Mining and Analyzing Occupational Characteristics from Job Postings

Abstract

Hiring/recruitment is key to an organization’s ability to position itself for success by attracting the right talent. Similarly, job search enables workers to connect to the right jobs in the right organizations. To assist in the hiring and job search processes, many technology solutions such as interest inventories, job recommendation models, job boards, and career pathway planning tools have been developed. However, solutions for preparing job postings are lacking. Job postings/ads play an essential role in hiring the right talent since they signal to the jobseeker the knowledge, skills, abilities, and other occupation-related characteristics (KSAOs) needed for a job. If the job ad does not convey the correct occupational characteristics, it is less likely that a well-qualified candidate will apply. Therefore, we present an interactive job ad visualization tool that analyzes the text in a job ad and matches phrases in it to a large occupational taxonomy of KSAOs. We combine O*NET, an occupational taxonomy, with natural language processing to perform semantic similarity matching between KSAOs for an occupation and ad text, and thereby assist jobseekers in their search process and recruiters in preparing job ads.