Interactive visualization for mobile visual analytics
Abstract
Mobility and situational awareness enabled by mobile devices make them attractive platforms for in-field operations. Until recently, form factor limitations have confined these devices to being just data-collection and display agents for in-field response and investigation operations. However, recent advances in mobile technology have enabled the usage of these devices for in-situ visual data analytics, enabling faster communication and knowledge discovery. Nevertheless, in order to facilitate this, intuitive mobile visualization and corresponding interaction techniques need to be developed and evaluated for effectiveness. This thesis presents a task-based approach to developing interactive visualization techniques for data exploration and analysis on mobile devices. According to this approach, application scenarios and common tasks are identified and categorized into existing taxonomies, novel mobile visual analytic solutions are developed, and these solutions are evaluated for effectiveness. Using this approach, three prototype solutions were developed, namely, SafeTogether, NetworkVis and MobileVALET. SafeTogether is a personal in-field response tool designed as a novel single-handed interface usable without prior training. Usability studies indicated users were able to learn and use the interface single-handedly without being trained. NetworkVis is an infield investigation tool designed for monitoring and analyzing network data. Field evaluations at the Ross-Ade stadium with network analysts demonstrated the utility of this system in discovering new patterns in network traffic. MobileVALET is an infield geospatial analysis tool designed using a novel focus + context-based technique for linked spatio-temporal and statistical visualization. User evaluation of this system revealed that this interface was significantly better over existing mobile geospatial visualization techniques for spatial analysis tasks. Finally, the task-based approach we followed is developed into a framework for designing generic task-adapted mobile visual analytics systems.
Degree
Ph.D.
Advisors
Ebert, Purdue University.
Subject Area
Computer Engineering|Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.