Cluster-Based Job Matching System

Abstract

A recruitment system comprises the processes, routines and elements essential to speedy and effective hiring for an organization or for a company. A good recruitment system includes all necessary features to aid management in hiring the best candidates for open positions. An "e-Recruitment" system is also available for recruiters to advertise jobs online where the applicants fill a form online and send or post their profiles. It is also a strong and effective recruitment system. However, there are many difficulties such as time-consuming and the lack of relevancy of job-matching. Online job recruitment platform is one of the most prominent channels for both job seekers and recruiters to hunt jobs and find suitable employees respectively. In the traditional job matching process, manually scanning the resume or profile of a job seeker and matching the resume of job seekers and requirements of job recruiters takes time-consuming and makes difficulties for both seekers and recruiters. Thus, nowadays, many job recommendation systems and cluster-based job matching systems appear. The studies applied k-means clustering for providing the similar clusters of data but gives less relevant data. This system has implemented a job matching system using k-means and word2vec that is to output the clusters with semantically similar words. As a result, using k-means clustering and word2vec model, recruiters can get the most relevant job seekers that fit employers‟ needs than k-means clustering only. And then, for the relevancy of job matching between job seekers and job recruiters, the classification accuracy method has been used in this system.