Germán Cobo Rodríguez

Development of Clustering Techniques and Algorithms

The main aim of this research line consists of the development of data mining tools to obtain knowledge based on large quantities of data, represented in multidimensional spaces and from diverse sources. One example is the design and implementation of parameter-free clustering algorithms. These algorithms do not require any prior knowledge about data, not even, and most importantly, the number of clusters to be identified. These techniques' and tools' scope of application is extensive and varied, as data can be obtained in a wide variety of research fields.

Student monitoring: assessment and feedback

The processes used by students are analysed at different levels: session (what do they do when they log on?), semester (how are they progressing in the course?) and programme (how do they acquire and develop skills at different times of life?). On all three levels, it must be possible to keep a record of what students do, why they do it (motivations, expectations, etc.) and how they do it (academic performance, satisfaction), and be able to intervene by means of adequate feedback (positioning, interaction, etc.).