Automating Testing Case Generation and Prioritization Using Generative AI

Naufal, Muhammad (2025) Automating Testing Case Generation and Prioritization Using Generative AI. Undergraduate thesis, Politeknik Negeri Jember.

	Text (Abstract) Abstract.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (14kB)
	Text (Bab 1 Pendahuluan) Bab 1 Pendahuluan.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (300kB)
	Text (Laporan Lengkap) Laporan Lengkap.pdf Restricted to Registered users only Download (2MB) | Request a copy
	Text (Daftar Pustaka) Daftar Pustaka.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (394kB)

Abstract

This study proposes an innovative approach to automating software test case generation and prioritization using advanced artificial intelligence (AI) techniques, specifically Retrieval-Augmented Generation (RAG), Large Language Models (LLMs), and LangChain, integrated with ChromaDB for efficient code embedding management. Traditional test case generation methods are often time- consuming, error-prone, and struggle to adapt to evolving codebases, leading to incomplete test coverage and inefficiencies. The proposed system addresses these challenges by leveraging RAG to dynamically retrieve relevant code context from ChromaDB and generate context-aware, prioritized test cases tailored to the programming language and code structure. A user-friendly web-based interface enables developers and QA engineers to input code, select functions, and generate test cases compatible with JUnit 4.12 or JUnit 5, streamlining integration into existing workflows. Evaluated through controlled experiments, the system achieves up to 97% code coverage, significantly reducing manual effort and improving testing efficiency. By combining retrieval and generation capabilities, this research offers a scalable, adaptable solution to enhance software quality and accelerate development cycles, contributing to advancements in AI-driven software testing.

Actions (login required)

View Item