Skip to content

Graph Reasoning Over Enterprise Innovation Networks for Industrial Investment Risk Assessment

Sep 2026 · The Journal of Applied Engineering and Technologies · 0 citations · 30 references

Abstract

Regional industrial funds invest in technology-oriented manufacturing firms, advanced materials companies, intelligent equipment producers, energy enterprises, and digital infrastructure providers. Investment risk is influenced not only by financial indicators but also by patent quality, R&D collaboration, government subsidies, customer concentration, supply-chain stability, executive experience, and related-party transactions. This study proposes a graph reasoning model over enterprise innovation networks for industrial investment risk assessment. The method builds a knowledge graph linking target firms, shareholders, patents, research partners, suppliers, customers, subsidy programs, financing rounds, executives, litigation events, and revenue records. A graph neural network is used to represent innovation capability and business dependency, while knowledge inference identifies hidden investment risks from weak patent commercialization, subsidy dependency, unstable key customers, supplier disruption, and shareholder debt exposure. Experiments are conducted on a regional industrial investment dataset containing 12,800 candidate enterprises, 96,000 patents, 38,400 R&D collaboration links, 520,000 supplier-customer edges, 18,600 subsidy records, 24,300 financing events, and 3,940 post-investment risk cases over 54 months. The proposed model shortens median risk-identification time from 132 days to 57 days after investment screening compared with a due-diligence checklist baseline. It discovers 4,260 innovation-commercialization risk paths and 1,780 subsidy-dependency risk chains. Portfolio simulation shows that graph-informed screening reduces expected impaired investment exposure by 118 million RMB during the validation period. Full quarterly assessment of candidate enterprises is completed in 8.7 minutes. The results demonstrate that enterprise innovation network reasoning can provide interpretable support for industrial fund investment risk assessment.

View source

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.