NBEP: Nature-Based Ensemble Prediction Framework for Intelligent Bug Report Classification

Abstract

In modern software development, the rapid accumulation of bug reports demands intelligent systems capable of accurate classification and prioritization. This paper introduces NBPMBR, a Nature-Based Prediction Model of Bug Reports that leverages ensemble machine learning methods integrated with nature-inspired optimization algorithms such as Genetic Algorithms (GA), Particle Swarm Optimization (PSO), and Ant Colony Optimization (ACO). These algorithms are strategically embedded within an ensemble framework comprising models like Support Vector Machines, Random Forests, and Logistic Regression to enhance the model’s predictive performance and robustness.NBPMBR is trained on real-world bug datasets, employing advanced text preprocessing, TF-IDF feature extraction, and dynamic training-testing splits to simulate practical software environments. Experimental results show that the proposed model significantly outperforms traditional classifiers, achieving high accuracy, precision, and recall. By automating bug classification with nature-inspired optimization and ensemble voting, NBPMBR streamlines the software maintenance lifecycle and improves defect resolution efficiency, offering a scalable and adaptable solution for real-time bug report prediction.