A Novel Job Similarity Index for Career Transition in the Mining Industry

Abstract

In this study, with the primary goal of capturing ongoing digital transformation and automation impacts on the mining industry and its workforce, we conduct several interviews with mining industry experts in the USA and analyze our survey reports qualitatively and quantitatively through exploratory analysis. After the interpretation of the insights of industry experts, we proceed to generate a personalized and customized data analysis through a novel metric based on skills, knowledge, competencies, and occupational requirements, which quantifies the job similarities for occupations in the mining industry based on the publicly available database of the United States Department of Labor. We utilize text analytics to tokenize and classify the interviews to capture a better understanding of major response categories. The temporal analysis shows that the critical competency needs in the data science and autonomy category increases from 28% in current demands to 43%. In defining our metric, we also calculate Kullback–Leibler (KL) divergence for each job profile that enables determining whether and to what extent that job is transitionary in our test set based on the mean, standard deviation, and kurtosis of each job of interest. Our analysis reveals that the in-group job transitions are significantly easier than the between-group transitions, proving our initial assumptions and common sense. The generated heat maps provide the opportunity to present the gap between the current job and desired job profiles that provide feasible career change options, among others, offering individualized career paths for job seekers and promoting potential job transitions. Through the collection of industry-specific individual employee data, the AI system is envisaged to continue to learn as end users engage with the system, thus creating a central data hub specifically for the future workforce in the mining industry. Although the study has limitations on generalizability for qualitative assessments, it presents itself as a valuable application of how qualitative and quantitative approaches could be of value for future worker training in the mining sector.