Enhancing Community Awareness in a Large and Distributed Organization
- Thesis Proposal Addendum -

azhao@cc.gatech.edu

Feb. 1, 1999

Outline

Introduction
System Design
Implementation Plan
Evaluation Plan
Research Contribution

Introduction

The goal of this research is to examine ways of supporting and enhancing community awareness in a large and distributed organization. More specifically, the research framework can be captured as the following:

Goal	Mechanisms	Evaluation Methods
To enhance people's general knowledge about the community: its states, activities, and public opinions.	Provide ubiquitous access to information about the community through interactive displays in the background of the work environment.	Community surveys and informal interviews before and after the introduction of proposed technologies.
To gain a sufficient user population.	Add incentives such as interesting content and public recognition.	Participatory observation and technology supported statistical information recording and analysis.

Next, I will describe an overview of the system design guided by a set of high-level design directives. Then, I will present the implementation and evaluation plans. Finally, I will discuss the contribution of this research.

System Design

High-level design decisions:

On-desktop vs. off-desktop
I will primarily focus on on-desktop technologies. In the College of Computing, people have easy access to desktop computers and there are a great number of these computers. On-desktop technologies may easily reach a large number of people and are the first steps toward a ubiquitously accessible information environment.
However, this does not preclude exploring off-desktop alternatives when resources become available. Information in the real world may be more convenient to access, especially to those who are disconnected to the electronic world during most parts of a working day.
Information content
The community awareness information can be categorized by its gathering methods:
- automatically collected information as basic content: upcoming events extracted from weekly article in the git.cc.talk newsgroup; news articles pulled from popular Web sites, such as slashdot.org.
- authored information supporting flexibility and creativity: people can submit news or events by sending messages to a well-known address (to be set up), for example community-news@cc; or, people can fill out a short form in the desktop community application(s).
On the other hand, information delivered can be categorized by its level of detail:
- 1st-level detail: for example, the time, title, and author of an article; broadcasted to all information displays, can be queried from small-space displays.
- 2nd-level detail: for example, a brief description of a story; broadcasted to full screen displays, can be pulled from other visual displays.
- 3rd-level detail: for example, a Web page describing the full content of the story, with graphics and animations; linked from all visual displays and accessible from an existing Web browser (separate application).
System capabilities
I will focus on building the community awareness supports. More specifically, the system will have:
- a calendar of events: showing scheduled community events of the day and remind people of the events.
- a community bulletin board: showing news articles, informal announcements (e.g. "Going to lunch in 30 minutes -- anybody wants to join?"), and informal polls (e.g. "Are you for or against the SAC-II fee increase?"); it will also have a form interface for people to submit new items.
- a community stage: cycling through a set of Web pages with displays of the calendar and bulletin board inserted in between. It broadcasts items on the playlist on root windows and screen savers.
- a community radio: cycling through a set of sound files with short announcements and reminders inserted in between. It broadcasts items on the playlist onto the networks and through radio waves.
- backend data processing and storage units: handling system-level data.
One idea of how to attract people to use the system is to let people host their own "shows": for example, "Bob's Blues programming every Tuesday 1-3pm on the community radio", or "Joe's funniest Web sites this Friday lunch time on the community stage".
In terms of information displays, the system will have:
- a small-window display: small ticker-tape style application that resides on the edges of users' desktops (approx. 128 pixels thick; raises to the top of the desktop when the mouse pointer enters and lowers to the bottom when the pointer leaves; fades in/out short information items); displays 1st-level-detail information and pops up 2nd-level-detail information when requested.
- full-screen displays: large size visualizations that sit in the desktop background or as a screen saver; display 1st- and 2nd- level-detail information on the bulletin board and community stage.
- receivers and playlist editors: controllers and status viewers of the community stage and community radio.
Interpersonal communication features such as awareness about on-line activities (e.g. chat-room crowdedness) and presence (e.g. a person's availability for interaction) may bring in more incentives and increase the level of convenience/usefulness of the system. However, they do not directly fit in the research framework and will be de-emphasized. Access to these features will be provided as links to applications outside of the scope of this research.
System instantiation
Due to availability of expertise, I will focus on system development on Unix platforms. Since computers in graduate student offices are changing to Solaris and IRIX systems, my target audience will include a large portion of the graduate student body and some faculty members who use the same Unix systems.
People can choose to automatically start those community tools when they login. Alternatively, the screen saver tool can be run when nobody is using the console; and when a user comes by, the screen saver lowers itself to the bottom layer of the desktop, hence revealing the login/password dialog prompt. This way, all computers can be set up as information kiosks and provide easy access to community information.
Computer systems can also be placed in a non-office setting. The same kind of applications can be displayed on large screens in the common areas, making community information access more convenient.

Implementation Plan

System implementation will be incremental, in the order of:

Community news backend and basic ticker tape.
Full-screen support, root-window/desktop-background support, and screen saver support.
Community stage backend, basic player, and basic playlist editor.
--- the "I'm done with programming" line ---
Community radio backend, basic receiver, and basic playlist editor.
--- the "I'm really done with programming and the rest will be implemented only if I want the fun" line ---
Optional: links to other applications.
Optional: on-line activity and presence awareness support.

The implementation process will also be iterative. For example, the transition from one information item to the next, or from one full-screen display to the next, will need to be experimented and revised many times throughout the implementation process in order to better satisfy the design guidelines (see "do not distract").

Evaluation Plan

To evaluate the proposed technologies, I will need to know whether people use them or not, and whether they have any effect on community awareness or not. More specifically, the evaluation process includes:

Before- and after- surveys and interviews
The surveys will quantitatively measure people's general knowledge about the states and activities of the community. Some potential questions are:
- Name three recent community events. Briefly describe what they were about and who the primary participants were.
- How often do you attend these events? How often do you miss the events that you intended to go to?
- How are you familiar with people in research group XYZ? How are you familiar with their research?
The informal interviews will help gather qualitative opinions about the community. They may provide a more subjective evaluation of community awareness and insights into problems in procedures and technologies: more specifically, whether people shift certain amount of communication load to the proposed system from other media, whether people expect or rely on certain rhythms of interactions, whether informal messages are more valuable than official announcements, and in addition to all these, why do these happen. They may also generate new ideas on how to solve these problems and improve community awareness.
To avoid people answering favorably to be polite, I may need to have someone not related to this research project to deploy the surveys and conduct the interviews.
By comparing the before- and after- survey and interviewing results, I will be able find out the changes and assess the influences of introduced technologies.
Participatory observation and logging
The system will log statistical data related to its use, such as the number of submissions it receives during a period of time, the size of the playlists and how often they are modified, how many people use the desktop tools and for how long, what component(s) they usually use, etc.
Further more, I will participate in using the system and observe what type of content people generate and of what level of quality, etc. These analyses will help me discover patterns of use and common practices (for example what type of messages are more often), and verify whether there is increased use.

Research Contribution

This project will build on technologies partly similar to those of PointCast but with an emphasis on pushing information delivery harder to the background of the work environment. It will also extend these technologies to include community event reminders and community artifacts such as the community stage and community radio. In addition, it will study the impacts that these technologies make in an organization.

Furthermore, this project distinguishes itself from previous research such as that at Apple by focusing on the desktop because a single shared space is not available. This project will also study alternative forms of information delivery and authoring. In addition, the local College of Computing community is larger and more diverse, probably posing more complicated, unanticipated problems.

In summary, the contributions from this project will lie in the design of the right mix of both existing and new technologies to enhance community awareness, and the evaluation of the effects on community awareness that these technologies may have. Ultimately, this research will help gaining a better understanding of how to use computer technologies to improve community awareness in a large and distributed organization.

Copyright © 1999 Q. Alex Zhao	Original Proposal Document (PDF)
azhao@cc.gatech.edu	Original Proposal Presentation (PowerPoint Show)

Enhancing Community Awareness in a Large and Distributed Organization - Thesis Proposal Addendum -