Spark-based Indicator Execution in OpenLAP
Open Learning Analytics (OLA) is an emerging field, which deals with learning data collected from various environments and contexts, analyzed with a wide range of analytics methods to address the requirements of different stakeholders. OLA introduces a set of requirements and implications for LA practitioners, developers, and researchers. These include data aggregation and integration, interoperability, reusability, modularity, flexibility, extensibility, performance, scalability, usability, privacy, and personalization (Chatti et al., 2017). The Open Learning Analytics Platform (OpenLAP) is a framework which addresses these issues and lays the foundation for an ecosystem of OLA. It follows a user-centric approach to engage end users in flexible definition and dynamic generation of personalized indicators. The generated indicators are executed by querying data, applying filters, performing analysis, and generating visualization to be rendered on the client side. To meet the requirements of diverse users, OpenLAP provides a modular and extensible architecture that allows the easy integration of new analytics modules, analytics methods, and visualization techniques (Muslim et al., 2018, 2017, 2016).
Recently, the tools and solutions available for big data (e.g. Hadoop, Spark, NoSQL DBs) have rapidly increased providing an efficient way to store and analyze a large amount of data. In order to address the performance and scalability challenges in OLA, the goal of this thesis is to leverage Apache Spark in the indicator execution process of OpenLAP to perform analysis in a distributed manner.
- Design, implement, and evaluate a spark-based data analysis extension for OpenLAP.
- Good programming skills in Java and Web Technologies
- Prior knowledge in big data technologies (e.g. Hadoop, Spark, NoSQL DBs) is helpful
- Interest in data science and learning technologies
- Chatti, M. A., Muslim, A., & Schroeder, U. (2017). Toward an Open Learning Analytics Ecosystem. In Big Data and Learning Analytics in Higher Education (pp. 195-219). Springer International Publishing.
- Muslim, A., Chatti, M. A., Bashir, M. B., Barrios Varela, O. E., and Schroeder, U. (2018). A Modular and Extensible Framework for Open Learning Analytics. Journal of Learning Analytics, 5(1):92–100.
- Muslim, A., Chatti, M. A., Mughal, M., and Schroeder, U. (2017). The Goal - Question - Indicator approach for personalized learning analytics. In Proceedings of the 9th International Conference on Computer Supported Education - Volume 1: CSEDU (pp. 371-378). ScitePress.
- Muslim, A., Chatti, M. A., Mahapatra, T., and Schroeder, U. (2016). A rule-based indicator definition tool for personalized learning analytics. In Proceedings of the Sixth International Conference on Learning Analytics & Knowledge, pages 264–273, New York, NY, USA. ACM.