Abstract:

A student-centered learning (teaching) methodology has been introduced in the subject “User Centered Design. Design for multimedia” (4º course, Computer Science Degree). The different teaching-learning activities carried out in the course promote the collaborative and autonomous learning of the students.

From a methodological point of view, these activities blend different teaching techniques, such as flipped classroom, autonomous learning, just-in-time-teaching or problem-based learning, and they have two features particularly interesting. On one side, the learning resources used in the activities are previously created by the students, instead of by the teachers. Since the course begins, students work in groups and collaborate for producing these learning resources. On the other side, teachers have a secondary role in this teaching-learning model, being necessary to give students a continuous feedback about their level regarding the expected learning results. The assessment must be automatic in order to reduce the teacher’s workload, especially in large groups of students. The evaluation method proposed here is innovative and is based on the use of Topic Maps (TM) and semantic techniques.

A new class of TM, called “Labelled Topic-Maps” (LTM), has been formally defined. LTM’s are used to represent the two points of view of the teaching-learning process. The expected learning results related to a resource or to a specific activity are represented by a LTM (generated by the teacher or the resource’s creator). On the other hand, when a student uses a resource or completes an activity, he/she also expresses with a LTM what he/she learned. Then, a set of semantic similarity algorithms inspired in graph theory are used to calculate different indicators to “compare” both LTM’s (expected learning results- vs students’ learning achievements). These indicators help students to obtain an interpretation about their individual and autonomous learning process and the teacher about the course development.

A software system, called M-eRoDes, has been developed in order to support this new teaching-learning method. It makes easy the creation and store of the learning resources, the management of the different kind of activities, and the assessment based in the use of semantic and LTM’s. M-eRoDes is a web oriented system and is based on last generation Web technologies. The functionality of the system has been integrated in a Web application that allows teachers to program and control the course activities and that allows students to participate in these activities and to access to their evaluations. The similarity algorithms use the WordNet database. Moreover, M-eRoDes semantically labels the available resources attending to their contents, and classifies them according to the IEEE-LOM standard, enabling their management in the different activities.

In this paper, we present the results of the last two courses and a discussion about the pros and cons of this methodology, the semantic tracking system and the learning assessment. Results reflect that the students consider that have a leading role in this process, are more motivated and feel more involved in the subject. Furthermore, the very nature of the activities promotes the development of transversal skills that are crucial in their training as computer science professionals.

Keywords:

Autonomous learning, semantic-based evaluation, knowledge representation, learning technologies.