Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.29742 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914613328609280 |
|---|---|
| author | Ju, Yeong-Joon Lee, Seong-Whan |
| author_facet | Ju, Yeong-Joon Lee, Seong-Whan |
| contents | Deploying Large Language Models (LLMs) for regulatory compliance demands rigorous traceability via comprehensive citations across multi-tiered authority structures. Unlike traditional multi-hop or legal QA, this task requires structured procedural lookups and evidence-set closure rather than entity resolution or case-law reasoning. Existing RAG systems struggle here due to flattened citation edges, fragmented retrieval expansions, and fragile post-hoc attribution. We formalize Regulatory Compliance QA with RegOps-Bench, a novel benchmark featuring an Operational Knowledge Graph derived from complex national R\&D regulations. To address these bottlenecks, we propose RefWalk, a unified framework driven by a shared topic anchor. RefWalk traverses cross-document citations, fuses multi-view candidates via max-based aggregation, and enforces per-rule attribution to explicitly map claims to sources. We establish a strong baseline with substantial improvements in retrieval recall and citation accuracy. Finally, a contrastive evaluation on a U.S. health compliance dataset (HIPAA) reveals that existing systems exhibit saturation on flat-structure rules, underscoring the need for RegOps-Bench. Our code is available at https://github.com/yeongjoonJu/RefWalk. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_29742 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Citation-Closure Retrieval and Per-Rule Attribution for Real-World Regulatory Compliance Question Answering Ju, Yeong-Joon Lee, Seong-Whan Artificial Intelligence Deploying Large Language Models (LLMs) for regulatory compliance demands rigorous traceability via comprehensive citations across multi-tiered authority structures. Unlike traditional multi-hop or legal QA, this task requires structured procedural lookups and evidence-set closure rather than entity resolution or case-law reasoning. Existing RAG systems struggle here due to flattened citation edges, fragmented retrieval expansions, and fragile post-hoc attribution. We formalize Regulatory Compliance QA with RegOps-Bench, a novel benchmark featuring an Operational Knowledge Graph derived from complex national R\&D regulations. To address these bottlenecks, we propose RefWalk, a unified framework driven by a shared topic anchor. RefWalk traverses cross-document citations, fuses multi-view candidates via max-based aggregation, and enforces per-rule attribution to explicitly map claims to sources. We establish a strong baseline with substantial improvements in retrieval recall and citation accuracy. Finally, a contrastive evaluation on a U.S. health compliance dataset (HIPAA) reveals that existing systems exhibit saturation on flat-structure rules, underscoring the need for RegOps-Bench. Our code is available at https://github.com/yeongjoonJu/RefWalk. |
| title | Citation-Closure Retrieval and Per-Rule Attribution for Real-World Regulatory Compliance Question Answering |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2605.29742 |