Leveraging Web Scraping in Predictive Modelling of Supply Risk Detection

• Main Author: Ng, Shi Ya
• Format: Final Year Project Report
• Language: English; English
• Published: Universiti Malaysia Sarawak (UNIMAS) 2022
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://ir.unimas.my/id/eprint/44092/1/Ng%20Shi%20Ya%20%2824pgs%29.pdf; http://ir.unimas.my/id/eprint/44092/2/Ng%20Shi%20Ya%20%28fulltext%29.pdf; http://ir.unimas.my/id/eprint/44092/

Supply risk is caused by interruptions to the flow of product in a supply chain whether it is economic, environmental, political, or ethical. These temporary events may cause a decrease in a supply chain’s performance in terms of inventory costs, production process, flexibility, and responsiveness. Despite these events faced by companies in the past few years, the consideration of the digital implementation in supply risk strategies is still not significant. This may be due to lack of budget and needing additional guidance to transition to more advanced technologies. For these purposes, a web scraping program is developed to identify the supply risks caused by these temporary events. It is a low-cost solution for the temporary events to be evaluated so that the risks can be detected in real-time for better decision-making. The important aspects for the temporary events are collected such as the date, description, title and link. The significance of the extracted output is evaluated with the topic modelling algorithm (LDA model algorithm) for the purpose of predictive supply risk. By collecting data from news sites, this system is aimed to provide data for predictive modelling so it can be used to detect patterns in supply risks to create and predict the probability of a temporary event occurring in the future