Power Enterprise Supply Chain Risk Evaluation and Design of Early Warning System Based on Semantic Clustering

Aiming at the problem that the risk early warning effect of traditional power enterprise supply chain is not good, a power enterprise supply chain risk evaluation and early warning model based on semantic clustering is constructed. Firstly, SAT word vector model and K-means clustering algorithm were used to obtain risk word vector and semantic clustering respectively. Then, on the basis of TextRank word graph model, semantic clustering information is introduced to extract power supply chain risk evaluation indexes. Finally, the risk indexes are input into the gray prediction model for risk evaluation and early warning. The example analysis shows that the model can accurately warn capital risk and customer satisfaction reduction risk, and the risk warning values are 92.3% and 86.2 %, respectively. The risk warning accuracy of this model is up to 99.65%, which is higher than that of traditional risk warning model. It shows that the model can realize power enterprise supply chain risk evaluation and early warning, and meet the system design requirements.