CURRICULA CUSTOMIZATION WITH THE READERBENCH FRAMEWORK

Abstract

Providing customized curricula tailored to learner's needs became a stringent problem while relating to the increasing number of people attending Massive Open Online Courses (MOOCs) and eLearning platforms because the same content is provided to all students. This study presents a Moodle plugin created on top of an eLearning course that enables curricula customization based on the learning needs of a high number of participants. With the help of the Mass Customization approach, two categories of attendees were identified in a previous research and imposed multiple filtering criteria, out of which the first one refers to participants’ profession. The second criterion, topics of interest, allows learners to select keywords of interest from a predefined two-level word list, but also to enumerate their own terms using natural language. With the support of ReaderBench, an advanced Natural Language Processing framework, the most relevant lessons are retrieved in descending order of semantic relatedness. Third, an additional specific parameter allows participants to establish what kind of learning materials they require - i.e., theoretical and background oriented, practice and counseling documents, or guidelines. Our collection of documents is composed of lessons with a short description and their title, together with lists of pre- and post-requisite lessons. Our tool provides a comprehensive list of recommended lessons that best match the input criteria, corroborated with the list of related pre- and post-requisite lessons. Moreover, we provide information in terms of the duration of each lesson, as well as potential Continuous Medical Education points gained after finishing all selected lessons.