CS6604 Assignment 1 - Abstracts - partial set - more later

1. • Name: Ghaleb Abdulla
   • Email: abdulla@csgrad.cs.vt.edu
   • Title: Models and Tools for Users of Digital Libraries
   • Abstract:
     Keeping in mind projects ENVISION, EI, SUCCEED and BEV -- in the era of digital libraries and distant learning where the student and the teacher can interact using computers -- can we develop a model for this interaction? What are the tools needed by the teacher to prepare and organize his material in advance? What are the interactive tools needed by the teacher to update and change the information dynamically during the course? What tools are needed by students to interact with the teacher, solve homework, write reports and do their projects? Does every student need to have a computer that is equipped with a video camera, audio/video capture board, network connection to a digital library, and a set of tools for browsing editting, linking, composing, and synchronizing information? How does this differ from field to field? How is it similar in different fields? How do we model the new world of information? Can we have a digital information library that can be shared by everybody via computer network? Can we model storage capacity, network load, physical shape of the library, location of the libray, etc.? How do we represent, link, access, and edit the objects that compose the digital library? What is better in such an environment: WYSIWYG tools or descriptive tools?

2. • Name: Karen Baker
   • Email: kabaker@vt.edu
   • Title: Comparison of cursor positioning input devices and methods for improving target selection accuracy
   • Abstract:
     Cursor positioning is an integral requirement of today's graphical user interfaces. Several researchers have explored the relative performance and ergonomic advantages of cursor positioning input devices, including the mouse, trackball, lightpen, digitizing tablet, and touch screen. The touch screen has been shown to provide several advantages over other input devices because it provides direct spatial compatibility between display and control, utilizes simple pointing as a natural and intuitive positioning/selection method, and eliminates the need for a separate input device, thereby reducing space requirements and improving durability.

     Despite these advantages, the touch screen as gained limited application because of the relative difficulty users have in accurately selecting small targets. Various researchers and interface designers have proposed methods for increasing touch screen accuracy, making it possible, for example, for users to reliably select targets as small as one pixel.

     A review of the advantages and disadvantages of alternative cursor positioning input devices will be presented, along with an analysis of the touch screen selection accuracy problem. A proposal will then be presented for research comparing alternative means of improving selection accuracy and assessing the relative performance of these methods under various user and task conditions.

3. • Name: Carolyn A. Bussi
   • Email: bussi@vt.edu
   • Title: Investigation of Information Display for Route Guidance and Navigation
   • Abstract:
     Intelligent Vehicle Highway Systems (IVHS) refer to an integrated group of technologies which will save lives, save time, and save money. This research will be dealing with In-vehicle Routing and Navigational Systems (IRANS). IRANS provide drivers with information about how to get from one place to another, information on traffic conditions, and recurrent and non-recurrent urban traffic congestion. This study will investigate the factors relating to information presentation which allows the driver to access pertinent information. There are several ways to present information to drivers: via computer screen, Head-Up Display (HUD), and/or aurally. The visual information can be text-based or graphic.

     The experimental apparatus used in this study will consist of a computer program that simulates movement and a navigator program that will present navigational information to the driver. The movement simulator program will be programmed in VREAM (a virtual reality software package) to look like a city; it requires participants to start at one point and, with the help of navigational information, end at the desired destination. The navigator program will present the navigational information to the participant in a variety of ways: text-based on computer screen, graphic on computer screen, text-based on HUD, graphic on HUD, aurally, or a combination of these. This navigational information will be presented on either a separate computer screen, a reserved area at the bottom of the main computer screen (for the HUD), or through speakers.

4. • Name: Jose Castillo
   • Email: josec@csgrad.cs.vt.edu
   • Title: Remote Evaluation of a User Interface Using the Network
   • Abstract:
     Increasing use of networks puts physical distance between users and developers/evaluators. Actual evaluation techniques involve a close interaction between the observer and the user. The role of the observer is to identify if the interface satisfies the user needs while conducting usability tests in a place where users can be observed directly. The kind of observer that evaluates an interface in the same physical location of the user can be called "local observer". This concept is limited by geographical means, and not by technology. We propose to take advantage of the distributed nature of a network to perform remote evaluation of a user interface.

     Our hypothesis is that the user can have the ability to identify usability problems that may impact usage of any application and report them to a "remote observer", an observer in a remote location. We believe that the user can be trained quickly to identify problems on the design of the user interface. Any missing functionality, usability problem or specific comment should be reported immediately before it is forgotten.

     Several tools or mechanisms could be used to report any usability problem to a "remote observer". First, the developer can provide a notepad where the user can write any general comments about the interface. The user can press a SEND button to transfer the information to the "remote observer". The developers could also create a window for a group of screens to handle specific information. Another way to evaluate a user interface is using video cameras (video-teleconference). Finally, another method to report any design problem could be the use of Mosaic and the World Wide Web. We believe that providing such mechanisms would gather enough data for the developers to identify usability problems on a user interface and help them take the corresponding actions to improve it.

5. • Name: Jon Kamler
   • Email: jkamler@vt.edu
   semester.
   • Trace File Analysis as an Internal Validation Tool: Identifying Typical Responses as a Basis for Interface Design and External Validation
   • Abstract:
     The use of artificial intelligence (AI) applications in natural resources has grown steadily over the last decade. These applications include (among others) intelligent GIS systems, vision-aided image analysis systems, and particularly expert systems. In this study the performance of a specific expert system developed for a classic problem in natural resources will be investigated; that being landscape or aesthetic quality (a.k.a., scenic beauty). Expert, or knowledge-based, systems are capable of carrying out reasoning and analysis functions in narrowly-defined subject areas at proficiency levels approaching that of human experts. As a specific subfield of AI, expert systems "use human expertise to attain high levels of performance in a narrow domain. They typically represent knowledge symbolically, have a good user interface, and are able to explain their reasoning processes" (Engel et al., 1992, p. 5). Consequently in terms of visual quality assessment, expert systems have characteristics that can be used to develop not just a method of predicting quality, but also a system that explains why certain levels of quality exist. A prototype Expert Visual Assessment system (EVA) was developed in VPI PROLOG which incorporates human expertise, represents its rules symbolically, and offers a reasonably good explanation of its reasoning process. But EVA falls short in terms of the final component as it does not yet have (by today's software standards) a good interface. Initial testing of the system shows that it gives very reliable visual quality estimates and based on cursory testing appears to be internally valid. In other words the system appears to be functioning the way the designers intended it to function. But in reality it is still unclear as to exactly what is going on within the system in terms of the patterns of the actual responses; which paths, if any, are people consistently following through the decision tree and which paths are seldom or never used? Internal validation through trace analysis using a tool such as CHITRA is a proposed approach to answer this question. A further result intended from the trace analysis is identification of design parameters for interface development. The interface development is necessary to take the system to the next phase by testing its external validity. The system will be taken into the external-validity-testing phase by correlating the ratings it provides with real-world results from visual assessment methods currently in use. It would, however, make little sense to test the validity of (supposed) real-world responses if the system does not have a real-world interface. Therefore interface design based on the internal validation is part and parcel of external validation. Conjunctive to the external validation will be testing slight variations in the interface in order to determine: 1) any possible dependencies that the system output may have on the respective interfaces, and 2) if it is possible to greatly alter the user workload required to use the system with only slight changes to the interface.

6. • Name: Andrew Gellatly
   • Email: gellatly@vt.edu
   • Title: Investigation of Virtual Reality as an Effective Medium for Testing User Interfaces
   • Keywords: Virtual reality, immersion/presence, user interface design evaluation
   • Abstract:
     Virtual reality (VR) has the potential to become a test platform for interface designs. Typically, interface designs have been evaluated under ecologically-valid conditions. However, situations arise that lead to potentially hazardous design-evaluation conditions. Testing a new vehicle-information display interface under actual traffic conditions is one such example. Simulation provides a method for evaluating an interface in a hazard-free environment.

     Research has demonstrated the potential of virtual reality for training. The data suggest that skills learned in a virtual environment transfer to the real world. However, questions about the transferability of simulated interface evaluation results to the real world remain unanswered. There remains the possibility that particular interface designs may prove successful in artificial environments but fail in their intended use environment.

     The purpose of the proposed research is to evaluate various levels of reality to determine how results from a user-interface design evaluation are affected, if at all. Levels of reality would include the "real" world, fully immersive VR, motion platform simulation, fixed-base simulation, and desk-top VR. A user interface design would be evaluated under each level of reality. Results from the user interface design evaluation would be compared for each level of artificial reality to the results derived from the "real" world testing condition. Information gained may prove useful in decisions of whether to test user interfaces in real or artificial environments.

7. • Name: Austin Hicklin
   • Email: hicklin@csgrad.cs.vt.edu
   • Home Page
   • Title: Commercial Development on the World-Wide Web
   • Abstract:
     The development of the World-Wide Web and its massive growth are creating an opportunity for business that only existed in theory two years ago. Use of tools like Mosaic have made the commercial distribution of electronic goods (such as software, text, and images) practical. A great deal of interest has been expressed in the recent past regarding such ventures, and a "land rush" is starting of small companies that hope to be successful in this new market.

     Commercial development of the World-Wide Web is proceeding on a variety of paths, such as publishing, maintaining libraries of information or electronic goods, developing catalogs of material goods that must be shipped by traditional mail, and advertising. Currently I am investigating the issues involved in developing on the World-Wide Web, focusing on three possible business ventures: maintaining WWW home pages for customers, developing a consignment library that could be used to disseminate software components for reuse by developers, and developing a consignment library that could be used to disseminate programs that would otherwise be sold as shareware.

     My Master's thesis research is to analyze the issues involved in developing such a commercial Web site. The research (admittedly much broader in scope than is the norm) includes evaluating security concerns and solutions, library retrieval methods, and reuse and componentware theory, then developing a working prototype of such a system.

8. • Name: Johathan Kies
   • Email: jkies@vt.edu
   • Title: A Framework for Evaluating Communication via Electronic Media
   • Keywords: Video Conferencing, Networked Multimedia, Communication Evaluation
   • Abstract:
     The goal of this research is to develop a framework to facilitate the evaluation of communication via networked multimedia. The first stage of the research will involve the determination of communication influences (social, technical, and interface) by a comprehensive review of the literature as well as scenario enactment and case study evaluation.

     The second stage will involve the development of a communication task taxonomy to understand how different formats (persuasion, information gathering, information sharing) are executed by potential users and what cognitive models they employ.

     The third stage will be to test critical social issues (group configuration, group size, social status) and technical variables (frame rate, resolution, image size, video view, and camera control). Both objective measures such as conversational analysis, and subjective measures will be used to measure the effects of the various communication influences. One possibility is to test different communication modalities and compare them to the video-mediated modality. Sequential experimentation techniques may prove useful in reducing the experimental space of the study.

     The final stage will be to synthesize the findings into a comprehensive framework for evaluating communication through networked multimedia. The most appropriate evaluation techniques will be discussed as well as critical levels of socio-technical variables.

9. • Name: S. Gregory Micheal
   • Email: gmicheal@vt.edu
   • Title: Development of an Augmented Reality System for Encumbered Users
   • Abstract:
     A project is proposed which will examine the viability of creating an integrated augmented reality display system for use by persons engaged in activities which require the use of both hands. It is expected that the system also will be useful for training individuals engaged in construction or disassembly of complex systems. The user interface will include elements such as a head-mounted stereoscopic visual display, advanced auditory displays, and voice recognition technology in order to assist users with tasks such as assembly line work, bomb deactivation, surgery, computer repair, or in other environments which necessitate "on-line" assistance. The visual system will involve the use of a head-up display for enhancing the real-world scene with textual or graphic overlays. It is also conceivable that animation would be of great benefit to users attempting construction of components. The auditory system will include earphones for three-dimensional (3D) localization of objects (to complement visual schematics) and integrated external directional microphones to allow the user to request information without interruption of the task being performed. The system could also be manipulated through the use of a dataglove, flying mouse or other input device if only one hand were needed to complete the task. Topics of concern include the occlusion of real-world visual field, masking of real-world auditory signals, and limited movement of the user (due to the number of physical connections between the devices and drivers).

10. • Name: Dennis C. Neale
    • Email: dneale@vt.edu
    • Title: The Effects of Metaphorical Design on 3D Virtual Environment Information Visualization and its Impact on User's Mental Models During Navigation
    • Keywords: Virtual reality, data visualization, metaphors, mental models, navigation
    • Abstract:
      Virtual metaphors are beginning to replace the WIMP system or desktop metaphor for information access systems. In fact, database management interfaces are starting to pictorially represent query specification and portrayal of the conceptual result schema. Specifically, it is hoped that this research will address the effectiveness of systematically manipulating graphical, virtual environment metaphors for retrieving, storing, manipulating, and understanding large information sets. An operational approach will be taken to determine the impact that metaphors have on performance. Examination of the user's mental model for the depth and breadth of domain of knowledge will be considered. In addition, the effect of graphical aids to facilitate orientation during navigation will be manipulated.

11. • Name: Vicki Lewis Neale
    • Email: vneale@vt.edu
    • Home Page
    • Title: A cost/benefit analysis of drowsiness detection training methods
    • Abstract:
      Training program design, development and implementation is a costly budget item for any organization. Roscoe (1971) proposed a method of measuring transfer of learning in which the effectiveness of successive increments of training on one task, as measured by the relative incremental savings in learning a second task, is a negatively decelerated function of time devoted to pretraining or interpolated training on the former task. Roscoe's theory, which has not been substantiated with empirical data, is that transfer effectiveness functions will change according to how training is influenced by all the well known facts concerning conditions favorable and unfavorable to learning. As Roscoe states, "The most useful single measure for the educational strategist is the cost effectiveness of any educational experience." The idea of the diminishing returns of training for the cost is particularly relevant when one considers the array of training resources that are available, especially the vastly expanding resources of computer-assisted instruction.

      Roscoe's theory of transfer effectiveness related to costs can be considered in a broader setting. Effectiveness of a training program can be evaluated in terms of all costs relevant to the program. The proposed study would evaluate the effectiveness of three drowsiness detection training programs which are:

      1. Viewing video segments with the assigned level of drowsiness given to the trainee
      2. A lecture style training program, and
      3. A computer-based multimedia format.

      Costs to be evaluated are development, dissemination, performance measures, time to reach criterion performance levels, and reliability. Furthermore, individual differences will be evaluated for impact on cost. The practical implication is that training may be more effective for individuals who do or do not possess certain skills.

12. • Name: Suvit Nopachai
    • Email: snopacha@vt.edu
    • Title: The Development of International User Interfaces
    • Abstract:
      As computer usage increases throughout the world, there is a need for software that runs these computers to accommodate the growing number of users from foreign countries. Up till now, most software has been developed for U.S. markets. However, as countries from Asia, Africa, and Europe purchase more computers, they will also need software to run them. Thus, the special needs of users from different countries and cultures must be considered.

      At first, it would seem the usage of icons, accommodating number, date, and time formats of each country, and direct translation would solve the problem. Unfortunately, this is not the case. Icons and color connotations are not universal. What might be familiar or neutral in one country might be puzzling and insulting in another. Thus, as companies become more global, they will need to think about how their products will be used throughout the world. This global design of software has important implications on usability testing and its impact on the product development cycle.

13. • Name: Ray A. Reaux
    • Email: reaux@csgrad.cs.vt.edu
    • Title: Exploration of the Three Dimensional Object Design Space
    • Abstract:
      Current object oriented design techniques for HCI primarily focus on one dimension, the user's task with the HCI artifact. However, this perspective only scratches the surface of the interaction and only indirectly addresses the three major dimensions which make up an interaction: user actions, user cognitive processes, and system dynamics. The strength of explicitly using a three dimensional object design space rather than a single dimensional space is that each dimension provides a unique perspective on the interaction. The cognitive process dimension provides insight into a user's expectations, his or her mental model of the system, and the points at which the mental model differs from the system and task model. A system process dimension explicitly specifies the system's state model and identifies the system's expectations with the interaction and its responses. The user action dimension bridges the gap between the cognitive and the system dimension. The proposed research explores the three dimensional object design space, with emphasis on the cognitive process design space, perhaps the least explored of the three design space dimensions. It also looks at mechanisms for binding together artifacts derived from each of the three dimensions. The intent is that a design object can be represented by a "coordinate" in the three dimensional design space.

14. • Name: Will Schmidt
    • Email: schmidt@csgrad.cs.vt.edu
    • Home Page
    • Title: User Interface Issues and Information Storage and Retrieval Problems in the Building of a History of the Blacksburg Electronic Village
    • Abstract:
      The Blacksburg Electronic Village (BEV) has brought the power and resources of the Internet to many members of the Blacksburg community. Although other cities and towns have experimented with community networks, few projects have matched the scope of the BEV. Much will be learned from the evolution of the BEV and many different groups will benefit from the knowledge produced.

      This project's ultimate goal is to design and develop a multi-media design history of the BEV. The final product will be an interesting "browse mode" tool to learn about the BEV, as well as a powerful tool for analysis of the factors that helped shape the project.

      There are many problems to be solved in this project. One is the selection or the development of a user interface. Because of popularity of the World Wide Web (WWW), one possible interface may be Mosaic or another Web browser. This way project information could be desimminated to the world via the WWW. Will this type of interface be sufficient or should we design a new one?

      Other problems revolve around the large numbers and types of documents related to the BEV. What is the best way to organize the hundreds of documents and thousands of email messages? What type of indexing and searching techniques should be used? Should we use algorithms to automatically create hypertext links within and among the documents?

15. • Name: Michael P. Snow
    • Email: msnow@vt.edu
    • Title: Charting Presence and its Effects on Performance
    • Abstract:
      Virtual reality involves an attempt to create an illusion that the user of the VR system is actually present in a synthetic (usually computer-generated) environment. Not much is known about how various system parameters affect the illusion of presence in a virtual environment. In particular, there seem to be no quantitative data on which to base VR system design decisions. Also, while presence (or immersion) in virtual environments is a primary goal of VR, little is known about how this variable affects task performance. The goal of the proposed research is to provide a ratio-scale measure of perceived presence in a virtual environment, to explore the effects of a number of environmental parameters on this measure, and to relate perceived presence to user performance.

      This will be done using the psychophysical technique of paired comparisons in a sequential experimentation paradigm. Sequential experimentation will be used because of the large number of VR system parameters investigated. Yet to be determined are the users' task(s), associated performance measures, and the identity of the various VR system parameters to be manipulated. Tentatively, three experiments are proposed: a display parameters experiment, a system parameters experiment, and an environmental parameters experiment. All experiments would likely employ fractional factorial designs because of the difficulty in interpreting higher-order interactions and the need for efficient collection of data. Variables under consideration for inclusion in the display parameters experiment are field of view, visual display resolution, stereopsis, and color. Variables under consideration for inclusion in the system parameters experiment are scene update rate, texture-mapping, head-tracking, and sound. Variables under consideration for inclusion in the environmental parameters experiment are environmental complexity (number of objects in the environment), representation of the user within the environment, time spent within the environment, and the presence of another user within the environment. The final determination of which variables will be included and what task(s) will be performed awaits the completion of a literature review.

16. • Name: Yanchun Su
    • Email: anansu@csgrad.cs.vt.edu
    • Home Page
    • Title: A Distributed Multimedia/Hypermedia Learning Environment
    • Abstract:
      This project proposes developing a distributed multimedia/hypermedia environment which concentrates on educational purpose. Human computer interaction, network and operating system support for digital audio & video, and information storage and retrieval research will be applied to form a interactive learning, teaching environment. The service will accommodate distributed database access, automated information/media retrieval/transfer, messaging, conferencing, multimedia and hypermedia.

      Various new methodology and technology will be integrated in the design and implementation. A task-oriented interface with self-designing facility will be designed to fit user's specific needs; Real-time operating system and networking protocol support for distributed multimedia will provide a suitable transport mechanism to convey and synchronize multimedia information between workstations in a distributed system; Advanced ISR methods will make it possible to have intelligent search among large collection of hypertext information. Resource in the database will be carefully selected to meet the particular educational purpose. Both professors and students would take advantage of the environment to improve the education quality.

17. • Name: Madhan Subhas
    • Email: madhan@csgrad.cs.vt.edu
    • Presentation Date: 17 April 1995
    • Title: Automatic Building of Hypertext Links in Digital Libraries
    • Abstract:
      Vast amounts of text materials are now available in machine readable form for automatic processing. When stored documents are long, the manual construction of hypertext links between related documents may not be practical. In such circumstances, efficient and effective results may be obtained by building the hypertext text links automatically using advanced information retrieval concepts. One possible approach would be to index the documents at different levels - document, section, paragraph, sentence - and then query the different levels against each other. We then create links between those levels that have similarity above a certain threshold (this threshold is set by us). Clustering concepts may be introduced to speed up the query process. The success of this approach relies on reducing the time and the space complexity involved in the processing.

18. • Name: Raghu R. Thanjavur
    • Email: trraghu@vt.edu
    • Home Page
    • Title: Interactive Visual Information System: Wheel Chair Accessible Routes on Virginia Tech Campus
    • Abstract:
      Attributes (like names) of physical features like buildings dominate University or College campus maps. Information about departments, laboratories, and their locations is usually what the majority of the users (students, faculty, staff, visitors, and especially newcomers) look for in these maps . Often such information is very difficult or impossible to obtain from these maps. These maps are of some use only in combination with indexes like telephone directory neither of which is readily available when the user needs it the most.

      Over time the user develops a mental map (or a perception) of the different locations (that she/he happens to visit frequently) and route(s) between these locations . Even these mental maps may not be accurate enough to make the user take the shortest route between any two locations. A wheel chair bound handicapped user cannot traverse most of the routes a normal user can. Also such a user can have access only to departments or laboratories or other places that have elevators or special ramps.

      My Master's thesis research will focus on,

      1. computer screen display (effect of factors like color, hue, etc., on the formation of user's mental map) design,
      2. user interface (like pop-up menus, visualization sliders, touch screens, etc.,) design and,
      3. performance measures of these designs in the development of a wheel chair accessible route information system for the Virginia Tech campus that will
         1. enable the user to query with great ease,
         2. rapidly retrieve query results and,
         3. help form a very accurate mental map of the shortest route (the user has to traverse) and of the important landmarks on the way.

19. • Name: Stephen Williams
    • Email: williams@csgrad.cs.vt.edu
    • Home Page
    • Title: Characterization and Simulation of Digitized Video from Different Genre'
    • Presentation: 4/10
    • Moderator: Greg Micheal
    • Abstract:
      As video-on-demand and other digital video services move from concept to realization, network designers find themselves in the position of having to create new protocols and switching schemes to make the transfer of high-bandwidth, low-latency video as efficient as possible. Simulation of these systems will require a standarized, realistic suite of data traces to use. This system should generate synthetic video traces that have a high correlation to a random sampling of actual videos in the same genre' -- in terms of frame size distribution and length. This thesis will define the genre' categorize and sample a variety of actual video sources within those genre', and create a system to generate synthetic frame traces for use in network simulation and analyis.