CoRT: Transformer-based code representations with self-supervision by predicting reserved words for code smell detection

Computer Science

• Semantic Feature 100%
• Detection Performance 100%
• Structural Feature 66%
• Multiple Task 33%
• Source Codes 33%
• Machine Learning-Based Approach 33%
• Supervised Learning 33%
• Annotation 33%
• Baseline Model 33%
• Representation Learning 33%
• Self-Supervised Learning 33%
• Intermediate Representation 33%
• Generalizability 33%