Knowledge Discovery and Machine Learning in Construction Project Databases

A KDD framework was developed to convert construction project data into knowledge. This paper shows the nine steps in the KDD process: (i) understanding and defining the problem, (ii) collecting data, (iii) exploring data, (iv) cleaning data, (v) enhancing data, (vi) selecting data attributes, (vii) mining data, (viii) analyzing the result, and (ix) evaluating the result. In this methodological procedure, the complexity of the construction data was considered to optimize the opportunities to discover valuable knowledge. To test the feasibility of the proposed approach, the KDD process framework was validated and tested with a database, RMS (Resident Management System), provided by the U.S. Army Corps of Engineers. Obviously, knowledge cannot be obtained from a database if the data have been collected inconsistently. In this thesis, the validation was conducted by comparing the results from the KDD process with estimations from a publication reference (RSMeans 2001) and project-control software (Monte Carlo simulation) used frequently by construction experts in industry. The result of the validation showed that the developed KDD framework would provide the construction manager the ability to identify possible project problems, such as causes of delays in activity, and to predict duration for dealing with the delayed activity.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Civil Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2015

Author dc:creator

• Kim, HyunJoo

Contributors dc:contributor

• Lucio Soibelman

• Artificial Intelligence

Language dc:language

eng

Identifier

(MiAaPQ)AAI3070350

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/83195