Text Analysis for Job Matching Quality Improvement

Abstract

Recently employment options based on fulltime and part-time models have increased. Companies may outright hire fulltime and part-time staff or hire temps through a staffing agency. In this case staffing agencies will facilitate the candidate to company match. Each individual has their own personalities and as a result, each individual has favorable and unfavorable job situations. When the staffing agency, taking into consideration characteristics, introduces the optimal job match, both the company and the individual reap benefits. In the case a suboptimal match takes place, the individual and company may end up in a damaging, unfavorable situation. Therefore, it is important not only for the staffing agency but also for our society to improve the quality of candidate to company match. In this paper, we analyzed data actually used in matching to understand the key factors for better matching. The data includes commute time, job location, job type, hourly rates, skill set of candidate, and so on. Of these attributes, using a text-mining tool, we specifically analyzed text data which had been written by recruiters at a staffing agency. As a result, we could extract positive and negative keywords affecting the matching result.