Text mining for supply chain risk management

• Main Author: Liu, Xi
• Other Authors: Chen Lihui
• Format: Final Year Project
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/64180
• Institution: Nanyang Technological University

Within close-interconnected globalized workplace, supply chain management is encountering risks from all aspects. One local event may impose a regional or even global impact. It has been a critical factor for a company to succeed that it could make the prompt response to the unexpected event and potential risk. Industries have always been aspiring to minimize the risks of operation with preventive measures to alleviate situation before it goes the worse end seriously. However, with tons of information generated from different source, supply chain managers demand a tool to track and evaluate the relationship among the event, specific industry and enterprises. Therefore, a web-based application is developed in this study enabling to estimate events’ potential consequence based on historical news articles and mitigate the risk with prepared solutions. In this study, we have combined data analytics, crawling engine and visualization techniques for presenting our supply chain risk management framework. Text mining has been shown to be a powerful approach for data analytics and trend predictions. It could identify the pattern among large volume of historical data; it could be used to extract useful information based on initial requirement and present in a readable and intuitive visualization form. In this study, we have employed text mining as an application to significantly increase supply chain efficiency by responding faster to unexpected event in daily operation. This report involves the literature review of text mining, risk management, and discusses the applications of text mining on supply chain risk management. It documents the process to design and program the web-based application for event-driven supply chain risk management tool. The algorithm design is elaborated with figures and tables; the visualizations of results are presented for case studies; extensive trouble shooting and debugging are executed for program optimization. Further explorations and recommendation are discussed for future developments.