Investigation of Forming a Framework to shortlist contractors in the tendering phase

Abstract

The aim of this research is to create a framework that can predict the best contractor to be awarded a construction contract by a consultant/client using a different set of variables known as “Decision factors.” This research was conducted to improve the traditional tendering process, the model was used to predict the “Success Rate” for the project by assessing each contractor’s possibility of completing the project successfully using their compatibility with the project. The model creation was divided into multiple phases which started with finding the decision factors through an extensive literature review, and then determining the weights of each decision factor by conducting a survey that professional experts took. After obtaining the weights of the decision factors, a model using Machine Learning algorithm on Google Colab was written using the Python language. The model to shortlist contractors in the tendering phase was created using machine learning to enable more contractors to submit for a project without having to waste time and money on the tendering process; if they are compatible with the project, then they have a high chance of getting it by being short-listed for the project, which they can then submit their tender package for; this will also ensure that the best company gets the job for the client which will act as a great step towards improving the tendering in construction projects. For the consultant, it will decrease the load of going through numerous tender packages and ensuring that the best companies will tender for the project. This research has generated a base model that can be altered depending on the project requirements which can assist all parties involved within the tendering process to save time and money and improve the success rate of projects. The limitation of this research is that to use the framework to its full extent, it needs a huge database that includes data from numerous previous projects to be able to accurately predict the success rate of the upcoming project; however, if it could be regulated through governmental institutes then the database can be quickly collected within a relatively short period of time.