Traceable Knowledge Graph Reasoning Enables LLM-Assisted Decision Support for Industrial VOCs in the Steel Industry

2026-05-26

AI Summary

This study addresses the challenge of fragmented and unstructured knowledge regarding volatile organic compound (VOC) control in the steel industry, which often leads general large language models to generate hallucinations on infrequent industrial queries. To tackle this, the authors propose Chat-ISV, a multi-agent question-answering system that uniquely integrates traceable knowledge graph reasoning with a multi-agent architecture. The system constructs a Neo4j-based knowledge graph comprising 27,180 nodes and 81,779 edges through prompt-constrained extraction, enhanced by topological optimization that reduces the proportion of isolated nodes from 57% to 4.08%. It further incorporates multi-agent routing and source-traceable retrieval mechanisms. In expert blind evaluations, Chat-ISV achieved a score of 1.69 out of 2.00, with a precision of 96.93%, recall of 72.63%, and an F1-score of 0.830, establishing a novel, high-confidence paradigm for industrial environmental informatics.

Abstract

Key knowledge for steel-industry volatile organic compounds (VOCs) governance is scattered across unstructured scientific literature, making it difficult to integrate process, pollutant, and control-technology evidence and increasing the risk of hallucination when general large language models (LLMs) answer low-frequency industrial questions. Here we developed Chat-ISV, a knowledge graph (KG) enhanced multi-agent Q&A system that parses a curated steel-industry VOCs literature corpus, constructs a Neo4j KG with 27180 nodes and 81779 semantic edges, and combines prompt-constrained extraction, chunk-centered topology optimization, multi-agent routing, source-backtracking retrieval, local literature retrieval, open-domain knowledge access, and interactive subgraph visualization. Benchmark tests and 400 expert blind evaluations showed that topology optimization reduced isolated nodes from 57% to 4.08% and that Chat-ISV achieved high factual reliability, with 96.93% precision, 72.63% recall, an F1-score of 0.830, and a mean score of 1.69/2.00. By converting fragmented environmental-engineering literature into traceable, queryable, and decision-support-oriented knowledge, Chat-ISV establishes a scalable environmental-informatics paradigm for reliable LLM deployment and intelligent pollution-control decision support in specialized industrial domains.

Problem

• VOCs governance
• knowledge fragmentation
• large language models
• industrial decision support
• environmental informatics

Innovation

• knowledge graph reasoning
• topology optimization
• multi-agent routing
• source-backtracking retrieval
• LLM-assisted decision support