SOFTWARE DEFECT PREDICTION TRENDS: A BIBLIOMETRIC ANALYSIS OF MACHINE AND DEEP LEARNING

This study provides a comprehensive bibliometric mapping of global research trends and emerging frontiers in Software Defect Prediction (SDP), emphasizing the integration of machine learning (ML) and deep learning (DL) approaches. Unlike previous bibliometric surveys that focused narrowly on metric-based or short-term analyses, this work offers a broader and more integrated perspective on the intellectual evolution, collaboration patterns, and thematic directions in SDP research. Using data retrieved from the Scopus database and analyzed through Bibliometrix and VOSviewer, the study systematically applied the PRISMA protocol to ensure transparency and replicability. A total of 1,549 publications were examined, revealing a steady increase in scientific output dominated by China, India, and the United States. Thematic and keyword analyses identified five core clusters that trace the paradigm shift from traditional statistical models to advanced ML- and DL-driven predictive frameworks. Emerging topics such as transfer learning, cross-project prediction, and explainable AI (XAI) were identified as promising frontiers shaping the next phase of software quality prediction research. Beyond mapping academic progress, this study contributes strategic insights for researchers seeking to identify research gaps, industry practitioners developing intelligent defect prediction tools, and policymakers designing AI-driven software quality initiatives