A socio-recruitment system using multi-criteria decision analysis for candidate ranking

Abstract

The large amount of time online users are spending on social networking sites, has encouraged organizations and industries to study the possibility of using these sites to enhance and improve their business practices. Nonetheless, recruitment through social networking sites has, until now, not been exploited to its full potential. Recruitment and social networking sites share one common element; they both revolve around people. These sites contain millions of users which could potentially become future employees of an organization. A substantial problem organizations face is the handling of applications they receive for a job vacancy - the evaluation of each and every application, as well as the screening of candidates takes a significant amount of time which could be otherwise used. This study evaluates and implements a solution that enables candidates to apply for a job by simply using their Linkedln log in credentials. Upon applying for the job their profile data is extracted and strored in a database. This was made possible by accessing Linkedln's API's. A ranking system was then developed using MATLAB to process the applications. This system automates the screening process in recruitment and selection by ranking the applicants which are then shortlisted for an interview. The applicants are ranked upon their performance in the 'Education', 'Work Experience' and 'Skills' criteria. The recruiter inputs the candidate scores in each criterion and assigns weights to the criteria. The system then returns a ranked list of candidates to the recruiter. The proposed system produced satisfactory results. The data extraction tool retrieved the desired data and was stored in a secure database. The ranking system also prooved to be successful. When implemented in a real-world scenario to rank candidates of a job vacancy, the system produced almost identical results when compared to making use of traditional methods. The system also reduced the time spent on screening the applications and shortlisting the candidates for an interview.