|
|
| Home | Products & services | Support & downloads | My account |
 |
|
IBM Research HCI Symposium Program Titles, Abstracts, Bios
September 28th
Judy Olson,
Cost and benefits of being remote from your co-workers.
Abstract:
We know from personal experience and the literature that it is hard to work with distributed co-workers. Various organizational researchers have theories about what kinds of things can alleviate the difficulties. To examine this phenomenon in more detail, we have created an “organizational simulation,” a supply-chain competitive game with 10 players. We began our exploration by putting 5 of the players in one room (collocated) and 5 in separate rooms (remote players), just measuring the pattern of play over the course of two hours (five simulated days of activity). This gives a baseline of people’s performance, strategies and preferences about with whom they will work. Sure enough, remote people are ignored, but compensate by developing strong trading relations among themselves. One remedy for difficulties remote people have is for the collocators to have had experience being in their situation, empathizing, and paying more attention to remote people. In the second study with the same physical set-up, after three rounds of play, two collocators switched places with two remote players. We test people’s allegiances, and again who does well.
Bio:
Judith Olson is the Richard W. Pew Professor of Human Computer Interaction at the
Darren Gergle, CMU
The Value of Shared Visual Information for Collaborative Tasks
Abstract:
While recent advances in computing technologies have increased the availability of systems to support collaborative tasks, a sizeable body of work suggests that such tasks are often more difficult and less successful when carried out remotely than when collocated. These failures are due in part to a lack of understanding the ways in which task goals, task objects and features of the media interact to influence communication and performance. This talk presents work investigating one of these media features—visual information—and its effect on collaboration. In a series of experiments done in our lab we demonstrate that visual information is a critical media feature to consider when exploring collaboration. Our work takes a decompositional approach to this problem, in which we strive to specify the different types of visual cues available to collaborators in face-to-face and mediated settings and identify how these cues influence situation awareness, communication, and task performance. I will first describe the basic components of our theory. Then, I will present a series of studies examining people’s use of visual information in different contexts. I will conclude by highlighting the implications of our findings for the design of successful technologies for remote collaboration.
Bio:
Darren Gergle is a fourth year doctoral student in the Human Computer Interaction Institute at
John Karat User-Centered Design Research on Organizational Privacy Technology
Abstract:
The rapid advancement of the use of information technology in industry, government, and academia makes it much easier to collect, transfer, and store personal information (PI) around the world. This raises challenging questions and problems regarding the use and protection of PI. How do organizations currently address privacy and the management of PI stored on their systems? Many organizations store PI in heterogeneous server system environments and currently do not have a unified way of defining or implementing privacy policies. This makes managing data privacy difficult for organizations working to put in place proper management and control of PI, the data users who access and work with the PI, and the data subjects who have rights regarding use of their PI. In this presentation, I will discuss two years of privacy research our team (Clare-Marie Karat, Carolyn Brodie and myself) has completed to move from developing an understanding of organizational requirements for privacy management to designing and iterating on a prototype of that privacy capability. This research included (1) identifying privacy needs within organizations through email survey questionnaires, (2) refining the needs through in-depth interviews with privacy-responsible individuals in organizations, (3) designing and validating a prototype of a technology approach to meeting organizational privacy needs through on site scenario-based walkthroughs with customers, and (4) collecting empirical data in a controlled usability laboratory test to understand the implications of different approaches for privacy policy authoring.
Bio:
John Karat is a Research Staff Member in the Security, Networking and Privacy group at IBM TJ Watson Research. Over his career with IBM Development and Research he has worked on the development of guidelines and principles for user interface design, researched and advised on design collaboration, researched and developed speech-based systems, researched and designed electronic medical record systems, and is currently carrying out research in privacy enabling technologies. He served as chair of national and international standards development committees, on the editorial board for 3 Journals and as editor of one (BIT), as co-editor-in-chief of the Kluwer HCI Book series, and held executive positions in several professional societies.
Bongwon Suh,
Semi-Automatic Personal Photo Annotation Using Event and Clothing Identification
Abstract:
Annotation is important for personal photo collections because acquired metadata plays a crucial role in image management and retrieval. Bulk annotation, where multiple images are annotated at once, is a desired feature for photo management tools because it reduces users' burden when making annotations. I will describe an approach for automatically creating meaningful image clusters for efficient bulk annotation. These techniques are not perfect and so are integrated into a bulk annotation interface where users can manually correct errors. We present hierarchical event clustering and clothing based human identification techniques. Hierarchical event clustering provides multiple levels of "event" groups. For identifying people in images, we introduce a new technique which uses clothing information rather than human facial features.
Bio:
Bongwon Suh is a Ph.D. candidate in Computer Science at the
Yan Qu,
Sources of Structure in Sensemaking
Abstract:
A critical aspect of sensemaking is finding appropriate representations for information important to a task. Understanding where and how people get ideas about structure as they make sense of things will draw useful implications for designs of sensemaking supporting systems. In this study, a topic comprehension task was set in which subjects are required to gather information and build external representations. Results show that representation construction and information seeking are closely coupled, as people get aspects of structure top down deducing from their previous knowledge, bottom up inducing from facts they find, and by borrowing from previous sensemaking efforts of others. The findings suggest revisions of previous sensemaking theories and new opportunities for system design.
Bio:
Yan Qu is a Ph.D student majors in HCI at the
Heidi Feng, U.
A power/reliability model for spatial navigation
Abstract:
The significant level of system errors distinguishes recognition-based technologies from other applications. The traditional speed and accuracy tradeoff model for navigation tasks, which derives from human motor function theories, is incompetent in addressing system-induced errors. I propose a power and reliability model that addresses not only user errors, but also system errors and context-induced errors. The model also includes a systematic approach to quantify the reliability of a technique. A preliminary empirical evaluation has been conducted and the results were quite promising.
Bio:
I'm a Ph.D. student from UMBC. I worked with Dr. Andrew Sears on the topic of Universal Accessibility and speech interface. During the past two summers, I've worked with the privacy group in
James Fogarty, CMU
Constructing and Evaluating Sensor‑Based Statistical Models of Human Interruptibility
Abstract:
A person seeking a colleague’s attention is normally able to quickly assess the colleague’s interruptibility, but the computer and communication systems we currently use often interrupt at inappropriate times or unduly demand attention because they have no way to consider human interruptibility. I will present my work on developing reliable sensor‑based statistical models of human interruptibility that perform as well as or significantly better than human observers. My work starts by using a Wizard of Oz approach to examine the potential usefulness of a variety of sensors, without first requiring their development. Given the high costs associated with developing hardware and software sensing systems, this approach allows informed decisions about what sensors to implement. I then present the results of deploying real sensors into the normal working environments of office workers. This deployment shows that the simple sensors already in laptop computers can support statistical models of human interruptibility that perform as well as or even significantly better than human observers.
Bio:
Claudio Pinhanez
Exploring the Challenges to Create a Personal Medium for Imagery Expression
Abstract:
Imagine that the human body could produce imagery as effortlessly as it produces language, so that to generate a map on a street corner is as naturally done as it is to give verbal directions. In the first part of this talk, we explore and discuss devices and systems that could begin to give human beings this new vehicle for expression, such as personal projectors and ecologies of symbiotic displays. The second part looks into the interface paradigms that might allow a person to collect, produce, alter, and control the appearance and flow of imagery, using speech processing, personal displays, and direct brain interfaces. Finally we discuss the desirability of a personal imagery medium and explore possible short and long-term applications.
Bio:
Claudio Pinhanez is a computer scientist and media artist. Since 1999 he has been a research scientist at IBM T.J. Watson, where he designs and develops interactive spaces and investigates physical interfaces to information. For the last four years he has been leading the Everywhere Displays project, developing novel ways for interaction in ubiquitous computing environments based on steerable interactive projectors. Claudio got his PhD. from the MIT Media Laboratory in 1999 working on the design and construction of physically interactive environments and in computer theater. Claudio has also been a visiting researcher at ATR-MIC laboratory (
More information and publications at http://www.research.ibm.com/people/p/pinhanez/
Andrew Sears, U.
Usability, disability, and the functional continuum
Abstract:
Understanding the users, their tasks, and the environments in which they will be working is central to the design of effective information technologies. Many terms are used when discussing the goals that drive design including usability, accessibility, universal access, universal design, and universal usability. This talk explores these terms and the associated foci, expectations, and philosophies as well as gaps and assumptions that may limit the efficacy of each approach to design. It is argued that understanding the continuum of human abilities, and how these abilities move through the continuum as a function of time, environment, technology, and activities, provides a foundation for designing effective information technologies. However, this is not sufficient as we must also consider the metrics we use to assess effectiveness. As with terms themselves, it is argued that traditional metrics often mask important differences and hindering our ability to continuously improve IT solutions.
Bio:
Andrew Sears is a Professor and Chair of the Information Systems Department at UMBC. Dr. Sears' research explores issues related to human-computer interaction with recent projects investigating issues associated with mobile computing, speech recognition, information technology (IT) accessibility, and the difficulties IT users experience as a result of their work environment or tasks. His research projects have been supported by a variety of corporations and government agencies including Aether Systems, IBM, Intel
Corporation, Microsoft Corporation, Motorola, Platinum Technologies, NASA, NSA, and NSF. He earned his BS in Computer Science from Rensselaer Polytechnic Institute and his Ph.D. in Computer Science with an emphasis on Human-Computer Interaction from the
September 29th
Scott Hudson, CMU Managing Human Attention Through Situationally Appropriate Interaction
Abstract:
As noted by Herbert Simon many years ago: in an information rich world, human attention is the scarce resource. Yet most of what we know about how to build interactive systems today is still implicitly centered around the notion that the interface is our center of attention. We know little about how to build interfaces which support secondary tasks, which share our attention judiciously, or which can move from the center to the periphery of our attention as needed. In this talk, I will consider a number of efforts with my colleagues in the managing human attention project. This interdisciplinary project combines behavioral, technology, and design approaches in an effort to better understand issues of human attention, create basic technologies to allow it to be consumed appropriately, and design new interaction approaches which can take advantage of these technologies. Much of the work has concentrated on the creation of situationally appropriate interactions, which sense human activities, model the situations these represent, and act appropriately to those situation – particularly demanding appropriate amounts of attention.
Bio:
Scott Hudson is an Associate Professor of Human Computer Interaction in the HCI Institute at Carnegie Mellon, where he serves as the director of the HCII PhD program. He has been the General and Program Chair for the ACM UIST conference, has served as an Associate Editor for the ACM Transactions on CHI, and has published extensively on
technology-oriented HCI topics. His research interests cover a wide range of topics in user interface software and technology.
Helen Zhao,
Interactive Sonification of Geo-referenced Data for Auditory Information Seeking
Abstract:
For people with vision-impairment, audio is an important alternative or supplementary information channel. The current support for vision-impaired users to access geo-referenced abstract data (e.g., the population distribution throughout the
conducted with blind-folded sighted subjects, more studies with vision-impaired users are currently being planned to compare to our findings with blind-folded sighted subjects.
Bio:
Haixia (Helen) Zhao is a PhD student in the Department of Computer Science,
Wendy Kellogg
Research and Emerging Trends in Social Computing
Abstract:
Social computing has emerged as a broad area of research in HCI and CSCW, encompassing systems that gather, store, process, re-present, and disseminate social information that is distributed across social collectivities such as teams, communities, organizations, cohorts, populations, and markets. Social computing systems are likely to contain components that support and make visible social features such as identity, reputation, trust, accountability, presence, social roles, expertise, knowledge, and ownership. We see social computing transforming the net by creating a pervasive technical infrastructure that includes people, organizations, and their activities as fundamental system components, and enabling identity, behavior, social relationships, and experience to be used as resources for activity. In this talk, I selectively review emerging trends in social computing, and discuss current research in our group and beyond that is driving and is driven by the emerging vision of social computing.
Bio:
Wendy A. Kellogg is Manager of Social Computing at IBM's
| ||||||||||||||||||||||||||||||||||||||||||||||
|
| About IBM | Privacy | Terms of use | Contact |
