Overview

I spend much of my thinking about how interfaces can and should change as we move away from working with a single computer and begin working with a variety of devices. While today's users already work with a variety of devices, there is little support for coordinating activities across those devices. In fact, most computers are completely unaware that a user might own other computers. We are exploring ways to make the devices (desktop PCs, laptops, PDAs, cellphones, etc.) owned by a single user aware of each other and to make it easier for users to coordinate their activities across multiple devices.

I also spend time thinking about how devices can adapt not only to a specific user, but also to that user's ever-evolving needs. I am particularly interested in exploring how to balance the responsibility for action between the user and the system across a variety of different domains.

Current Projects

Since joining IBM Research I've been working on extending my research on multi-device interaction and personal information environments.

Previous Projects @ Georgia Tech

		Opportunistic Annexing Small, portable devices such as PDAs and cellular phones provide a convenient way for users to access data anytime and anywhere. Inherent to the portability of these devices are small displays and the limited input capabilities of pen input and phone keypads. To overcome these difficulties we propose a system which would allow small device users to opportunistically annex input and output devices that they encounter in their environment. Users could annex a keyboard to quickly type an email on their PDA or a monitor to view their calendar while talking on their cellular phone.
		DIAMOND Despite the increasing prevalence of physically proximate computing devices, current interfaces remain largely limited to single computing devices because of the prevailing assumption that interfaces can only draw on input and output (I/O) resources attached to the same device. That assumption has led previous research to emphasize transferring interaction to the computer with the best available I/O resources, but that approach introduces security and privacy risks. We propose to instead allow users to divide interfaces across multiple devices so that they can allocate functionality and information appropriate across trusted and untrusted devices. DIAMOND is a prototype framework that we are developing to support Dividing Interfaces Across Multiple Opportunistically aNnexed Devices.
		SEREFE At work and at play, people need access to the right information, and they frequently need to share that information with others. While current tools such as electronic mail and USB flash drives provide powerful mechanisms for managing and sharing information, they too often require that users anticipate what information they might share (e.g. so they have it on their flash drive) and when they might share it (e.g. so they bring the flash drive with them). These tools thus provide excellent support for planned sharing but inadequate support for serendipitously sharing digital information with others. SEREFE is a new architecture for SEREndipitous File Exchange that extends an instant messaging architecture to allow users to use any of their devices, including a cell phone, to share information stored on any of their devices with other users or to copy it to another of their devices.
		Translucent Computing Spyware, phishing, and malware are both a serious threat and a significant risk in part because of the emphasis in commercial operating systems on making computer systems more opaque: they have worked hard to hide the details of operation in the name of usability. While making computers more opaque is effective when they work exactly as desired, opacity is problematic when the actions of those computers can differ from their users' intentions. We propose translucent computing as an approach to improving computer security and user privacy by reducing the impact and incidence of spyware, phishing, and malware. Translucent computing emphasizes selectively presenting information about a computer's actions to its user to allow him to verify that those actions match his intentions.
		SWIMM SWIMM (SWimming In My Music) is a next-generation digital music player that provides adaptive music streams. A stream, while analogous to a radio station, draws from a user's digital music library, preferences expressed by the user, and the user's empirical listening habits to provide an unbounded stream of music customized to the user's tastes, even as those tastes evolve over time.