AI-Based Supply Chain Sentiment and Risk Forecasting Using Real-Time News Analytics

Abstract

Supply chain disruptions have become increasingly frequent because of geopolitical events, climate change, government regulations, and market volatility. Conventional supply chain risk management systems are mainly dependent on historical information and static variables, making them less capable of recognizing potential risks in real-time. This paper proposes an AI-powered supply chain sentiment analysis and risk prediction system that uses real-time news information to detect possible risks at an early stage. The proposed system uses real-time news articles from various news channels and transformer-based sentiment analysis to determine the underlying sentiment linked to supply chain events. The sentiment values are then converted to dynamic risk scores, making it possible to monitor risks continuously. At the same time, time-series forecasting methods are used to forecast the short-term trend of risk scores based on sentiment-driven values. The experimental outcome shows that the proposed model is capable of accurately detecting negative cues linked to possible risks while performing well on neutral and positive news articles.