The derived data approach to support the construction and consumption of explorable visual narratives

Abstract

This thesis proposes a novel approach called the derived data approach that supports the construction and consumption of explorable visual narratives in Technology Enhanced Learning (TEL). Like many other domains, TEL environments generate valuable data which can be difficult to interpret. As TEL continues to grow in popularity, the challenge of addressing high student dropout rates remains, which is due to a low level of student involvement in the learning process. Learning Analytics aims to motivate learners and enhance their engagement by using the student-logged data generated by TEL environments and also enables educators to monitor students. It uses Information Visualisation to present interactive views using student-logged data that learners and educators can analyse. Information Visualisation research has demonstrated the value of visual narratives in communicating a message, by highlighting facts and making the message more memorable. In addition, visual data exploration can support users in gaining deeper understanding of data. However, in many domains and specifically in TEL, there are deficiencies in the presentation of explorable visual narratives which could enable the analysis of data related to it. In addition, there is limited support offered during the construction of explorable visual narratives, specifically in the automatic generation of sequences which present visualisations that can be applied in such narratives. The derived data approach supports the creation of explorable visual narratives by automatically generating and presenting sequences of visualisations for a narrative under construction. It also supports the exploration of these narratives by deriving and presenting data related to it, which can be further investigated. A framework consisting of a set of models and components based on this approach has been designed and implemented in a technical infrastructure called VisEN (Visual Explorations through Narratives). The implementation of VisEN is tightly coupled to the design requirements to realise this approach, hence its evaluations validate the usefulness and effectiveness of this approach. VisEN has been evaluated through a number of user trials, both in a TEL context and as a stand-alone system. In TEL, it was shown that the approach supported the majority of ‘improving students’ to engage with course content. As a stand-alone system, the evaluations showed that the approach supported both the construction and consumption of explorable visual narratives. VisEN was also used with non-TEL data (Irish economic data), which showed that the approach can be effectively used in other domains as long as the data adheres to the same format used in TEL.