Claim tenability assessment in Indian real estate projects using ANN and decision tree models

Purpose Claims have become an inseparable part of construction projects across the world. Construction claims often tend to result not only in time and cost overruns but in case of a dispute arising from the claim, it may result in erosion of the brand value and the working relationship between the parties. Thus, construction claim prediction is important but is complicated because of a large number of dependent factors and the complex inter-relations between them. With the aid of machine learning techniques, claim tenability assessment for real estate projects in India is attempted in this paper. Design/methodology/approach In this research, artificial neural network (ANN) and decision tree models are used for assessment of claims in the Indian real estate sector using project and claims data from 275 real estate projects. Findings The developed ANN model assesses the claim tenability in a project with a high degree of accuracy. Both ANN and decision tree models identify that “inconsistency between drawings and specification” as the most influencing factor in claim tenability assessment. Research limitations/implications Notwithstanding the claim tenability assessment, the model, in its current form, cannot be used to predict the “extent of claim” in the real estate projects. Originality/value Claim tenability assessment in real estate projects, especially in India, is scantily discussed in literature. This research, by adding to the body of knowledge, helps in both claim assessment and identification of factors that need to be controlled to reduce the claim tenability in real estate construction projects in India.