Graph Mining to Characterize Competition for Employment

Abstract

In this paper, we discuss a novel application of graph analytics to characterize competition in the workforce. We propose a methodology that relies on finding communities in a graph representing prospective employees (with edges connecting people who interviewed for the same job) and communities in a graph representing available jobs (with edges connecting jobs that interviewed the same person). We then apply the proposed methodology to a real dataset corresponding to cooperative internships offered to undergraduate students at a North American post-secondary institution, illustrating the benefits of our approach.