A Novel Approach for Learning How to Automatically Match Job Offers and Candidate Profiles
- PDF / 640,130 Bytes
- 10 Pages / 595.224 x 790.955 pts Page_size
- 47 Downloads / 159 Views
A Novel Approach for Learning How to Automatically Match Job Offers and Candidate Profiles Jorge Martinez-Gil1
· Alejandra Lorena Paoletti1 · Mario Pichler1
© Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract Automatic matching of job offers and job candidates is a major problem for a number of organizations and job applicants that if it were successfully addressed could have a positive impact in many countries around the world. In this context, it is widely accepted that semi-automatic matching algorithms between job and candidate profiles would provide a vital technology for making the recruitment processes faster, more accurate and transparent. In this work, we present our research towards achieving a realistic matching approach for satisfactorily addressing this challenge. This novel approach relies on a matching learning solution aiming to learn from past solved cases in order to accurately predict the results in new situations. An empirical study shows us that our approach is able to beat solutions with no learning capabilities by a wide margin. Keywords Human resources management systems · Knowledge engineering · e-Recruitment
1 Introduction One of the most challenging problems in the Human Resources (HR) domain is to deal with scenarios with a high amount of job applications. This problem has a number of direct and indirect consequences, including but not limited to the waste of resources on processing all these applications. For this reason, researchers and practitioners have focused on finding ways to reduce the cost associated to situations of this kind (Mirizzi et al. 2009; Yi et al. 2007). Additionally, the field of Human Resources carries an old problem on not giving a fair treatment to the job candidates who have spent their time on preparing an application, and however, no feedback on the reasons for not being finally hired for a given position is provided (Suerdem and Oztaysi 2014). Organizations do not usually send this feedback to the candidates since this task has little profit for them (Martinez-Gil 2014). However, we think that if we were able to provide some automatic mechanisms to do so, both sides could benefit, i.e. recruiters can improve their branding reputation, and the unsuccessful applicants can easily know
Jorge Martinez-Gil
[email protected] 1
Software Competence Center Hagenberg GmbH, Softwarepark 21, 4232 Hagenberg, Austria
the reasons behind the decisions taken by the recruiters, and take decisions leading to succeed in the future. To overcome this situation, researchers and practitioners from this field often remark that accurate methods for the automatic matching of applicant profiles and job offers could partly alleviate the problem (Malherbe et al. 2015). Therefore, the design of new automatic approaches that can improve the recruitment processes is an important challenge (Malinowski et al. 2006). Additionally, such an automatic approach could be of great interest for employment agencies and many educational organizations around the
Data Loading...
