Abstract:

The main objective of this scientific article is to illustrate the behavioral and interaction differences of students from face to face courses and online courses through learning analytics techniques combined with Big Data technology and analysis and pattern recognition algorithms in texts.

The experiments were carried out with classes from a massive and open online course on student career management and coaching, where two large groups were created, by means of teaching, in person and online, the texts and interactions of the students were analyzed in the forums.

Discussion of both classes, through the integration of the virtual learning environment and a Big Data platform with implementations of text analysis and clustering algorithms.

It was observed through the analysis different behaviors and interactions between students of online and face to face courses, in which the generated clusters demonstrated that the discussions promoted by the students of the online modality were more in-depth and relevant to the subjects of the course, while the discussions of the students in the face to face courses were more superficial and of diverse subjects.

These analyzes allowed tutoring to adopt different student engagement approaches for each of the teaching modalities and for the clusters generated, so as to increase the perception of value in the course and in their future careers.

Keywords:

Big Data, MOOC, learning analytics, clustering, text analysis.