services. A “provision-centric” view of legal order indeed has been considered of primary importance to define strategies and tools for the upkeep of legal systems and to provide facilities to access norms. Such strategies are essentially based on the identification of the provisions within a normative text and on providing them with an explicit semantic description in terms of provision types and their arguments, namely the actions and the entities, with their roles, which are regulated by the provisions. This activity can be carried out both manually and automatically. This paper explores automatic methodologies for helping the human activities of detecting the typologies of provisions contained in a normative document and extracting the related arguments. This paper is organized as follows: in Section 2 the main components of a model of provisions are introduced; Section 3 presents the standards established by the NIR project and the tools developed to make their adoption easier; among such tools in Section 4 a module able to automatically describe a normative document in terms of the contained provision types is described and tested, as well as in Section 5 a module able to automatically extract the arguments of the detected provisions is shown and evaluated. Finally, in Section 6 some conclusions are discussed.
2. A Model of Provisions
The entire body of the law, with its articles and paragraphs, may be seen as a set of provisions, intended as rules and carried by linguistic acts, and therefore propositions, whether simple or complex, endowed with meaning 1 [25]. Basically, a normative text can be viewed according to a structural or formal profile, and a semantic or functional profile. Following this perspective, fragments of a normative text are, at the same time, paragraphs and provisions, according to whether they are seen from a formal or functional viewpoint.
...
References
,