Un entorno para la creación de actividades de aprendizajeuso de estándares de interoperabilidad, analíticas de aprendizaje y automatización basada en modelos

Abstract

Producing online content is a difficult time-consuming task that requires the design and creation of course elements, normally using authoring tools. Once teachers have created the desired content, it must be published and linked to e-learning platforms. Moreover, other aspects must be considered in online content production: i) portability to different platforms to ease the maintenance by avoiding to have different versions. ii) the integration of content created with third-party learning applications, iii) the reuse of content in different courses, and iv) the ability of learning analytics to improve the learning/teaching processes. Interoperability between tools is a key factor to offer these functionalities. Higher education institutions must support teachers by providing an integrated environment to create content, which should include authoring tools and other tools intended to offer the above functionalities. Goals This PhD. thesis aims to develop an online content creation platform with the following functionalities: i) it should support the creation of interoperable online content based on standards, ii) the content created must be portable and independent from a particular platform or technology, iii) it should provide a content reuse mechanism for a complete course or individual units, iv) it should provide content sequencing mechanisms, v) content gamification, and vi) learning analytics customization and efficient visualization of dashboards. Methodology In order to achieve the goals introduced in the previous section, we have followed the Design Science Research Methodology. First, we identified the problem to be addressed: automating the production of online content for teaching in universities. We note that current authoring tools have some limitations that make it difficult for them to be widely used by teachers. Then, an iterative process was followed to build the tools, which consisted of four stages: design, development, evaluation, and dissemination. We first created a basic services infrastruture UPCTforma). Next, we developed an authoring tool (INDIeAuthor) that overcome the limitations identified in existing tools, and finally we tackle the construction of the learning analytics component (CustomLA). UPCTforma was evaluated using a simulator and a case study, INDIeAuthor was evaluated with metrics and a survey given to teachers, and CustomLA by means of a case study of a university course. Results All the proposed goals have been achieved. (I) Development of an infrastructure that supports the basic functionalities required to create content, in particular, content and tool interoperability through the LTI standard. (II) Model-driven development of a family of textual domain-specific languages (DSL) to produce interoperable content. These languages make it possible to specify content, assessment, gamification, and sequencing. (III) Content reusability and portability. (IV) An efficient component to support learning analytics in different scenarios. (V) A textual DSL that enables teachers to customize learning analytics and dashboards. Conclusions In this PhD. thesis, novel features in online content production have been successfully explored. We have shown an interoperability-based approach to produce reusable and portable content. We have also presented a novel approach to customize learning analytics. Finally, this work has evidenced the benefits of using model-driven development techniques to build e-learning tools as well as the advantages of representing content as models.