Email: asiancomnet@usssociety.org
Website & IT Support: hi@aconf.org
[Online]Human-AI Comparative Evaluation of Test Case Generation in Scenario-Oriented Software Testing
Human-AI Comparative Evaluation of Test Case Generation in Scenario-Oriented Software Testing
ID：142
Submission ID：458 View Protection：ATTENDEE
Updated Time：2025-12-23 13:18:11 Hits：287
Online
Start Time：2025-12-30 12:05 （Asia/Amman）
Duration：10min
Session：[S9] Track 5: Emerging Trends of AI/ML » [S9-1] Track 5: Emerging Trends of AI/ML
Presentation File
Tips: The file permissions under this presentation are only for participants. You have not logged in yet and cannot view it temporarily.
Abstract
This research evaluates the effectiveness of AI-generated test cases (using GPT-4) against test cases constructed using conventional manual testing approaches in scenario-driven software testing. Manual test cases developed by applying established black-box testing methods, while GPT-4 generated test cases through structured prompts. Three scenarios—easy, moderate, and complex—used to conduct the evaluation under equivalent conditions. The primary comparisons in the present study evaluated defect detection capability, test coverage, efficiency of execution, and scenario relevance. The results indicate that AI-generated test cases provide better coverage, are faster to generate, and more effectively detect edge case faults; notably when evaluating the complex scenario. Procedural/manual testing found to be stronger in contextual reasoning and for safety critical interpretation. Overall, this research concludes that AI-generated testing is a complement to procedural/manual testing methods not a replacement. The results support a "hybrid" testing approach for modern software testing and quality assurance.
Keywords
Artificial Intelligence,GPT-4; Software Testing; Test Case Generation; Scenario-Driven Testing; Quality Assurance
Speaker
Rawan Habarneh
Student zarqa universitySubmission Author
All comments
Comment submit