AlignSum: Data Pyramid Hierarchical Fine-tuning for Aligning with Human Summarization Preference
Han, Yang, Wang, Yiming, Wang, Rui, Chen, Lu, Yu, Kai
–arXiv.org Artificial Intelligence
Text summarization tasks commonly employ Pre-trained Language Models (PLMs) to fit diverse standard datasets. While these PLMs excel in automatic evaluations, they frequently underperform in human evaluations, indicating a deviation between their generated summaries and human summarization preferences. This discrepancy is likely due to the low quality of fine-tuning datasets and the limited availability of high-quality human-annotated data that reflect true human preference. To address this challenge, we introduce a novel human summarization preference alignment framework AlignSum. This framework consists of three parts: Firstly, we construct a Data Pymarid with extractive, abstractive, and human-annotated summary data. Secondly, we conduct the Gaussian Resampling to remove summaries with extreme lengths. Finally, we implement the two-stage hierarchical fine-tuning with Data Pymarid after Gaussian Resampling. We apply AlignSum to PLMs on the human-annotated CNN/DailyMail and BBC XSum datasets. Experiments show that with AlignSum, PLMs like BART-Large surpass 175B GPT-3 in both automatic and human evaluations. This demonstrates that AlignSum significantly enhances the alignment of language models with human summarization preferences.
arXiv.org Artificial Intelligence
Oct-1-2024
- Country:
- Asia
- China > Shanghai
- Shanghai (0.04)
- Middle East
- Iran > Tehran Province
- Tehran (0.04)
- UAE
- Abu Dhabi Emirate > Abu Dhabi (0.04)
- Dubai Emirate > Dubai (0.04)
- Fujairah Emirate > Fujairah (0.04)
- Iran > Tehran Province
- Singapore (0.04)
- China > Shanghai
- Europe > Ireland
- Leinster > County Dublin > Dublin (0.04)
- North America
- Canada > Ontario
- Toronto (0.04)
- United States
- California > San Francisco County
- San Francisco (0.04)
- New Jersey (0.04)
- Texas > Harris County
- Houston (0.04)
- California > San Francisco County
- Canada > Ontario
- Asia
- Genre:
- Research Report (1.00)
- Industry:
- Technology: