Group members



    Brief CV:

    In November 2000 I finished my studies in Computer Science and, from that date, all my working life has been linked to research and teaching.

    Upon completion of studies and for 1 year and 9 months I worked for the Servicio Regional de Investigación y Desarrollo Agroalimentario (SERIDA) of the  Principado de Asturias, a public body that runs the programs of agricultural and food research in Asturias. This led me to collaborate on various research projects in which AI techniques were applied to the resolution of agribusiness and genetic problems. At this stage it took shape what became my doctoral thesis (2003) and mark the course of my research: the problems of preference and ranking. During my stay in this research center I had the opportunity to have real data sets (something unusual in this field), which allowed me to work on finding solutions to current problems in the real world. My contract with the SERIDA came to an end, however, I maintain close cooperation with this organization in several researches that produce publications in leading journals.

    Later I joined as Assistant Professor (2002) at the University of Oviedo and I am still there, as Tenured Associate Professor since 2009, having gone through the figure of Contracted Doctor between 2006 and 2009. All these contractual changes were validated by the corresponding evaluations by the ANECA.

    Within the University of Oviedo, I participated in several research projects and contracts, all directed from the Artificial Intelligence Center of the University. Collaboration on these projects has resulted in publications in prestigious international Machine Learning conferences as NIPS, ICML or ECML and in prestigious journals such as Journal of Machine Learning Research, IEEE Transactions on Neural Networks and Learning Systems and Pattern Recognition.

    At present my research focuses on the search for automatic evaluation methods open answers in Massive Open Online Courses, which can have tens of thousands of students, which prevents the assessment of these works by the teachers. Our approach is based on trying to establish a ranking of exams, by means of matrix factorization techniques, from the preferences obtained through a peer evaluation, where each student orders a reduced set of exams.

    Another line of research is the quantification, where the important thing is not to know the class of each sample individually, but to know the percentage of occurrence of each class in future sets.