A context-based algorithm for annotating educational content with Linked Data ⋆ Estefan´ ıa Otero-Garc´ ıa 1 , Juan C. Vidal 1 , Manuel Lama 1 , Alberto Bugar´ ın 1 and Jos´ e E. Domenech 2 1 Depto. de Electr´ onica e Computaci´on, Universidade de Santiago de Compostela 15782 Santiago de Compostela, Spain {estefania.otero,juan.vidal,manuel.lama,alberto.bugarin.diz}@usc.es 2 Netex Knowledge Factory, 15172 Oleiros, A Coru˜ na jose.domenech@netex.es Abstract. In this paper we present an approach for annotating and enriching educational contents modeled as Learning Fruits (LFs). LFs are web books described in XML files and created to make more dynamic and flexible the learning process. A way to reduce the cost of creating a LF is to complete its content with information available in the Web. The solution described in this paper combines syntactic and semantic analysis techniques to enrich and annotate the LFs with relevant and reliable data retrieved from the DBpedia repository. 1 Introduction The education of youth has always been a major concern of society, and a great deal of effort has been spent with the aim of getting the right tools to facili- tate the learning process. However, the information society in which we live has brought about the need to adapt traditional methods of learning to new habits and requirements. One of the new trends are the Learning Fruits 1 . Learning Fruits (LFs) are learning pills and interactive activities on topics of the school curriculum. More concretely, LFs are web books that stand out for their innovative format which provide a friendly and interactive interface to fa- cilitate the access and the navigation through the course contents. An important feature of LFs is that they provide links to other content that complement and extend the information to the student and the teacher. This step is expensive when implementing a course: (i) it involves identifying which parts need to be supplemented with external information, that is, it is necessary to determine the important topics of the LF; and (ii) it involves selecting and analyzing the external links in order to determine if they contain accurate information. In this paper we propose a semantic and context-based approach to minimize the cost of enriching and annotating LFs with external contents. Our solution ⋆ This work was supported by the Ministerio de Educaci´on y Ciencia and the Xunta de Galicia under the projects TSI2007-65677C02-02 and 09SIN065E respectively. 1 http://www.netex.es/santillana/eng/index.html