Patterns of Skill Sets for Multiskilled Laborers Based on Construction Job Advertisements Using Web Scraping and Text Analytics

Abstract

The major objective of this research is to investigate what skill sets are desirable to utilize a multiskilled labor force in the construction industry using web scraping and text analytics. The construction industry has been suffering from skilled labor shortages for decades. To resolve the shortage and pertinent problems, multiskilled labor utilization has been presented in construction. Although extensive research has been conducted over multiskilling, it is still uncertain what skill sets are desirable to foster and utilize multiskilled labor force in construction projects. It is also difficult for laborers to know which skill sets will be competitive in the job market. To fill this gap, this research investigated skill sets for multiskilled labor development in the construction industry using web scraping techniques. The results were derived from large data collected over a year. We analyzed texts of construction job advertisements and provided the trends in what skill sets are beneficial for laborers to get employed using text mining techniques with categories from US Bureau of Labor Statistics. The results provided construction workers with the information of what kind of skill sets are required to meet industrial needs. This study provided two types of dual skill sets matrices from text analytics, analyzing only the texts that are essential to job advertisements and analyzing all the texts in job advertisements. The research also provided an example for multiskilling strategies with the matrices and an additional matrix from professional surveys. This study will contribute to the body of knowledge by identifying high-demand dual skill sets and presenting the multiskilling through the data-driven analysis of real construction job market.