Application of KDD Techniques to Extract Useful Knowledge from Labor Resources Data in Industrial Construction Projects

Abstract

Improper management of labor resources is one of the main causes of schedule delays and budget overruns in industrial construction projects. During management of these projects, a vast amount of data is collected and discarded without being analyzed to extract useful knowledge. To address this issue, an integrated proposed methodology is developed based on a five-step knowledge discovery in data (KDD) model. First, a synthesis of previous research is presented. Second, an inclusive analysis of the industrial construction domain and labor resources data is performed. Third, the concept of predefined progressable work packages is introduced for consistent data collection. Fourth, a prototype data warehouse is built using the snowflake schema to centrally store the collected data and produce dynamic online analytical processing (OLAP) reports and graphs. Fifth, data mining techniques are applied to extract useful knowledge from large sets of real projects’ data. Results show that the developed methodology is capable of gathering valuable knowledge from previously unanalyzed data that significantly improves current resource management practices.