A review on massive e-learning (MOOC) design, delivery and assessment 1 Department of Cultural Technology and Communication (D.C.T.C), School of Social Sciences, University of Aegean, Mytilini, Greece 2 Department of Computer Science, Multimedia and Telecommunications, Open University of Catalonia, Barcelona, Spain Abstract— MOOCs or Massive Online Open Courses based on Open Educational Resources (OER) might be one of the most versatile ways to offer access to quality education, especially for those residing in far or disadvantaged areas. This article analyzes the state of the art on MOOCs, exploring open research questions and setting interesting topics and goals for further research. Finally, it proposes a framework that includes the use of software agents with the aim to improve and personalize management, delivery, efficiency and evaluation of massive online courses on an individual level basis. Keywords: MOOC, OER, agents, e-learning, virtual learning environment (VLE) I. INTRODUCTION UNESCO considers that universal access to high quality education is the key to the building of peace, sustainable social and economic development, and intercultural dialogue. Open Educational Resources (OER) provide a strategic opportunity to improve the quality of education as well as facilitate policy dialogue, knowledge sharing and capacity building. [1]. A recent popular trend in e-learning are MOOCs or Massive Online Open Courses. According to Siemens [2] MOOCs are a continuation of the trend in innovation, experimentation and the use of technology initiated by distance and online learning, to provide learning opportunities for large numbers of learners. The Commonwealth of Learning [3] indicates that MOOCs are a means of facilitating the efficient creation, distribution and use of knowledge and information for learning by taking advantage of freely available online resources such as OER, and that they can be used to support social networking and other forms of “connectivity” among the participants. Therefore the power of the MOOC is based on the active engagement of large numbers of self-organizing learners and the connections they build amongst them using the course platform and other available social interaction tools. In the first part of this paper (Section II) we develop an analysis of the state of the art on MOOCs in the areas of design, delivery and assessment. In Section III, we set up the context of our research by identifying issues related to improved and personalized management, delivery and assessment of massive online courses. In particular, we define our research problem, which we break down into specific research topics. We then analyze the topics into open research questions and we establish our research goals. In Section IV, we contribute our own proposal for improving MOOC design, delivery and assessment using software agents. II. STATE OF THE ART ON MOOCS The first open e-learning course for a large number of participants was launched about five years ago [2], but the concept became widespread in 2012 when several large Universities started their own MOOCs. According to Siemens [2] open online courses, sometimes called “massive” (MOOCs) due to their high enrolment numbers, offer a middle ground for teaching and learning between the highly organized and structured classroom environment and the chaotic open web of fragmented information. Therefore MOOCs have characteristic – and challenges – of its own, due mostly to the large number of simultaneous participants, with some courses offered by Coursera and Udacity exceeding 100,000 registrants [2]. Based on research work conducted so far, in this section we present some of the distinctive features of the design, delivery and assessment of MOOCs, and discuss the main challenges in each of these areas. A. Design MOOC models are evolving quickly, according to Siemens [2] they can be currently classified as xMOOCs, cMOOCs and quasi-MOOCs. xMOOCs replicate online the traditional model of an expert tutor and learners as knowledge consumers, with saved video tutorials and graded assignments. cMOOCs are based on a connectivist pedagogical model that views knowledge as a networked state and learning as the process of generating those networks, in this case using online and social tools. Lastly, the category of quasi-MOOCs encompasses a myriad of web-based tutorials as OER that are technically not courses but are intended to support learning-specific tasks and consist of asynchronous learning resources that do not offer the social interaction of cMOOCs or the automated grading and tutorial-driven format of xMOOCs. Although many MOOC formats exist, according to Glance et al. [4] who made a sample of them in 2012, most courses exhibit common defining characteristics that include massive participation, online and open access, lectures formatted as short videos combined with formative quizzes, automated assessment and/or peer and self–assessment, and online forums and applications for peer support and discussion. MOOC design should favor a Learner-Centered Thanasis Daradoumis 1,2 Roxana Bassi Fatos Xhafa 2 Santi Caballé 2 adaradoumis@uoc.edu rox@roxanabassi.com.ar fxhafa@uoc.edu scaballe@uoc.edu © 2013 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. DOI 10.1109/3PGCIC.2013.37