Authors:
K. Pradeep, S. Sindhu

Addresses:
Department of Data Science and Business Systems, School of Computing, SRM Institute of Science and Technology, Kattankulathur, Chennai, Tamil Nadu, India.

Abstract:

Modern supply chains have made procurement risk management even more critical due to demand uncertainty, supply disruptions, and price volatility. Traditional methodologies may not be able to combine diverse data sources and deliver timely, operational insights. The proposed Hybrid Artificial Intelligence-based Procurement Risk Control is implemented for Procurement Risk Analysis. The suggested framework combines structured data, such as Demand Net Shortage (DNS), inter-site supply availability, open-market sourcing, and financial exposure indicators (such as IPDR and OPOR), with unstructured information in the form of buyer comments. An inference engine based on fuzzy logic is used to characterise uncertainty and categorise procurement risk into meaningful categories, enabling effective decision-making in uncertain situations. Simultaneously, a Natural Language Processing (NLP) component takes textual inputs, applies Term Frequency-Inverse Document Frequency (TF-IDF) vectorisation, and a Logistic Regression classifier to present qualitative risk insights. To increase the system's reliability during the first stages of deployment, a confidence-based switching mechanism is presented that dynamically balances rule-based logic and machine learning predictions. Continuous retraining allows adaptive learning and improved performance. The experiment shows that the hybrid method is more efficient and successful in procurement risk identification, decision-making, and supply chain resilience. Its scalable architecture, computational efficiency, and corporate implementation provide a comprehensive approach to intelligent procurement risk management. 

Keywords: Procurement Risk; Scalable Structure; Operational Insights; Logistic Regression; Fuzzy Logic; Textual Inputs; Financial Risk Modelling; Self-Learning System.

Received on: 28/10/2025, Revised on: 19/12/2025, Accepted on: 04/02/2026, Published on: 19/08/2026

DOI: 10.69888/FTSFDS.2026.000743

FMDB Transactions on Sustainable Finance and Data Science, 2026 Vol. 1 No. 3, Pages: 172-186

• Views : 25
• Downloads : 5

Download PDF