A Data Pipeline Approach for Building Learning Analytics Dashboards

In the era of data abundance, the ability to leverage data to assess the learning process is of great importance. Learning Analytics has been widely used and different approaches of deployment methods have been proposed, aiming to improve teaching and learning. Learning Analytics Dashboards (LAD), as the most dominant method to communicate results to the educational stakeholders, are found to be very effective. However, building a flexible and informative LAD is a complex procedure that incorporates several consecutive steps. The data pipeline framework which is used as a blueprint for generating LADs in this paper offers an important abstraction that helps the non-technical users to appreciate the effectiveness of the approach, as for every insight and report that is generated by a data scientist, there are most probably large such pipelines that implement the underlying functionality. This paper discusses the utility of data pipelines and presents the implementation of a LAD based on a data pipeline in Distance Learning students' data for summative assessment, along with some preliminary results.