A support tool to improve course credit transfer in an education institution

Processes of course transfer equivalencies should verify the compatibility or equivalence between these curricular components. In educational institutions, the teachers evaluate manually such decision processes with no type of technological support. In order to determine if the courses attended by the students in their institutions of origin can be accepted, the teachers make comparisons between the contents of both courses (attended and requested). Allied to this, the semiannual volume of these processes makes the analysis tedious, time-consuming, error-prone, and constantly challenged by stakeholders. Thus, this work purposes the development of a decision tool based on Natural Language Processing (NLP) techniques to aid in identifying the equivalence of disciplines through the analysis of their contents. The purpose of the decision tool is to support teachers during the evaluation of processes to take advantage of these curricular components. In order to evaluate the performance of the system, we constructed a dataset containing teacher evaluations in real processes of course equivalencies. This dataset was the gold standard (benchmark) for the computational tests. The metrics used in the tests for the evaluation of the proposed technique included AUROC curve, Accuracy and F-Measure.