CONKER: CONtrastive knowledge enhanced retrieval for app review bug classification

Key Points

• Accuracy in bug classification improves with the use of contrastive knowledge algorithms.
• The system achieves a notable performance with 85% accuracy on a dataset of app reviews.
• Assessment involves natural language processing techniques and machine learning methodologies to enhance retrieval.
• This new approach may enable more efficient identification of app-related bugs, providing clearer user feedback.