User-centric Visualization of Data Provenance

The need to understand and track files (and inherently, data) in cloud computing systems is in high demand. Over the past years, the use of logs and data representation using graphs have become the main method for tracking and relating information to the cloud users. While it is still in use, tracking and relating information with ‘Data Provenance’ (i.e. series of chronicles and the derivation history of data on meta-data) is the new trend for cloud users. However, there is still much room for improving representation of data activities in cloud systems for end-users. In this thesis, we propose “UVisP (User-centric Visualization of Data Provenance with Gestalt)”, a novel user-centric visualization technique for data provenance. This technique aims to facilitate the missing link between data movements in cloud computing environments and the end-users’ uncertain queries over their files’ security and life cycle within cloud systems. The proof of concept for the UVisP technique integrates D3 (an open-source visualization API) with Gestalts’ theory of perception to provide a range of user-centric visualizations. UVisP allows users to transform and visualize provenance (logs) with implicit prior knowledge of ‘Gestalts’ theory of perception.’ We presented the initial development of the UVisP technique and our results show that the integration of Gestalt and the existence of ‘perceptual key(s)’ in provenance visualization allows end-users to enhance their visualizing capabilities, extract useful knowledge and understand the visualizations better. This technique also enables end-users to develop certain methods and preferences when sighting different visualizations. For example, having the prior knowledge of Gestalt’s theory of perception and integrated with the types of visualizations offers the user-centric experience when using different visualizations. We also present significant future work that will help profile new user-centric visualizations for cloud users