Treatment of Legal Sentences Including Itemized and Referential Expressions – Towards Translation into Logical Forms

Abstract

Abstract. This paper proposes a framework for analyzing legal sentences including itemized or referential expressions. Thus far, we have developed a system for translating legal documents into logical formulae. Although our system basically converts words and phrases in a target sentence into predicates in a logical formula, it generates some useless predicates for itemized and referential expressions. We propose a front end system which substitutes corresponding referent phrases for these expressions. Thus, the proposed system generates a meaningful text with high readability, which can be input into our translation system. We examine our system with actual data of legal documents. As a result, the system was 73.1% accurate in terms of removing itemized expressions in a closed test, and 51.4% accurate in an open test.

Treatment of Legal Sentences Including Itemization Written in Japanese, English and Vietnamese-Towards Translation into Logical Forms-

“…In the preceding study (Kimura et al 2009), focusing on Japanese legal sentences, we showed that the system worked well using some simple regular expressions. In this paper, we apply it to the Vietnamese Law on Enterprises and the United States Code: Title 39-Postal Service.…”

“…These words should be replaced with appropriate phrases before the process of translation. Since Japanese legal documents have strict rules concerning its description and modification, we succeeded to extract conventional expressions in the documents by some regular expressions (Kimura, Nakamura, and Shimazu 2009). In order to investigate whether the proper method depends on the language or the nation establishing laws, we try to apply our approach to the English and Vietnamese versions of Vietnamese Law and the United States Code.…”

Acquisition of hyponymy relations for agricultural terms from a Japanese statutory corpus

“…Therefore, we do not rely on a syntactic parser and minimize the use of natural language processing (NLP) tools. Kimura et al [13] acquired knowledge from itemized expressions in legal texts. Their experimental result showed that only a few surface patterns following itemization successfully extracted itemized expressions and represented semantics.…”

A Two-Phase Framework for Learning Logical Structures of Paragraphs in Legal Articles

“…Authors presented a method to evaluate the importance of words in the task and found that words that have strong relations to the logical structure of law sentences are very important for machine learning models. Kimura et al [2009] focus on dealing with legal sentences including itemized and referential expressions. They presented a rule-based method for substituting referent phrases.…”