Goals & Tasks
Our Mission
Our mission is to incorporate AISE into quality assurance processes, demonstrating that it can significantly enhance the ability to ensure software quality, with a focus on code dependability, and security.
Objectives
Use GenAI/LLMs to leverage syntatic and workflow knowledge for accurate bug and vulnerability identification.
Use novel GenAI/LLM-based techniques to systematically generate code patches according to best practices and principles of software development
Produce tools for automated test case generation, generating representative test cases on demand for fault/failure/attack/intrusion injection
Tasks
Literature related to AI-augmented tools for code analysis, bug detection, automated patch generation, and testing techniques
Identifying and selecting a set of representative software projects and existing tools for code analysis (from the systematic literature review that is going be conducted in Task 1, partially overlaping), bug detection, patch generation, and test cases generation
Development and implementation of LLM’s and deep learning techniques for code analysis and bug detection.
Develop AI-based methods for generating and validating patches for identified bugs (...), developing automated testing frameworks to validate these patches against a variety of test cases
GenAI techniques to create realistic scenarios that test the resilience of software under various fault conditions (...), identifying/using common fault types, developing algorithms to generate corresponding test cases
Generating test cases that simulate attacks and intrusions (...) integrate existing knowledge into GenAI solutions to create realistic and diverse attack scenarios that test the security defenses of software systems
Integration of tools and techniques (...) combining the code analysis, bug detection, patch generation, fault/failure test case generation, and attack/intrusion test case generation modules (...) into CI/CD pipelines