Each learning object represents a self-contained topic that may be consumed separately from others. The related learning object metadata thoroughly describes the content and is based upon open specifications, such as IMS Common Cartridge, Content Packaging, LOM or the outdated ADL SCORM. Quizzes, exercises and tests, in turn, can be developed based on the IMS QTI specification.

The contents are stored in a separate content repository in order to be reusable for different learning environments, re-sellable for third party stakeholders and provide content-protection mechanisms. Once created, learning objects can be grouped into didactic blocks for micro-learning or offered as a hierarchical combination to be presented in interactive courses.

The emergence of online courses has enabled new research opportunities in characterizing individual learning behaviors in different ways, by taking into account the specificities of learning management systems. X-means clustering, for example, is used to extract typical learning patterns, such as completing, auditing and disengaging from distinct university courses. The behaviors found in these courses may provide insights into typical patterns that can lead to better grades.

For learners, Fraunhofer FOKUS developed a Smart Learning Recommender (SLR), where students can keep track of their personal predicted knowledge level on different learning objects at any point in time and obtain personalized learning recommendations to overcome individual learning weaknesses. In addition to content metadata, such as exam relevance, lecture times and prerequisites, SLR takes different factors into account for each student and learning object, such as the user’s self-assessments, interactions with the content, excercise performances, as well as individual forgetting curves and the learning progress of classmates. At the same time, teachers can make use of this data to get an overview of the students’ overall progress and be aware of any potential knowledge gaps.

Moreover, the researchers designed a generator to provide personalized learning paths through knowledge networks. These paths are constructed by considering the time at which they are requested and alternate routes are suggested to provide the user with a selection of preferred learning items. Another promising approach in utilizing artificial intelligence methods is to offer virtual learning assistance in terms of chatbots, with which users can communicate in textual natural languages. This approach reduces barriers to accessibility because learners do not have to log into a platform. Instead, they can access their learning content, definitions and content recommendations directly, e.g., through a text messenger.