A recommendation system for predicting dependencies among software changes insights from an empirical study on OpenStack

Keyphrases

• Actionable Recommendations 33%
• Agentic AI 100%
• AI-driven Solutions 33%
• Average AUC 33%
• Brier Score 33%
• Build Failure 33%
• Builders 33%
• CI/CD pipeline 33%
• Code Review 33%
• Dependency Management 33%
• Engineering Techniques 33%
• Exact Couple 33%
• Failure Feature 33%
• Incomplete Features 33%
• Large-scale Software System 66%
• Machine Learning Approach 33%
• Machine Learning Models 33%
• Modern Software System 33%
• Multiple Components 33%
• OpenStack 100%
• Practitioner Tools 33%
• Practitioner-researcher 33%
• Prompt Engineering 33%
• Recommendation System 100%
• Regex 33%
• ROC-AUC 33%
• Search Tools 33%
• Semi-automated Approach 33%
• Software Change 100%
• Software System 33%
• System Dependency 33%
• Time Change 33%

Computer Science

• Artificial Intelligence 75%
• Machine Learning-Based Approach 25%
• Modern Software 25%
• Multiple Component 25%
• Prompt Engineering 25%
• Software Change 100%
• Software System 100%