PREDICTION OF CLIENT SATISFACTION LEVELS WITH GHANA’S LAND ADMINISTRATION SYSTEM: USING A MACHINE LEARNING APPROACH

Abstract

Ghana’s first digital Land Administration System (LAS), known as the Client Service Access Unit (CSAU) managed by the Lands Commission (LC) is responsible for the provision of all land documentation procedures and is the one-stop shop for land transactions. However, since its creation as a result of Ghana’s Land Administration Projects (LAP 1 and LAP 2), there is no system available to determine the satisfaction levels of clients to maintain client loyalty to the system. This study aims to develop predictive models to forecast client satisfaction (CS) with the CSAU using machine learning (ML) algorithms including Random Forest (RF), Decision Tree (DT), Naive Bayes (NB) and Support Vector Machine (SVM) using Jupyter NoteBook Software and to determine the most important factors influencing client satisfaction with the services rendered by the CSAU, based on 199 responses. Among the ML algorithms, Random Forest and Decision Tree had the highest predictive accuracy of 85%, and among the eleven (11) latent factors used in the study, Value for Money had the strongest influence on CS with the services provided by the CSAU. These findings provide the basics of incorporating machine learning algorithms into real-time client satisfaction analysis dashboards to enable the service providers of the CSAU to identify areas for improvement and the overall satisfaction of clients with the services they provide. This study further contributes to the general world of research on leveraging advanced ML algorithms and Artificial Intelligence (AI) to solve problems in the field of service delivery assessment and client satisfaction.

Keywords: Land Administration Systems (LAS), Client Service Access Unit (CSAU), Client Satisfaction (CS), Classification, Machine Learning (ML) Algorithm