Predicting defects with latent and semantic features from commit logs in an industrial setting

Keyphrases

• Latent Feature 100%
• Semantic Features 100%
• Industrial Setting 100%
• Commit Logs 100%
• Process Metrics 50%
• Prediction Model 25%
• Machine Learning Algorithms 25%
• Detection Probability 25%
• Challenging Tasks 25%
• Topic Modeling 25%
• Feature Selection Methods 25%
• Lack of Data 25%
• Noisy Data 25%
• Naïve Bayes 25%
• Defect Prediction 25%
• Software Project 25%
• Probabilistic Model 25%
• Data Preprocessing 25%
• Software Defect Prediction 25%
• Matrix Factorization 25%
• Two-machine 25%
• Commit Messages 25%
• Level Process 25%
• Hidden Features 25%
• Multiple Resources 25%
• Log Filtering 25%
• Extreme Gradient Boosting(XGBoost) 25%

Computer Science

• Industrial Setting 100%
• Semantic Feature 100%
• Prediction Model 50%
• Defect Prediction 50%
• Extreme Gradient Boosting 50%
• Process Metrics 50%
• Machine Learning Algorithm 25%
• Feature Selection 25%
• Topic Modeling 25%
• Collecting Data 25%
• Defect Predictor 25%
• Software Project 25%
• Data-Preprocessing Technique 25%
• Software Defect 25%
• Multiple Resource 25%
• Matrix Factorization 25%
• Feature Extraction 25%