cyberquantic logo header
EN-language img
FR-language img
breadcrumbs icon
Knowledge Representation

Ontologies & Taxonomies

Semantic analysis of legal documents

Semantic analysis of legal documents

For:
Legislative institutions, management institutions
Goal:
Other
Problem addressed
Machine understanding of the meaning of legal documents.
The acquisition of semantic visual images of documents; the detection of
contradictions and inaccuracies in legal documents describing similar objects of
law for the task of classifying documents; quick document comprehension; and
analysis of the consistency of the legal base.
Scope of use case
Semantic analysis of legal documents in the course of their development,
verification and improvement.
Description
The most important condition for ensuring the integrity of
the legal base is the identification and elimination of
contradictions, which are often found when using existing or
developing new legal acts and documents relating to various
aspects of the same objects of law. To solve this problem, a
software tool has been developed to control the integrity of
the legal base in the development and use of legal
documents.
The software tool accepts an initial set of legal documents as
input, and performs its syntactic and semantic analysis. For
parsing and determining the morphological characteristics
of words, a grammar dictionary of the Russian language, the
WordNet thesaurus, and the SyntaxNet library for
determining syntactic relationships, are used.
The result of the analysis of the document is a weighted
semantic image of the document, which is a semantic
network of concepts and relations between them.
A fragment of this network related to documents and
concepts of interest to an expert is described as a set of OWL
expressions - an ontology of documents. As the base of this
ontology - ontology for legal knowledge representation -
LKIF-Core is used.
On the resulting ontology, logical inference by using the
JFACT++ reasoner is performed to identify possible
inconsistencies and notify the expert.
At the same time, the generated semantic image of the
document is visualized in various ways, in which vertex
incidence (degree of detail of the description of a concept in
the document), the weight of the edges (the importance of
the relationship between the concepts), and paths (structure
of the definition of concept in the document) are presented
using visual effects. The expert can therewith quickly
comprehend the document, identify documents that are
similar in meaning, and identify possible problem places in
the definition of legal concepts and relations between them.
understand frame img
Ontologies & Taxonomy
Interested in the same or similar project?
Submit a request and get a free project evaluation.