Tool support for domain analysis of the software specification in natural language

Yoichi Omori, Keijiro Araki

Research output: Chapter in Book/Report/Conference proceedingConference contribution


Most software specifications are discussed on natural language among stakeholders. There is, however, a semantic gap between natural language and programming language, which sometime causes discrepancies between them and misunderstanding among stakeholders. Formal language compensates the gap so that realizes early verification in software development process. We developed a dictionary tool to support the translation from natural language to formal language. The tool provides functionalities those are easy registration of keywords to the dictionary and exhaustive marking of the keywords. The dictionary represents a map between equivalent keywords in natural language and formal language which gives defined semantics, and the tool enhance the coverage of the keywords in the formal specification. In short, the dictionary contains conceptual keywords and specific semantics in the problem; therefore, it corresponds to a domain of the system. A system generally belongs to multi problem domains and also contains multi sub domains, consequently some dictionaries should be dynamically alternated through the translation process. APIs related to dictionary operation are proposed and verified in this paper, and the dictionary class is introduced to map the relation among domains.

Original languageEnglish
Title of host publicationTENCON 2010 - 2010 IEEE Region 10 Conference
Number of pages6
Publication statusPublished - Dec 1 2010
Event2010 IEEE Region 10 Conference, TENCON 2010 - Fukuoka, Japan
Duration: Nov 21 2010Nov 24 2010

Publication series

NameIEEE Region 10 Annual International Conference, Proceedings/TENCON


Other2010 IEEE Region 10 Conference, TENCON 2010

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Electrical and Electronic Engineering


Dive into the research topics of 'Tool support for domain analysis of the software specification in natural language'. Together they form a unique fingerprint.

Cite this