Development of Sentence Analysis Methods for Intelligent Regulation Management System

Abstract

The growing complication of regulations requires intelligent regulation management system, with support of information technology. In this research, we develop basic technology for analyzing rules and policies expressed in natural languages, in order to extract the regulatory attributes and automate data input process for regulation database. We analyze the frequency of verbs occurring in the regulation database, and focus on the most frequent types of regulations: ‘should submit,’ ‘should issue,’ and ‘may not’. Regex based approach and syntactic analysis approach are used for sentence structure analysis.