Resumen de Estrategias de recomendación basadas en conocimiento para la localización personalizada de recursos en repositorios educativos
La abundancia de recursos disponibles en repositorios educativos plantea un reto: la necesidad de proporcionar soporte a la localización de aquellos recursos que se adapten a las necesidades, objetivos, preferencias, etc. de los usuarios, en definitiva, a la localización de los recursos que les resulten más convenientes según el contexto. Además es conveniente que esta localización sea capaz de proponer listas de recursos que no contengan muchos elementos y que estos sean lo más variados posibles. Finalmente los usuarios echan en falta la existencia de mecanismos de interacción que permitan explorar el espacio de los recursos y que reduzcan el esfuerzo a realizar para localizar un recurso. Los sistemas de recomendación, que actúan sugiriendo productos a usuarios, nacen con el propósito de facilitar la toma de decisiones en dominios y situaciones en los que las posibilidades de elección son muchas y variadas. Aunque tradicionalmente los sistemas de recomendación se han aplicado al campo del comercio electrónico, su uso se ha extendido a otros campos entre los que se encuentra el dominio educativo. El trabajo presentado en esta memoria de tesis se engloba dentro de la línea de investigación que afronta el traslado de técnicas de recomendación al dominio educativo. En concreto, este trabajo aborda el diseño y el uso de estrategias de recomendación basadas en conocimiento como soporte al acceso personalizado a recursos educativos existentes en repositorios electrónicos. Las estrategias presentadas en este trabajo hacen uso de una representación del dominio rica en conocimiento, promueven la personalización haciendo uso de la información contextual de la actividad y del estudiante, introducen variedad en los recursos recomendados y exploran un modelo de interacción proactivo sobre el repositorio de recursos educativos que se complementa con un modelo de navegación por propuesta. ABSTRACT. The development of electronic repositories for the storage of educational resources has been intensified during the last years and in most educational disciplines. The availability of these educational resources eases and motivates student self-learning as a complementary activity to lectures. However, the high number of resources that exist in these repositories makes the access difficult to those adapted to the individual knowledge, goals and/or preferences of the students. It is necessary to provide support for personalized searching functionalities, which retrieve resources that fit the needs, goals and preferences of the students. Hence, one of the goals of our research is to design recommendation strategies that support locating educational resources adapted to the student knowledge. Furthermore, this recommendation must be intended to propose a set of resources that are appropriate to the student so that she cans take full advantage of a study session. It means that the proposals may not contain a lot of resources and it would be also desirable that the proposals be as varied as possible, in order to prevent the student get resources that are very similar among them. Finally this recommendation should explore mechanisms of interaction that allows to navigate through the space of resources and reduce the work load of the users. Research work on recommendation technologies helps to alleviate the aforementioned information overload by supporting users in pre-selecting information they may be interested in. The goal of our work on recommendation technologies in e-learning is to provide smart support for accessing to the Learning Objects that exist in repositories. This proposal has lead to the definition of three recommendation strategies that make use of existing knowledge of the domain, as well as additional information from both the student and the activity, with an ontology-based semantic representation. One of the strategies provides an student with a recommendation, list of educational resources that are adapted to the student's learning needs. The second one promotes diversity in the recommendation list. The third strategy explores a proactive model for user interaction based on a navigation-by-proposing model. The implementation of these three strategies has lead to the proposal of a developed framework for the rapid prototyping of knowledge-based recommenders to the learning field. These three strategies have been implemented and evaluated in a computational way and in a real learning field, where teachers and students have shown their satisfaction with the recommendation strategies designed. These three strategies presented will come to address the weaknesses identified in recommender systems in education.
