D-SCoRE: Document-Centric Segmentation and CoT Reasoning with Structured Export for QA-CoT Data Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915779416424448 |
|---|---|
| author | Zhou, Weibo Li, Lingbo Liang, Shangsong |
| author_facet | Zhou, Weibo Li, Lingbo Liang, Shangsong |
| contents | The scarcity and high cost of high-quality domain-specific question-answering (QA) datasets limit supervised fine-tuning of large language models (LLMs). We introduce $\textbf{D-SCoRE}$, a training-free framework that leverages LLMs and prompt engineering to automatically generate diverse, rich QA datasets with Chain-of-Thought (CoT) from arbitrary textual sources. By integrating $\textbf{D}$ocument-centric processing, $\textbf{S}$egmentation, $\textbf{Co}$T $\textbf{R}$easoning, and structured $\textbf{E}$xport - along with multi-dimensional controls such as semantic role transformation, question type balancing, and counterfactual augmentation - D-SCoRE produces tailored QA pairs with enhanced diversity and relevance. LLMs fine-tuned on D-SCoRE-generated datasets outperform those trained on human-annotated QA data across most evaluated domains. Its efficiency and scalability enable rapid, high-performance domain-adaptive fine-tuning on consumer-grade hardware, generating over 1,100 high-quality QA pairs per GPU-hour end-to-end. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_01309 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | D-SCoRE: Document-Centric Segmentation and CoT Reasoning with Structured Export for QA-CoT Data Generation Zhou, Weibo Li, Lingbo Liang, Shangsong Computation and Language Artificial Intelligence The scarcity and high cost of high-quality domain-specific question-answering (QA) datasets limit supervised fine-tuning of large language models (LLMs). We introduce $\textbf{D-SCoRE}$, a training-free framework that leverages LLMs and prompt engineering to automatically generate diverse, rich QA datasets with Chain-of-Thought (CoT) from arbitrary textual sources. By integrating $\textbf{D}$ocument-centric processing, $\textbf{S}$egmentation, $\textbf{Co}$T $\textbf{R}$easoning, and structured $\textbf{E}$xport - along with multi-dimensional controls such as semantic role transformation, question type balancing, and counterfactual augmentation - D-SCoRE produces tailored QA pairs with enhanced diversity and relevance. LLMs fine-tuned on D-SCoRE-generated datasets outperform those trained on human-annotated QA data across most evaluated domains. Its efficiency and scalability enable rapid, high-performance domain-adaptive fine-tuning on consumer-grade hardware, generating over 1,100 high-quality QA pairs per GPU-hour end-to-end. |
| title | D-SCoRE: Document-Centric Segmentation and CoT Reasoning with Structured Export for QA-CoT Data Generation |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2508.01309 |