매일 아침, 어제의 AI를 한 통으로 정리해 보내드립니다메일로 받아보기

METAL LAB

ConstructCIE: A Dataset for Extracting Causal Information from Construction Accident Narratives

arXiv:2608.064952026-08-10

arXiv:2608.06495v1 Announce Type: new Abstract: Construction accident narratives contain rich causal information, but the evidence is often implicit, long-span, and distributed. We introduce ConstructCIE, a manually annotated dataset for Causal Information Extraction from OSHA construction accident reports. The dataset uses a hierarchical schema for accident types, causal factors, sub-causal factors, and supporting evidence spans. We evaluate supervised sequence taggers and instruction-tuned LLM

저자 · Hung Nguyen, Jaehoon Lee, Namgyun Kim, Kuan-Hao Huang

arXiv에서 원문 보기