Software Bug Tracker using Deep Learning
K. C. Mohanraj, M. Prasanth, S. Rafith, A. Sriman · 2024
Software bug prediction is the process of locating and predicting potential issues in a software project by using a range of techniques before they become costly and cause disruptive issues. The 'Deep Learning based Software Bug Tracker' is a novel approach presented in this research study. The primary goal of the proposed system is to revolutionize traditional bug tracking systems by automatically identifying, categorizing, and prioritizing software defects through the use of cutting-edge deep learning algorithms. "Software bug prediction" is the technique used for determining which software components have the highest likelihood of containing problems. To detect software flaws, the proposed system primarily employs two predictive modeling techniques: Learning to Rank (LTR) and Linear Regression (LR). It entails obtaining information from relevant and historical defect sources, preprocessing the information to clean and convert it, selecting features to identify important signals, and teaching Learning to Rank (LTR) and LR models. When put into practice, this method makes use of attributes to calculate the likelihood that software modules would have errors. The system's output controls how resources and testing strategies are used, enhancing the quality of software and development productivity. It is possible to effectively address proactive defect control during software development by using ongoing monitoring and adjustments that provide accuracy over time.