Framework Based on a Fuzzy Inference Model for the Selection of University Program Applicants

Abstract

This study consisted in the construction of an integrated framework based on a fuzzy inference model for evaluating the competencies in the entry profile of applicants to university programs in order to evaluate the process and make it more efficient. This framework was designed to systematize the method for evaluating the applicant profiles based on specific requirements and to improve performance indicators in the evaluation process. This architecture framework has two components: (a) the evaluation system and (b) the fuzzy model; where the system collects input information from knowledge assessment tests and applicant interviews, which are processed by a web service and delivered as parameters to libraries embedded in the fuzzy model, which was designed in Matlab SDK, and as a result the level of the applicant is returned. Furthermore, the system can be configured to add linguistic variables to the fuzzy model and set membership functions, as well as other modifications. This study is important because it facilitates the evaluators and improves the efficiency of the process to evaluate competencies in a profile as well as the efficiency of results which serve as input for process improvements. Finally, the results of the implementation of the framework based on indicators improved time efficiency (80.3%), use of resources (45.7%) and increased evaluator satisfaction (25.7%) with a value of p < 0.05 for all cases.