火灾爆炸事故溯源知识图谱数据集

针对“基于知识图谱及多源信息的火灾爆炸事故溯源与重构系统”专题的研究任务：构建火灾爆炸事故调查知识图谱，典型起火物、起火原因、引火源等火灾要素≥3种，数据关系扩展图谱层级≥3层，在2022年11月-2025年7月期间，设计搭建了火灾爆炸调查知识图谱本体模型，开展了原始数据（包含火灾爆炸事故案例数据、火灾爆炸调查知识数据、火灾要素数据、火灾爆炸事故重构资产信息数据）采集和数据预处理工作（包括文本提取、格式清洗和内容校验）。根据抽取和人工审核结果，数据集包含5类结构化CSV文件：火灾爆炸调查知识图谱本体模型、火灾爆炸调查知识图谱、事故知识图谱、事故涉及物品知识图谱、火灾爆炸事故重构资产知识图谱。数据质量控制通过官方来源权威性保障及领域专家参与审核构建实现，支持事故调查中的因果推理与数字重构，对提升火灾爆炸事故溯源精准度具有重要应用价值。

For the research task of the special topic 'Fire and Explosion Accident Traceability and Reconstruction System Based on Knowledge Graph and Multi-source Information': Construct a knowledge graph for fire and explosion accident investigation, with no less than 3 types of fire-related elements including typical ignition materials, ignition causes, ignition sources, etc., and the number of extended layers of data relationship in the graph is no less than 3. Between November 2022 and July 2025, the ontology model of the fire and explosion investigation knowledge graph was designed and established. Data collection and preprocessing were then conducted, covering original data such as fire and explosion accident case data, fire and explosion investigation knowledge data, fire element data, and asset information data for fire and explosion accident reconstruction; the preprocessing work included text extraction, format cleaning and content verification. According to the extraction results and manual review, the dataset includes 5 categories of structured CSV files: the ontology model of fire and explosion investigation knowledge graph, fire and explosion investigation knowledge graph, accident knowledge graph, knowledge graph of items involved in accidents, and asset knowledge graph for fire and explosion accident reconstruction. Data quality control is achieved by guaranteeing the authority of official sources and conducting review with the participation of domain experts. This dataset supports causal reasoning and digital reconstruction in accident investigation, and holds important application value for improving the accuracy of fire and explosion accident traceability.

搜集汇总

背景与挑战

背景概述

火灾爆炸事故溯源知识图谱数据集是一个包含5类结构化CSV文件的数据集，数据量1.14GB，旨在构建火灾爆炸事故调查知识图谱，支持事故调查中的因果推理与数字重构，提升溯源精准度。该数据集由应急管理部天津消防研究所创建，属于国家重点研发计划项目的一部分。