Abstract:

Learning Management Systems (LMS) are widely used in all levels of education. In the last decades, the LMS usage have been increasing, leading to massive data creation. This massive data generation created a considerable potential for these systems and it has aroused interests in many stakeholders involved in educational process. Different techniques have been applied in educational data analysis.

Recommender systems and analytics systems are widely used in educational business as well as in other different business fields. Social Learning Analytics (SLA) is one of the techniques developed to support the learning improvement process. SLA is concerned on how users build their knowledge in networks, interacting with other users and with the environment.

Some SLA tools have been developed during the last years, following the growth of this field of studies. The tools are concerned with data analysis and data visualization and they help teachers and education institutions to improve their students learning process.

This work proposes the development of a SLA tool. The tool development is based on a reference model instantiation, along with the application of graph theory and SLA techniques. The system architecture relies on the main elements in this reference model The Generic Framework model, proposed by Greller and Drachsler in 2012, that considers six critical dimensions for Learning Analytics: stakeholders, data, objectives, instruments, external constraints and internal limitations.

Each of the dimensions can be instantiated in several different ways depending on the scenario of application. It is possible to convert concrete user cases into a model instantiation. Since SLA is a LA subarea, it is possible to instantiate a SLA scenario in the reference model. The model establishes relationships between the critical dimensions, making it possible to identify and react to impact caused by changes in any of the dimensions.

Basically, in the instantiation proposed in this work, the stakeholders are the students, teachers and tutors, and they generate interaction data by using the LMS (publishing posts, questions and answers). Social Network Analysis algorithms will be the instruments to achieve the objective of identifying students disconnected from the course network.

The user’s interaction data will be collected from the Wiki’s logs inside the LMS Moodle and data processing will performed by the tool. The data is organized in a graph database. All the data will be available and shown through a visualization tool integrated to the LMS.

The application of Social Network Analysis (SNA) algorithms, as a SLA technique, allows the logical and visual identification of the users’ participation in the course social network through the identification of connections between students and other entities. The connections representation and visualization will be achieved through a graph structure. A graph structure is a suitable representation for relationships between entities, in a context that the entities are represented by nodes and the relationships are represented by the edges.

Keywords:

Learning Analytics, Social Learning Analytics, Graphs, Social Network Analysis.