Current Journal of Applied Science and Technology

Smart construction project management integrates advanced technologies, data analytics, and digital tools to support the planning, execution, and monitoring of construction projects. This study aimed to identify critical success factors (CSFs) influencing construction project success in Tanzania and to examine how machine learning can be integrated into project management to support a smart construction approach. A mixed-methods design was adopted. Quantitative data were collected through structured questionnaires from 100 construction experts and stakeholders, while qualitative data were obtained through interviews with 10 professionals. The study identified 41 CSFs, which were rated and used as input features for machine learning model development. Data pre-processing and feature selection were undertaken, and the dataset was expanded using generative adversarial networks. Logistic regression, random forest, support vector classifier, and decision tree models were evaluated using accuracy, precision, recall, F1-score, and receiver operating characteristic metrics. Results showed that the random forest and support vector classifier models had identical performance across the evaluation metrics, with approximately 0.97 for accuracy, 0.99 for precision, 0.97 for F1-score, and 1.00 for ROC; both outperformed the logistic regression and decision tree algorithms. The developed model was deployed in a Streamlit-based web application that allows users to rate CSFs and obtain project success predictions. The framework demonstrates how locally identified CSFs and machine learning can support pre-implementation assessment of construction project success in the Tanzanian context.