AI-Assisted Debugging Tools for Enhancing Developer Productivity in Large Codebases Using Genetic Programming Algorithms

One of the key features of software design, mainly required in large scale software projects, is Debugging because the process of identifying and correcting errors in the large structured programmable code takes a lot of time, efforts and may also involve several errors. Existing_ad hoc_debugging techniques although useful are much less successful in scaling up with size and complexity of code. AI has brought new opportunities for development of new approaches to improving the debugging tools, in particular Genetic Programming algorithms. In this paper, we look at how developers can be assisted by AI tools for purposes of debugging with GP algorithms in large code systems. GP, a subclass of evolutionary algorithms, is capable of automatically developing solutions for such applications through the implementation of the process that mimics natural selection. Through applying GP algorithms, the bug detection and correction can be done with a computationally aided approach hence taking less time to discovery and possibly a better solution than traditional methods. The paper presents the use of GP in AI provided debugging, describes the approach and offers conclusions based on conducted experiments, which show the potential of these tools. The approach presented in this work reveals a more efficient and scalable approach towards debugging large code bases; thus, AI-based debugging is best suited for the modern development world.

Перевод пока недоступен