Analysis of Design-Build Contracts Utilising Machine Learning on the Axis of Time, Cost and Quality
Abstract
In construction projects, contract management processes necessitate achieving a delicate equilibrium among time, cost, and quality. This work aims to analyse Design-Build standard construction contract texts by the application of machine learning (ML) methodologies to attain a more effective equilibrium among these three essential factors, hence enhancing the decision support process in contract management. The suggested model offers comprehensive predictions regarding time, cost, and quality in the management and risk assessment of construction projects where conventional methods are inadequate. A classification model employing text mining and ML algorithms is proposed in the study. The standard contract clauses of the FIDIC Conditions of Contract for Design, Build and Operate, as well as the JCT Design and Build Contract, have been analysed. In the realm of text mining, natural language processing (NLP) methodologies, including Term Frequency-Inverse Document Frequency (TF-IDF), Word2Vec (Continous Bag-of-Words (CBOW) and Skip-Gram), and Bag-of-Words (BoW), have been employed. Various ML algorithms, including Support Vector Machines (SVM), Decision Trees (DT), and Ensemble Learning (EL) techniques (XGBoost), have been employed to evaluate the efficacy of various text representation techniques. The models' performance was evaluated using 70%-30% and 80%-20% training-test data splits. The study concluded that the integration of the Skip-Gram approach with the XGBoost model yielded the highest accuracy (Acc) and F1 scores. In the 80%-20% train-test split, the F1 score was recorded at 0.8858 and the Acc at 0.8779, highlighting the significance of capturing contextual information. The primary constraint of the study is the inability to make a precise separation in the dataset, since the time and cost aspects in contract texts frequently overlap. This circumstance has led the model to erroneously categorise certain words into both the cost and time classifications, hence diminishing accuracy rates. The variety of legal and technical terminology in contracts has hindered the model's ability to appropriately evaluate some expressions. The results demonstrate that ML provides a novel approach to analysing construction contracts and has the capacity to enhance decision-making in contract management and negotiations. The model presented in this paper provides a framework for the more efficient management of time, cost, and quality factors.
Keywords
Standard Construction Contracts, Natural Language Processing, Machine Learning, Ensemble Learning, Text Mining
Ethical Statement
This study does not involve human or animal participants. All procedures followed scientific and ethical principles, and all referenced studies are appropriately cited
References
- Ayhan, M., Dikmen, İ., & Birgonul, M. T. (2021). Predicting the occurrence of construction disputes using machine learning techniques. Journal of Construction Engineering and Management, 147(4). https://doi.org/10.1061/(ASCE)CO.1943-7862.0002019
- Bafna, P., Pramod, D., & Vaidya, A. (2016). Document clustering: TF-IDF approach. In 2016 International Conference on Electrical, Electronics, and Optimization Techniques (ICEEOT) (pp. 61–66). IEEE. https://doi.org/10.1109/ICEEOT.2016.7754750
- Başarslan, M. S., & Kayaalp, F. (2023). MBi-GRUMCONV: A novel Multi Bi-GRU and Multi CNN-based deep learning model for social media sentiment analysis. Journal of Cloud Computing, 12(1), 5. https://doi.org/10.1186/s13677-022-00440-z
- Başarslan, M. S., & Kayaalp, F. (2024). Sentiment analysis of coronavirus data with ensemble and machine learning methods. Turkish Journal of Engineering, 8(2), 175–185. https://doi.org/10.55560/tureng.1394101
- Ben Talha, M., Fatima-Zahrae, H., Benghabrit, A., & Eddine, R. (2024). A machine learning approach based on contract parameters for cost forecasting in construction. Journal of Construction Engineering and Management. Advance online publication. https://doi.org/10.1061/(ASCE)CO.1943-7862.0002664
- Bogus, S., Migliaccio, G., & Jin, R. (2013). Study of the relationship between procurement duration and project performance in design–build projects: Comparison between water/wastewater and transportation sectors. Journal of Management in Engineering, 29(4), 382–391. https://doi.org/10.1061/(ASCE)ME.1943-5479.0000207
- Cantarelli, C. C., van Wee, B., Molin, E. J. E., & Flyvbjerg, B. (2012). Different cost performance: Different determinants? Transport Policy, 22, 88–95. https://doi.org/10.1016/j.tranpol.2012.04.002
- Candaş, A. B., & Tokdemir, O. B. (2022). Automated identification of vagueness in the FIDIC Silver Book conditions of contract. Journal of Construction Engineering and Management, 148(4), Article 04022007. https://doi.org/10.1061/(ASCE)CO.1943-7862.0002283
- Çaki, M. B., & Başarslan, M. S. (2024). Classification of fake news using machine learning and deep learning. Journal of Artificial Intelligence and Data Science, 4(1), 22–32. https://dergipark.org.tr/pub/jaida
- Debero, D. W., & Sinesilassie, E. G. (2024). Machine learning model for construction time prediction: A case of selected public building projects in Hosanna, Ethiopia. Journal of Engineering, 2024(1), Article 5653690. https://doi.org/10.1155/2024/5653690