Semi-Automatic Code Generation through Analyzing Natural Language Requirement Specifications with Generative AI 


Vol. 14,  No. 10, pp. 804-812, Oct.  2025
https://doi.org/10.3745/TKIPS.2025.14.10.804


PDF
  Abstract

Recently, generative AI tools have been increasingly utilized in the field of software development. However, the reliability of AI-generated outputs still cannot be guaranteed. In particular, current AI-based code generation often lacks a clear reflection of requirements and sufficient traceability of the design process. To address these limitations, we propose a mechanism that partially applies generative AI and applies metamodel-driven model transformation across the software process, from natural language requirements to skeleton code generation via UML design. We define transformation rules that are automatically applied on a Metamodel Transformation Engine. This approach may provide an automated development process that ensures the consistency of natural language requirements, the traceability of design artifacts, and the quality of the generated code. In future research, we aim to compare human-written code with AI-assisted code through the proposed mechanism. The objective is to verify which approach ensures higher software quality and, by doing so, enhances the reliability of AI-generated code while supporting the development of high-quality software.

  Statistics
  Cite this article

[IEEE Style]

Y. Jin, J. Kim, R. Y. C. Kim, "Semi-Automatic Code Generation through Analyzing Natural Language Requirement Specifications with Generative AI," The Transactions of the Korea Information Processing Society, vol. 14, no. 10, pp. 804-812, 2025. DOI: https://doi.org/10.3745/TKIPS.2025.14.10.804.

[ACM Style]

Yejin Jin, Janghwan Kim, and R. Young Chul Kim. 2025. Semi-Automatic Code Generation through Analyzing Natural Language Requirement Specifications with Generative AI. The Transactions of the Korea Information Processing Society, 14, 10, (2025), 804-812. DOI: https://doi.org/10.3745/TKIPS.2025.14.10.804.

Forms

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)

Old Journals

Indexing

All publications of TKIPS is indexed in DOI, EBSCO, Google Scholar, Crossref, and CrossCheck.

TKIPS is also selected as the Journal for Accreditation by NRF (National Research Foundation of Korea).

Related Journals