Management Causes of Delays in Government Construction Innovations Using Data Mining Techniques

Abstract

This research had two main objectives: 1) Identify the factors that lead to delays in construction projects, a highly significant aspect, and 2) Assess the performance of models using data mining techniques to predict and analyze these delays. This work highlighted the value of stakeholder involvement. The research findings revealed significant factors contributing to construction project delays. On the contractor's side, the most influential factor was the lack of diligence and failure to inspect the construction site before commencing work, leading to project expenditure errors with a high confidence value of 0.99. On the employer's side, a critical factor was the delay in granting approval or permission for the contractor to access the construction site, with a confidence value of 0.99. Inadequate communication and coordination between the supervisor and the contractor were also identified as key factors causing delays, with a confidence value of 1.00. Additionally, it was found that political situations significantly impacted financial liquidity, which in turn caused project delays, with a confidence value of 1.00. For the second objective, the performance of five models in predicting delays was compared. After thorough testing, the Random Forest model demonstrated the best performance, with an accuracy of 89.74%. Based on these findings, we concluded that the Random Forest model was the most suitable for managing and analyzing the causes of delays in government construction projects.