Abstract:

Data mining is the process of discovering patterns in large data sets involving different methods at the intersection of machine learning, statistics, and database systems. This interdisciplinary subfield of computer science is currently being used in most of the industries and helps to discover different (sometimes unexpected) dependencies.

Over the recent years, the construction industry has been adapting the information technology (IT) in terms of computer design, construction documentation, maintenance, cost estimates, schedules (for example through BIM – Building Information Modelling), sales, procurement, etc. Gathered information about buildings, contracts, facilities, their use and maintenance generates a solid platform for the use of data mining.

Using data mining methods in construction can improve the management and maintenance processes. Stakeholders like: architects, building managers, contractors, engineers, and owners can obtain important information from the databases using one of the numerous tools mentioned in the article (for example CART, MARS, ANN, SVM).

Data mining can also allow for discovering of previously unknown dependencies in processes like construction training or decision-making.
The article is a review of the data mining and machine learning methods used in construction. It presents current applications and academic research projects.

Authors describe the phenomena, identify relevant publications, asses them, summarize and interpret significant findings. As a result, missing areas, trends and research opportunities are outlined.

Keywords:

Construction, data mining, machine learning.