Using a Learner-Topic Model for Mining Learner Interests in Open Learning Environments.

The present study uses a text data mining approach to automatically discover learner interests in open learning environments. We propose a method to construct learner interests automatically from the combination of learner generated content and their dynamic interactions with other learning resources. We develop a learner-topic model to discover not only the learner's knowledge interests (interest in generating content), but also the learner's collection interests (interest in collecting content generated by others). Then we combine the extracted knowledge interests and collection interests to yield a set of interest words for each learner. Experiments using a dataset from the Learning Cell Knowledge Community demonstrate that this method is able to discover learners' interests effectively. In addition, we find that knowledge interests and collection interests are related and consistent in their subject matter. We further show that learner interest words discovered by the learner-topic model method include learner self-defined interest tags, but reflect a broader range of interests.