A Data Mining Approach to Identify Associations Between Job Titles and Skills in Job Vacancies

Abstract

At present, job vacancies appear in online job vacancy repositories. They are available in the form of images or in the form of Portable Digital Format Documents. Text information is embedded in them. By mining that text information, current dynamics of the job market can be identified. As the Information Technology industry is dynamic, it is worth understanding what associations exist between job titles and technologies that frequently appear together, when applying for a job. For this purpose two algorithms can be used. They are Apriori algorithm and Frequent Pattern Growth algorithm. Among these two algorithms, this study emphasizes the importance of using Frequent Pattern Growth algorithm because it has eliminated the issue of performing so many scans in the database, which lead the Apriori algorithm less efficient. Frequent Pattern Growth algorithm used to mine association rules which exist between job titles and technologies required for them. The aim of the study is to mine how technologies appear associated with each other in job vacancies. Job seekers can be aware of what technologies have association trends in the job market and that would be helpful to reduce the gap that exists between skills of job seekers and industry demands.