Papers by ShengYong Ding
DiFRa: A Unified Framework for Harmonizing Semantic Diversity and Factual Consistency in Question-Answer Generation (2026.findings-acl)
Copied to clipboard
| Challenge: | Question-Answer Generation (QAG) is essential for domain-specific large language models post-training. |
| Approach: | They propose a framework that balances semantic diversity and factual consistency . they propose entropy and consistency scores that harmonize the trade-off between diversity and correctness . |
| Outcome: | The proposed framework outperforms baseline models in generating diverse QA pairs . the proposed framework harmonizes semantic entropy and consistency scores to quantify trade-off between diversity and correctness. |
Evaluation of Text-to-Image Generation from a Creativity Perspective (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Recent studies have assessed the creativity of T2I models, but little has been done on the quality of generated images and image-text alignment. |
| Approach: | They define the creativity of T2I models and propose metrics to test reliability . they also develop a pipeline capable of transforming existing image-text datasets into benchmarks . |
| Outcome: | The proposed method tests the reliability of the metric and a fully automated pipeline capable of transforming image-text datasets into benchmarks tailored for evaluating creativity. |