Toward Personalized Visualizations
Abstract: As we
enter the age of big and messy data, visualizations have emerged as an integral part of
data analysis and decision making. Visualizations serve as an extension to the analyst's
cognition, and the computer supports the human by managing large amounts of data,
performing computationally difficult tasks, and providing relevant feedback to the
human. As a result, analysts are now ... read moreable to solve historically challenging problems in
many areas including health, business and military. Unfortunately, no two humans are the
same, and a single human's behavior may even change from time to time due to fatigue,
emotional and situational changes. Despite the overwhelming evidence of the impact of
individual differences, most systems are built using a one-size-fits-all approach. To
better support the analysis, the computer must be able to adapt and adjust to its user,
and intelligently tailor situationally appropriate information to the user in a timely
manner. Such systems represent the next generation of user interfaces in which the user
is not forced to adapt to the system's interface and design, but is free to interact
with the system in ways that are natural and in tune with her individual needs. This
dissertation works toward creating such next generation visualization systems by
exploring: (1) how individual differences such as personality affect the ways people use
visualizations to solve problems, (2) how we can automatically detect these traits from
a user's interaction with visualization tools, and (3) how we can leverage users'
individual cognitive differences in a mixed-initiative
Thesis (Ph.D.)--Tufts University, 2016.
Submitted to the Dept. of Computer Science.
Advisor: Remco Chang.
Committee: Robert Jacob, Samuel Guyer, Holly Taylor, and Cristina Conati.
Keyword: Computer science.read less