Visual Data Mining in Virtual Reality: An Exploratory Study

Larsen, Anton Christian Mondrup ; Hammer, Jens Lindberg ; Lindhardt, Thomas Krarup

4. term

Medialogy, Master

2024

2024-05-29

55 pages

This report investigates the largely uncharted integration of Virtual Reality into Visual Data Mining. In the era of Big Data where vast quantities of complex information are abundant, this confluence presents a compelling potential to enhance the efficacy of Visual Data Mining processes by combining the visual and interactive capabilities of the two technologies. This approach has the potential to revolutionize how we view, interact with, and understand complex data, change how the complexities of data analysis are navigated, and transcend the way actionable insights are unlocked. We conducted an exploratory study to examine the implications of combining the technologies and to identify how Virtual Reality can aid different Visual Data Mining processes with the clinical domain as our foundational context. We developed a Virtual Reality Visual Data Mining prototype to aid in our endeavour and to facilitate what such a tool should entail. The results from the study show that the integration is very much viable as Virtual Reality can be incorporated in multiple stages of Visual Data Mining. Research must however be conducted with Data Mining experts to uncover if Virtual Reality can actually facilitate the visual mining of data. Additionally, future work for a more complete Visual Data Mining system should entail actual Data Mining algorithms working behind the scenes to complement and transform the visual aspect so that data management is handled seamlessly with little to no manual back-end human intervention.

This report investigates the largely uncharted integration of Virtual Reality into Visual Data Mining. In the era of Big Data where vast quantities of complex information are abundant, this confluence presents a compelling potential to enhance the efficacy of Visual Data Mining processes by combining the visual and interactive capabilities of the two technologies. This approach has the potential to revolutionize how we view, interact with, and understand complex data, change how the complexities of data analysis are navigated, and transcend the way actionable insights are unlocked. We conducted an exploratory study to examine the implications of combining the technologies and to identify how Virtual Reality can aid different Visual Data Mining processes with the clinical domain as our foundational context. We developed a Virtual Reality Visual Data Mining prototype to aid in our endeavour and to facilitate what such a tool should entail. The results from the study show that the integration is very much viable as Virtual Reality can be incorporated in multiple stages of Visual Data Mining. Research must however be conducted with Data Mining experts to uncover if Virtual Reality can actually facilitate the visual mining of data. Additionally, future work for a more complete Visual Data Mining system should entail actual Data Mining algorithms working behind the scenes to complement and transform the visual aspect so that data management is handled seamlessly with little to no manual back-end human intervention.

Download
View record in AAU Student Projects