A preliminary study on text mining-based human resource allocation in a construction project

Abstract

Even engineers who have the same job title and are working on the same project have different skills and backgrounds. However, despite the importance of assigning the right person to the right job to ensure successful project delivery, current human resource (HR) allocation practices are only concerned with the list of projects in which a candidate has been involved, years of experience in a discipline, and their previous job titles. As a result, some engineers lack the knowledge and experience to perform the tasks they are assigned on a project. This study demonstrates the need for a long, descriptive résumé rather than the commonly used short, brief résumé. When the job candidates who participated in this study were asked to describe their work experience in several sentences on their curriculum vitae (CV) instead of describing them using a few words composed of just the job title and roles, the contents of the CVs were clearer and showed less bias. This paper therefore presents a text-mining algorithm and a semantic résumé analysis system that we developed to automatically extract and analyze the work experience candidates write about on their résumés. This algorithm and analysis system can help HR managers control the significant amount of information found on prospective employees' long résumés. To validate the algorithm and the system, six sample résumés were collected anonymously. These résumés were then reviewed both manually and by the developed system. In a subsequent study, we expect to apply the text mining-based HR resource allocation algorithm to select construction engineers for a real project and measure the job-matching rate.