Machine Learning para la mejora de la experiencia con MOOC: el caso de la Universitat Politècnica de València

Abstract

The aim of this paper is to design a proposal for automated mechanisms based on machine learning to improve the experience of participants in MOOC courses at the Universitat Politecnica de Valencia and reduce dropout rates. Following a desing based research DBR design, in which pedagogical decisions have always been prioritised over data analytics, three iterations have been carried out with different methodological patterns (systematic literature review, machine learning based on data from 260 courses and 700.000 students, and creation of automated mechanisms) that always end with the presentation of results and feedback from the university team. The main conclusions of this work indicate that, of the twenty-five pedagogical dropout indicators referred to by the literature reviews in iteration 1, only ten of them are validated with UPV courses (no automated or automatable data are available for the others), and of those finally only six of them are possible predictors of student dropout, with the data used. Finally, a set of automated mechanisms are proposed to be applied in the university's EdX platform to improve the user experience and reduce the dropout rate in the courses.