TRANSPARENT ELECTRONIC PUBLIC PROCUREMENT MONITORING SYSTEM THROUGH THE USE OF MACHINE LEARNING AND ANALYTICS

Abstract

E-procurement can largely enhance the processes of public procurement by allowing more competition, reducing corporate transaction costs through digital technologies, and instilling confidence in decision-making within public institutions. These systems, when integrated with machine learning and analytics, could go one step further in the detection of fraudulent behaviors, thus enabling data-driven decisions.

This paper discusses the adoption of sophisticated technologies, such as Machine Learning and Analytics with the view to increase efficiency, accountability, and transparency in public procurement in Zambia. It also analyzes the current electronic procurement system operated by the Zambia Public Procurement Authority (ZPPA) in order to identify the systems shortcomings in which Machine Learning and Analytics could provide substantial support. The proposed system makes use of Machine Learning for corrupt-risk analysis of market prices, classification of tenders and bids, data analytics for predictive procurement strategies, and better decision-making.

This research goes on to assess such technologies in terms of their impact on the performance and maturity of electronic public procurement processes and consequently marks those as potential solutions to inefficiencies and inequities. Key highlights reviewed of 178 Procuring Entities, 5,356 Tenders, and 8,709 Bids bring out some interesting patterns: 35.68% of the tenders were single bid closing, 6.05% had huge price gaps in bids submitted by only two suppliers, and 3.58% had price outliers in cases of three or more bids. These anomalies were identified through Machine Learning-driven methods of percentage difference and interquartile range analysis to make sure that procurement practices are closely monitored and evaluated at large.

Beyond uncovering system inefficiencies, the study underscores the broader impact of Machine Learning and Analytics in optimizing sourcing, transaction management, payment processes, and bid evaluations. These technologies have been pivotal in elevating the operational maturity of Zambia’s e-Procurement system. The findings demonstrate that a strategically aligned, data-driven adoption of ML and Analytics not only delivers notable performance improvements but also enhances trust, fairness, and integrity within public procurement operations.