New SR-BERT model automates conflict and duplicate detection in software requirements

Researchers have developed a novel transformer-based architecture called SR-BERT to automatically identify conflicting and duplicate software requirements. This approach utilizes Sentence-BERT and bi-encoders, combined with supervised multi-stage fine-tuning, to classify requirement pairs. Experiments across four datasets showed that SR-BERT performed best on larger datasets, demonstrating the effectiveness of transformer-based natural language processing strategies for automating conflict and duplicate detection in software engineering. AI

IMPACT Enhances automation in software development by improving the detection of requirement conflicts and duplicates.

RANK_REASON The item is an academic paper detailing a new model and methodology for a specific software engineering task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alphaXiv
• arXiv
• Bi-encoders
• CatalyzeX Code Finder for Papers
• CORE Recommender
• DagsHub
• Garima Malik
• Gotit.pub
• Hugging Face
• Influence Flower
• ScienceCast
• Sentence-BERT
• SR-BERT

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →