User centric AI-based similar defect recommendation system

Software defect is unwanted but unavoidable side product of software development process. Managing the defect from origin to fix is long cycle and consume lot of effort. Resolving the defect is one of crucial task in the software defect lifecycle. There are many defects which have been fixed in past can give hint for resolving the new defect. However, it is very tedious task to keep track of all the defects in system and get a list of similar or technically duplicate defect. In this paper, we have explained our novel work of retrieving lexically and technically similar defect based on user’s focus. We have leveraged technique of Natural language processing (NLP) to analyze the historical defects in the system and build a model to retrieve similar defects. We have taken user-centric approach to develop the software. This work also opens various possibilities for further research and apply AI on software defects to improve the process of software development overall.