Social learning in large online audiences of health professionals: Improving dialogue with automated tools

Abstract

In reply to the reviewers' comments, the following changes were made: -References were added and the text was completed in relation to these references, particularly about communities of practice and the revolution of interaction triggered by the new information and communication technologies. -The results were summarized and the discussion was extended. -A clarification was made regarding social media in CME and the scope of the article. -A reference published in MedEdPublish about the same topic after the first version was added AbstractMassive open online courses (MOOCs) bring about the opportunity to reach large international audiences of health professionals. However, change in clinical practice eventually needs social interaction, to validate the new knowledge with trusted peers, in the agreement and adoption phases of change. How can meaningful dialogue take place without scaling up expert tutoring?The extensive experience from social network applications such as Facebook or Twitter provides an opportunity to improve dialogue among peers and with experts automatically and seamlessly, as part of what is called social learning analytics (SLA). Large amounts of data about prior relationships among participants in a course -similar to Facebook and other social applications-, among participants and course materials -similar to Netflix or Amazon -, as well as natural language processing, could be obtained, and then analyzed and used to improve the educational processes and outcomes.In this paper, a series of examples with pilot uses of SLA in the context of massive online courses for physicians and other health care professionals are described. They include: 1) Forecasting of academic accomplishment. 2) Teambased face-to-face learning as part of massive online courses. 3) Analysis of existing connections, to ensure the most connected discussion groups of course participants. 4) Facebook-like dialogue with other course participants who are previously related, as well as with the Course Faculty. 5) Crowdsourcing and friendsourcing, for recommending useful study materials or future courses. 6) Natural language processing, to classify posts in online discussions. It should be noted that the article does not address the use of Facebook or Twitter in continuing medical education (CME), but instead, the use of their approaches in CME.The intent of this manuscript is to create awareness in the medical education community that this type of analysis is possible and potentially useful, to receive feedback on the possible functionalities as well as critique these developments, and to create a space for collaboration in research and innovation projects with other interested parties.