Patterns in Context Activity Data (Iurii Ignatko)

In order to provide learners with personalized data, gathered context information has to be processed and analyzed. Usually, this is done by a specialized service that captures, persists, processes and present the data to the learner or any other application for further processing. As already mentioned, for presenting more personalized data, one can incorporate recommender systems or learning analytics. In this thesis the emphasis will be on learning analytics and, in particular, on data mining methods which are the part of learning analytics tools.
Data mining methods provide a powerful toolset for pattern discovering. Unlike simple statistics and/or filtering, which actually produce interesting and valuable results for the learner, data mining focuses more on investigating the hidden knowledge. It is particularly useful when datasets are very big and consist of different types of information. And this is true for the case with context data.