Computer Science Events
[PAST EVENT] Nancy Carter, Computer Science - Oral Defense
Abstract:
New personal computing devices such as smartphones, personal fitness trackers, and digital home assistants are moving towards fuller integration into user lifestyles. The usability of these designs by non-technologically sophisticated or physically disabled users is open to study and further enhancement. Humans vary widely in their psychological and physiological characteristics. Those who are challenged by technology may find themselves unable to easily utilize the full capability of their computing devices, or unable to participate in important aspects of technological society. By studying the effectiveness, efficiency and current usability of personal computing technology, we seek to improve usability for the non-technologically sophisticated, and physically disabled portions of the user population. We present our efforts towards studying the oldest users, designing and then testing an improved personal authentication mechanism. We study the emerging population of personal fitness tracker users, a new population with its own characteristics and dynamics. Finally, we study speech patterns, and response dialogue dynamics of the most popular digital home assistant, seeking to reduce current user frustration, and increase response accuracy rates.
Our interview-style study of older computer users revealed that they are often challenged by traditional text passwords, frequently reusing simple passwords or avoiding computer use altogether. In response, we created and tested a graphical password user authentication system on a touchscreen-equipped laptop that is based on selection of memorable images from each user?s personal past history. Our approach achieved a password entropy superior to traditional PINs, improved efficiency, and is also extendable to smartphones. Older or manually impaired users enjoyed using this authentication method, facilitating continued engagement with technology.
Personal fitness trackers measure user physical activities, streaming data to device manufacturers for analysis and retention. Manufacturers have not opened their designs or data for independent testing and verification, leaving consumers reliant on popular media for information. By exploring publicly accessible data records, we characterized real-world user motivations, reliability concerns, fitness activity levels, and fitness-related socialization patterns. Our work illustrates the privacy implications faced by fitness tracker users when activity data, social fitness records and personal annotations are exposed to public view.
Digital Home Assistants are currently placed in more than 10% of U.S. households, providing information, and device controller services in response to user verbal requests. We propose to study user request speech patterns with the Amazon Echo and the Alexa natural voice language engine. We design and test enhancements to improve request interpretation, recognition, and response action accuracy. The general user population exhibits a wide range of word and grammar utterance patterns. Investigating the cognitive and grammatical effects of these patterns will lead to improved speech and request intent recognition, along with more usable verbal dialogues between users and the digital home assistant device.
Biography:
Nancy Carter is a Ph.D. candidate at William & Mary in the Department of Computer Science. She is advised by Dr. Qun Li. Her research interests are human-computer interaction, and smart personal computing devices. Nancy received a bachelor's degree in Computer Science from the University of Maryland College Park, and a master's degree in Electrical Engineering from the Naval Postgraduate School in Monterey California.