The Learning Analytics and Knowledge Dataset

In this abstract we describe an initiative aimed at providing a public corpus of structured data covering the research literature on Learning Analytics (LA) and Educational Data Mining (EDM). This work is jointly promoted by the Society for the Learning Analytics Research (SoLAR) with the collaboration of the Institute for Educational Technologies of the National Research Council of Italy, and the LinkedUp project (http://linkedup-project.eu/). The resulting Learning Analytics and Knowledge (LAK) Dataset (solaresearch.org/events/lak/lak-dataset) contains more than 300 papers collected from: - the Proceedings of the ACM International Conference on Learning Analytics and Knowledge (LAK) for the year 2011 and 2013, - the special issue on "Learning and Knowledge Analytics" of the journal of Educational Technology & Society published in 2012, - the proceedings of the International Conference on Educational Data Mining (from 2008 to 2013). The dataset represents a so far unprecedented body of knowledge for researchers in the field of LA and EDM due to a number of reasons: - the LAK Dataset is published according to Linked Data principles in a machine understandable format using Semantic Web technologies - the corpus contains structured metadata as well as the full text body of all publications - some of the publications have not been publicly available before and were published with a dedicated license from the ACM Digital Library - the LAK Dataset is published using established (RDF) schemas and is publicly accessible via standard interfaces such as SPARQL - the dataset is being constantly updated to provide a living corpus for research into the emerging field of LA and EDM