World-first bughunter: ARC DECRA Fellow develops AI that can detect defects faster

Monash researcher Dr Chakkrit Tantithamthavorn has developed a world-first practical and explainable AI to help developers predict and find bugs faster – and understand why a program is likely to be defective in the future.

‘We are living in a software-driven society. However, software defects and technology glitches are very annoying and expensive, and they’re hard to detect and prevent. Errors in safety-critical systems could result in serious injuries and even death. We want to prevent this as much as possible. ’ said Dr Tantithamthavorn.

His recent work is the first to develop AI that can accurately predict which lines of code will be defective in the future. By leveraging recent advances in explainable AI, his project can accurately locate 61% of the actual defective lines, which is 91% better than the state-of-the-art approaches.

‘Imagine you’re a developer working on a software project with million lines of code. Developers have to spend years and years reviewing and testing every single line of code, which is very time-consuming and inefficient. This leads to project overruns and high costs,’ explained Dr Tantithamthavorn.

‘My project will help developers find defects faster, saving a significant amount of time and effort when identifying the precise locations of software defects.’

Understanding why files are ‘buggy’

Dr Tantithamthavorn and his Monash’s PhD student, Jirayus Jiarpakdee, are also the first to develop AI that can explain why a file is defective and what developers can do to mitigate the risks moving forward.

‘My ultimate goal is to provide actionable advice to developers on what they should and shouldn’t do so software products can be improved in the future.’ said Dr Tantithamthavorn.

In a recent study, 65% of participants considered this explainable defect prediction tool is useful. Dr Tantithamthavorn’s work is also published in the top-tier software engineering journal, IEEE Transactions on Software Engineering, Impact Factor 6.11.


1. Jirayus Jiarpakdee, Chakkrit Tantithamthavorn, Hoa Khanh Dam, John Grundy, An Empirical Study of Model-Agnostics Techniques for Defect Prediction Models, IEEE Transactions on Software Engineering (TSE), 2020.

2. Supatsara Wattanakriengkrai, Patanamon Thongtanunam, Chakkrit Tantithamthavorn, Hideaki Hata, and Kenichi Matsumoto, Predicting Defective Lines Using a Model-Agnostic Technique, IEEE Transactions on Software Engineering (TSE), 2020.