A road construction project involves the planning, design, and construction of roads, highways, and related infrastructure, such as bridges, tunnels, and drainage systems. Risks in road construction arise from various factors, including adverse weather conditions (e.g., heavy rain or flooding), soil instability, labor shortages, equipment failures, delays in material supply, safety hazards, and regulatory compliance issues. Managing these risks is crucial for ensuring the timely completion and safety of the project. This research addresses the critical challenges in risk management for road construction projects, where unforeseen risks often lead to delays, budget overruns, and safety hazards. The aim of the study is to develop a robust model for predicting and quantifying the severity of risks encountered in road construction projects, enabling better preparedness and decision-making. The research utilizes machine learning techniques, specifically the Multinomial Naive Bayes (MNB) classifier, to analyze and predict various risk categories such as Cost Overruns, Flooding, Design Errors, and Environmental Impact. The model was built using historical data on past construction projects, providing insights into the likelihood of risk events and their potential severity. The methodology involved data preprocessing, risk categorization, and the application of the MNB model to generate accurate predictions of risk occurrences. Evaluation metrics, including precision, recall, and F1-score, showed that the model performed effectively in predicting most of the risk categories with a high degree of accuracy of 93%. The results of the study indicate that the model can be a valuable tool for identifying and addressing critical risks like Community Opposition and Corrosion of Structures, allowing civil engineers and construction managers to allocate resources more effectively and mitigate risks before they escalate. We recommend the adoption of this risk prediction model by civil engineering teams, risk management units, and construction workers.
Edet et al. (Fri,) studied this question.