Representation of virtual choreographies in learning dashboards of interoperable LMS analytics

Abstract

Learning management systems (LMS) collect a large amount of data from user interaction, and it isn't easy to analyze this data in a reliable and context-independent manner. This research seeks to comprehend how virtual choreographies can be represented in interoperable LMS analytics dashboards. In order to gain a better understanding of the problem, this objective has been divided into three sub-goals: determining which interactions can be gathered from LMS contexts, identifying virtual choreographies from LMS logs, and representing virtual choreographies in learning dashboards. To achieve these objectives, we first conducted a Systematic Literature Review to comprehend the behaviors and interactions other authors have investigated in LMS contexts. Then, by applying these findings to this dissertation's case study, a methodical procedure for extracting valuable choreographies from the logs was outlined. The Design Science Research methodology was then applied to transforming logs into virtual choreographies and their representation in learning dashboards. It was implemented two services: one responsible for identifying virtual choreographies from data logs and transforming the logs into statements, recipes, and choreographies, following xAPI specification elements; and the other translates the information from the backend service into dashboard visualizations, allowing the user to view representations for statements, recipes, choreographies, and various visualizations. These artifacts provide a new flexible and cost-efficient solution for the identification of virtual choreographies, thereby facilitating the widespread adoption of their use.