3 Community Information Services
The primary goals of the Boulder Community Network are to develop a comprehensive information service, a distributed discussion mechanism, and specific training and access services for underrepresented groups. Supported in part as an NTIA demonstration s ite, BCN is building a large mass of community information, working with public and commercial information providers to insure production level services. The information services are extensive, ranging from the timely (e.g. weather updates every five min utes) to the archival (such as a library image database of local historical photos); from social services (United Way, YMCA, job training, etc.) to commercial information (e.g. restaurant and shopping guides); additional categories include education, gove rnment and current events. Access is via the Internet, public modem pools and kiosks in public sites.
Target communities such as seniors, at-risk youth, and low-income families have been given personal access accounts, training, and the opportunity to mount their own information servers. The community network has also initiated on-line public forums. Th ese discussions, built on a web-netnews bridge, permit wide-spread involvement in timely issues in the community.
Yet it is the relationships around the community network that are perhaps of greatest importance. These partnerships build on the common ground of the network to develop innovative K-12 curricula, reengineer social services, and bring new resources on-li ne. This paper discusses the issues raised and lessons learned in the development of the community network and focuses on the role of the network in creating and supporting new collaborations.
The breadth of participation in BCN is unusual. The University of Colorado at Boulder has had a major role in national and regional networking. Recently the university began a wide-ranging research initiative to understand the impact of networking on so ciety and commerce. The Boulder Valley School District (BVSD) is a national test bed for the systemic application of networking in K-12 education. Supported in part by the National Science Foundation and the Annenberg Foundation, BVSD has become a leadi ng model in deployment strategies for K-12 networking as well as training programs and curriculum development. The BATMAN project (which includes national labs at NCAR, NOAA, NIST and USWest as well as at CU, local schools, and high-tech corporations) ha s been running an ATM research network for almost a year. Applications ranging from videoconferencing to high-bandwidth networking are being used in educational and scientific projects. Beyond these organizational efforts are the individual volunteer co ntributions of many Boulder citizens, drawing from 250,000 residents, encompassing a broad spectrum of economic and ethnic populations, with a mix of urban and rural communities.
At the operational level, the BCN server and the coordinator staff are housed at the University of Colorado, where they can draw on a pool of expertise that mounts institutional web sites. Three committees, on Management, Policy and Technology, set direc tion and are composed of community volunteers and University faculty and staff. As indicated below, this is an extremely low-cost model. That it is also a high participation model appears beneficial in terms of the organizational and human relationship s that have been formed.
BCN has an extremely broad base of information. The county and several cities contribute governmental agendas, minutes, boards, regulations, maps, voter information and election returns, etc. Educational resources at the University and district levels i nclude calendars, announcements, governance information, student newspapers, etc. The Human Services Section includes United Way listings for every social service agencies, providing location, services offered, bus route access, languages spoken, and oth er vital information. The Weather Center updates conditions(temperature, wind speed and direction, pressure, etc.) every five minutes with ongoing twenty four hour plots, forecasts for fifty seven Colorado zones, and access to national servers. The Busi ness section has business newspapers, Chamber of Commerce information, and restaurant guides that include on-line menus for most County restaurants (searchable by geographic, alphabetic and culinary indexes), shopping guides, and maps of local golf course s. There are equally comprehensive listings in other categories such as the Media Center, the Transportation Center, and Community Centers.
On-line access is especially beneficial for three types of information: volatile, current and archival. The volatile information includes weather updates, election returns posted every six minutes, road conditions every two hours and significant environm ental events as they happen. No other medium can present such volumes of volatile information in as manageable a form. Current information ranges from city and county council meetings and minutes, through cultural information (plays, movies, concerts, e tc.) to school district material, neighborhood group listings, shopping and restaurant guides, business reports, etc. Aggregating this breadth at a single source and making it into hypertext (connecting, for example, movie schedules with reviews) adds re al value. The archival database includes ÒoldÓ current events, rules and regulations, library access, and historical photos. That such information is available on-line and keyword searchable greatly enhances its use.
The major challenge lies in integrating the information on a topical basis although it is being furnished on an organizational base. For example, hiking trail information is supplied by a variety of sources: city, county, state, or any of several federal agencies. Users are oriented only by location and want to see the information oriented by geography rather than ownership. This integration of information is essential but difficult, and requires both a common nomenclature and an ongoing maintenance. The nomenclature should establish classifications and key words to assist in this aggregation. While hypertext allows the inclusion of an item in multiple locations within information space, the confusion of navigation suggests keeping the complexity to a minimum. (For example, many pages have a local home button that will take user to a different location than the way they entered.)
The topical integration of information in BCN is done by coordination mechanisms called centers. Each center in BCN has an ÒownerÓ who is responsible for organizing their space. These owners in turn rely on existing information aggregators where possibl e. For example, the Human Service Center provides information aggregated by the United Way and the County Social Service Department; the Business Center links to a local restaurant guide and a business newspaper that already had gathered materials for ot her media.
While this approach is efficient and economical, it does raise some issues. Being an owner of a center can attract significant amounts of work and occasional discord. Keeping the number of information providers limited by restricting postings to aggrega tors limits those who do not wish to work with an aggregator. To remedy that in the business sector, BCN is developing Boulder County Business Cards, an on-line form (coupled with a review process) to post small business information into a searchable BCN database.
Much of the power of the Web is derived from the efficiencies obtained in using hypertext. Yet almost all archival information and most current materials are serial in structure. In mounting this information on-line, it is important to use appropriate s trategies for internal and external hypertext links The file structure should reflect the content and how it is generally accessed. Information that is likely to be used in its entirety (such as minutes, regulations and plans) should be kept as a single file with internal hypertext. Files that are accessed for limited pieces of information (movie reviews, medical facts, etc.) is best broken up to minimize needless delays from downloading.
As community networks proliferate, regional information providers will want to post the information once for community networks to share. The community networks want to orient the information for their own citizenry. For example, a regional bus provide r will want to upload schedules once, and then have individual communities restructure the data so that bus routes involving those communities are presented at a high level and in an appropriate fashion. These considerations suggest hypertext database ar chitectures that allow such pointing and reorienting.
Information architectures should be constructed with issues of data maintenance in mind as well. One of the keys to data maintenance is to insure that the owner of the information is also the mounter (so that updates are part of the ownerÕs standard chan ge procedure. This indicates that the owner should structure data with localities in mind. For example, at the state level, the Division of Wildlife should list its offices in a geographic fashion that permits communities to point directly to their own regional information.
While BCN displays most of its information in text to accommodate the large number of ascii clients, there is some information best conveyed in a graphic format. Road conditions are best indexed through a hot map; bus routes best displayed graphically ra ther than by text description. BCN will likely not convert such scanned information into textual equivalents.
It is essential that data be current and correct. Users have a higher expectation that, compared to hardcopy, on-line information is up-to-date. In the distributed architecture of BCN, responsibility for the integrity of information is spread over many people and organizations. Several basic approaches are being used.
Posting of information is being made part of existing work processes and automated as much as possible. Data can be sent and processed via email. Senders are being equipped with macros to make possible the inclusion of the documents in HTML format. It is very important that the agencies have network access themselves so that they can view and verify their own data.
Ultimately, there must develop a culture of maintaining data. One step in that direction is to make the posters themselves dependent on the quality of the information. For example, making internal governmental interaction dependent on shared on-line age ndas and regulations can insure the accuracy of that information for the broader community.
There are several options in dealing with stale data. One can mount a posting date along with the data, and leave the individual to interpret the accuracy of the data. Alternatively, one can have expiration dates and remove the data. The expiration dat e can be connected to a mechanism that informs the poster of the pending expiration. BCN utilizes both methods. Maintenance on links is an increasingly automated task, due to new web tools. When invalid links are found on BCN, there is still a policy i ssue in determining what to do with them, as removal of the link may introduce other problems.
There are several important indices of public information to compile for popular use. For example, a recent search produced four different lists of ÒK-12 schools in the state that have a network presenceÓ, with only limited overlap among the lists. It w ould be useful to have a single definitive list compiled by an appropriate resource. (In this instance, that might be the State Dept. of Education.) Similarly, for BCN, it would be helpful to have a single source (assembled, perhaps, by the State Libraria n) that pointed directly to the Colorado sections of the major Federal data repositories.
One area that has been particularly difficult for BCN has been the uncertain cusp between commercial information and economic development service. Tourism, transit and medical information may come from commercial sources, but they are valuable local reso urces and their dissemination can improve the economic and social health of a community.
However, in directly mounting commercial information, BCN does not want to compete with commercial services either in its current legal status as a research project of a public university or in its likely evolution into a 501c(3) not-for-profit. With a n umber of commercial Internet info-mall providers in the area, BCN is selective in the commercial information that it will accept. For services with unproven on-line economic potential such as the Boulder County Restaurant and Shopping Guides, BCN is will ing to act as an incubator, providing a temporary platform on its own server while the service is evaluated. (The restaurant guide demonstration was highly successful, and the information provider is now moving to a commercial server, which BCN will cont inue to reference.) BCN also mounts key commercial data that is widely used in the community. This includes movie and concert listings, commercial airport transportation schedules, and ski reports. Lastly, Boulder County Business Cards and similar proj ects are intended to explore how traditional community information services such as supermarket bulletin boards can be provided on-line.
Other issues concern freedom of speech rights and censorship obligations. To date, BCN has finessed these by focusing development efforts on groups and services that are clearly within the community norms. But it is anticipated that, for example, the So ciety for Hamster and Duct Tape will approach BCN at some point with a request to post their information. Currently, as a research project of the university, BCN likely has legal protection in making these decisions. As BCN moves into another organizati onal status, these issues will emerge. One solace is that, unlike the broad Internet, there is a physical context in community networks that does have Òestablished community standardsÓ.
While it is important to provide on-line search tools, the task is daunting. The tools themselves are still immature. BCN is looking at utilizing the Harvest search tool, but there are challenges in effective use. One problem is in the diversity of the information space being indexed. There is a wide range of domains that are included in BCN, and some of them (such as colorado.edu) are extremely extensive and include a lot of information that is not of interest to the general public.
Web browsers are a new and volatile area . While BCN has a standard browser for use in kiosks and public sites, there is a growing collection of browsers, with an increasing divergence of features. BCN periodically tests the major tools against its dat abase and, in an associated style sheet, recommends a lesser common denominator of constructs for information mounters to employ. One particularly vexing area has been tables. Almost all information providers use tables in their information and it is di fficult to preserve layout among browsers.
Another outgrowth of this immaturity is the need for training information providers in the use of HTML editors and style sheets. HTML tools found in common application packages (such as Microsoft Word and Lotus Notes) are of limited value and there is fr equently a need for HTML itself. BCN is intending to run periodic sessions on HTML and connects people to developersÕ lists.
One of the most vital aspects of community is the use of discussion to shape opinions and resolve local issues. While the Unites States has had a long and rich history of public discussion forum, the last several years has seen an erosion of such opportu nities. Many of the limitations have stemmed from logistical or financial constraints. Electronic discussion forums obviate some of those constraints and have other apparent values as well. For example, the distributed and asynchronous nature of on-lin e newsgroups could permit some groups, such as disabled, rural or otherwise isolated citizens, to now participate.
Inclusion of public discussion has been a basic tenet of BCN from the start. In the past few months BCN has implemented a web interface to netnews, allowing users to use forms and hypertext for posting and reading. Postings are generally anonymous, alth ough there are places for users to specify names and email addresses. The news group mechanism will permit the groups to be exported to a larger area as well as to link into national discussions such as the recent NTIA Town meeting.
The public discussion area has not yet become vigorous on BCN. Access is limited at this time, and there has not been a major issue that has drawn strong opinions recently in the area. It is likely to grow in the next few months, and that the mechanism will serve as the basis for some of the community curriculum outlined below.
BCN is built on a distributed model of access. The basic premise has been that in an electronic society many people will derive their network access from their work or educational relationships. (Perhaps 40% of the Boulder community today has such oppor tunities through their affiliation with the university, the school district, corporations, labs, etc.) Many other citizens buy commercial network access accounts from Internet access providers. By not dealing with large account and modem access issues, BCN has been able to focus on two other important access paths: kiosks and target community services.
Access kiosks are being placed in libraries, county buildings such as Motor Vehicle registration offices, recreation centers and public buildings such as hotel lobbies. The kiosks are hardwired to existing Internet access. The kiosks attract significant attention and provide means for public BCN access. In addition, the state library modem pool can be used to reach BCN.
The current state of on-line kiosk design is quite primitive. It is difficult to secure either the operating system or the browser against mischief. Connectivity options are limited to hardwired connections at this time. BCN is developing approaches to a dial-up kiosk that would hold initial screens and archived local data while connecting to a server via SLIP/PPP when needed for other information. BCN is also working with Cablelabs, the R&D arm of the cable industry, to deploy several cable-based kio sks into prominent hotel lobbies in the area.
The BCN approach of distributed access leaves out those segments of the community without the means or affiliation for obtaining network accounts. This is particularly unsettling since many of these underrepresented sectors can get particular benefit fro m networking. The net can help compensate for past inequities; the technology is amenable to certain physical limitations; the computer attracts many for whom other information resources (libraries, newspapers ) are not engaging.
BCN has therefore targeted several particular community sectors for additional services. These target communities include the local Senior Center and Project Self-Sufficiency. In each of these communities, BCN has provided training, personal access acco unts, and the opportunity for individuals and groups to mount their own information.
There are several values to this approach. It is clear that the access services are being used directly by the intended audience instead of being misappropriated by people that might have other alternatives for access. Moreover, the training can focus o n items of direct interest to the target groups, increasing the likelihood of continued use. Lastly, and perhaps most importantly, the opportunity to mount their own information has proven galvanizing for each of the groups. The seniors have developed a n extensive set of information, both of an organizational nature (such as activity schedules, club news, etc.) and of a personal nature (several seniors are developing their own home pages).
The early results from the target communities has been quite remarkable. For seniors, the technology is particularly powerful, in its ability to compensate for limited mobility. Even minor features, such as the ability to change font size in browsers, h ave benefit. For kids, the network is the preferred way to acquire information, with an attraction that neither newspapers or libraries have for them. The next generation, in its use of Nintendo, computers, and even clocks, prefers digital.
The specific purpose of BCN was to develop on-line community information and promote community discussion. Along the way, a remarkable number of new associations have sprung up among the members of the community. These connections include enhancements t o social service offerings, new educational opportunities, and reengineering governmental processes.
One of the more impressive serendipities concerns the Red Book, which has served as the definitive information source on Boulder County United Way agencies. The book has had internal audiences as well as external use; one of the regular activities of soc ial service agencies themselves is referral of clients to other agencies as appropriate. Unfortunately the book has always had limited value due to out-of-date information. BCN has put this information on-line and is instituting processes to update it r egularly.
In doing so, a number of unexpected benefits have occurred. An existing United Way service, called First Call for Help, had been intended to provide agency information to the general community. However, it went largely unused while people called either 911 and the Public Library with their social service needs. As a result of BCN, both the 911 dispatchers and the public libraries now have access to the Red Book and can answer requests that they previously had to turn down. Indeed, the United Way is in tending to save money by eliminating First Call.
Instead of providing the Red Book to the internal audience of agencies, United Way now provides them with floppies with the on-line information. Many of these agencies are turning on computers (which have been donated over the last few years) for the fir st time to use the Red Book. This has lead several groups to overcome their reluctance to use computers and ask for additional tools such as word processing and spreadsheets.
Another notable instance is the development of community curriculum, where the issues and the information of the community serves as the basic material for many K-12 activities. The curriculum builds on the data and the people of the area to add relevanc e, real data and collaborative mechanisms. Through links with government, senior centers, and scientific labs, subjects from mathematics and science to social studies and writing are being revitalized by using local material and engaging new community-ba sed processes.
Over the next two years, the Boulder Valley School District will be developing a series of curriculum modules that utilize local data sources, such as atmospheric conditions, water readings from across the Boulder Creek Watershed, census data, and local z oning and governance information. Curriculum will stretch across grades and disciplines, using local data as the basis for math and science activities. A second component will involve students working with the public discussion mechanisms on BCN to inve stigate community attitudes on current local issues. The internal support of each of these initiatives will be a central GIS system, and clients in the schools, that can be the hub for automating and scaling the delivery of information to classes. The s chool district will use both BCN information and discussion mechanisms to show the community at large the educational progress of the students. The same mechanisms will help air the issues of school governance and policy that are widespread in a site-bas ed system such as Boulder.
A third area of unexpected benefit is in the small partnerships that crystallize around the activities of BCN. Examples abound, in the work of business college students showing United Way agencies the uses of computers, in the enhanced communication betw een neighborhood associations and city planners, and in the senior high educational partnerships with the Senior Center. These associations may be temporary or long-term, but community members from opposite ends of the spectrum are pairing up to assist e ach other. In particular, people from the large technical community in town have found new opportunities, comfortable but satisfying, to contribute to the common wealth.
There are several ways to measure the impact of BCN. One of the most obvious is traffic. In early 1995, there were about 25,000 requests per week on BCN server. A statistical analysis shows a median of 4-6 minutes connect time per session with an avera ge of 8 requests. Recent favorite topics include: menu guide, weather, student newspaper and the library. There were about 600 request a week from public kiosks. These numbers continue to grow rapidly.
A second measure is the size of the database. Here, one must focus not on the direct megabytes on the BCN server, which has little relevance in a hypertext, hypermedia world, but on the depth of the tree of community resources. The depth of the BCN tree is around 30 links.
A third metric of a community network is its leverage to increase the ÒelectronificationÓ of the community. The underlying concept here is not a blind adoration of the on-line life, but rather a measure of network utilization by those interactions that c ould benefit from an on-line component. The benefits of the on-line component may be direct, in ease of access to accurate and timely information, or indirect. In the latter category, the city of Boulder spends $40,000 a year promoting a reduction of tr affic in downtown; however, the only place that many city services are currently accessible is downtown. Electronification of these transactions could be of considerable benefit. By this metric, BCN has had little success to-date. The gaps are small bu t real, in the inability to mount key databases or in the incompatibility of key software. Moreover, there is considerable inertia in doing things the existing way. To those of us used to the malleability of technology, the rigidity of institutions is u nexpected and frustrating.
Perhaps the hardest assessment is the measurement of the overall cost/benefit to the community. BCN has chosen a deliberately scalable and low-cost approach. Operating costs are less than $50,000 a year, mostly for a coordination staff position. Additi onal costs are largely hidden in late nights and bootstrapped systems. Some hidden costs, such as volunteer efforts, are also some of the benefits. The partnerships that have grown have significance beyond the project. City, county, university and distr ict officials are collaborating on a host of other issues for the first time. People who may not have otherwise met and enriched each otherÕs lives are coming together on the common ground of the community network.
The most immediate challenges for BCN are to stabilize the infrastructure and extend the offerings. Automation of information posting and new techniques for data maintenance are needed. An extensible nomenclature is being developed in concert with a sta te-wide on-line information effort, with the goal of having state managed data mounted in a fashion consistent with community networks in the state. Public access methods also must be enhanced. As indicated earlier, kiosks are important but highly probl ematic. Some form of broad modem access should be engineered, permitting public use of BCN while restricting general network access.
Target community efforts need to be solidified. While there are now pockets of strong interest and activity in both the Senior Center and Project Self Sufficiency, usage should be broadened to create the critical mass that is in turn self-supporting and self-growing. The disabled community in Boulder has been reluctant to adopt a technology that can help them. Among information areas, the Health Center has been slow to develop. Two works are in progress - a current local medical information service (a so-called flu du jour) and integration of local and national information. BCN is working with an NIH grant to develop such an integration of cancer information and service.
One of the most difficult and important tasks in front of BCN is to develop a long-term model that is financially sound. While there is significant capitalization cost for BCN, operating costs are quite modest. Much of that low cost is due to the volunt eer associations that are the heart of BCN. Some of the activities associated with BCN, such as the target community services, must likely receive their permanent support through the general social services of the community. But the core of BCN, the pro motion and coordination of local on-line information, needs a dependable core of funds. Fortunately, several viable models for long-term support are now emerging. Many local corporations have realized the contribution of BCN to the quality of life and feel it important for their employees. There are also strong possibilities for corporate sponsorship in return for links from a recognition area to their home pages. Perhaps the greatest potential is in a growing relationship with a local Chamber of Com merce. The alignment of missions of BCN and the Chamber is quite high, and the relationship between a strong community and strong local business environment is clear.
Lastly, there is a broad research agenda that the university has undertaken regarding the impact of technology on community. Multiple departments across campus are examining issues of information interface and organization, of electronic commerce and mar keting, of the social benefits and costs of network-based interactions. These studies will take years to develop the longitudinal data that can provide the critical analyses necessary to understand just how our world is changing.
The Internet has always been composed of two elements - a physical switching network that operates at the bit and packet level and a social network that shares a common belief that the switching network could be a powerful tool of great benefit to our wor ld. Clearly much of both the physical and the social fabric of the Internet has changed in the last twenty years. Yet threads persist, and the theme of harnessing the physical infrastructure to improve the social infrastructure persists today. Communit y networks are a part of that thread.
There have been some eloquent discussions recently about the issues associated with physical and virtual community. It is important to note that it is not an either/or choice. Virtual community can amplify the physical associations. On the common grou nd of the community network can be found, again, the sense of place that we all need.
Ken Klingenstein is Director of Computing and Network Services at the University of Colorado at Boulder. He is also a principal investigator on the Boulder Valley Internet Project and the Boulder Community Network. He has been chair of the Federal Netwo rking Council Advisory Committee, Vice-President of Farnet, on the board of CAUSE, and held other educational and networking positions. He is also an adjunct faculty member in Telecommunications.
Ken Klingenstein has been active in national and regional networking since 1985, serving as a member and former Chair of the Federal Networking Council Advisory Committee (FNCAC), co-founder of Westnet and Colorado Supernet, and in other similar roles. H e regularly presents papers and seminars to professional networking and computing groups, including Educom, INET and CAUSE and has testified before Congress on topics in networking. He teaches advanced courses in network management. He received his Ph. D . from the University of California at Berkeley in Applied Mathematics.