Hi all...I'm a new subscriber to this list. I just learned about Wikipedia for the first time a couple of weeks ago and think it's an impressive new step in the evolution of the web as an information commons. I have some thoughts about directions in which I think it could be taken further, and I hope that some people here could offer your reactions and advice.

I'll begin by giving some particulars that explain my personal interest in Wikipedia, after which I'll touch on ways that my specific interests might have more general relevance.

I edit a publication called "PR Watch" (www.prwatch.org) that monitors deceptive and manipulative public relations campaigns. My particular interest in Wikipedia stems from my long-standing desire to develop an information base about PR firms, PR campaigns and deceptive PR front groups -- organizations that claim to represent some public interest while actually serving the agenda of a client or industry whose sponsorship of the organization is often hidden. For example, the APCO PR firm set up a front group called "The Advancement of Sound Science Coalition" (TASSC), whose stated mission was to promote science-based decision-making about public health policies. In reality, TASSC was a front group for the Philip Morris tobacco company, whose primary mission was to attack the U.S. Environmental Protection Agency's risk assessment of secondhand cigarette smoke.

A number of organizations, including PR Watch, have tried over the years to develop databases, printed directories or other systems for tracking PR campaigns, front groups or industry-sponsored "think tanks." Typically, however, these efforts have run up against limitations of time and staffing. There are thousands of think tanks in the United States alone, with new ones forming all the time and ever-changing personnel. Awhile back we added a section to the PR Watch web site called the "Impropaganda Review" that serves as a "rogues gallery" of a few notable examples, but the Impropaganda Review remains in what Wikipedians might call the "Nupedia stage of development." We've developed a handful of articles, but nothing approaching a comprehensive directory. Rather than attempt the impossible task of trying to expand the Impropaganda Review all by ourselves, it seems to me that the Wikipedia model would be a great way to invite the world at large to contribute.

There are, however, a few issues with this idea:

First, the nature of the work we do at PR Watch is inherently controversial. Not everyone would agree with our characterization of TASSC as an "industry front group," and I'm wondering if attempting to develop an open-source directory of information on controversial topics could succeed.

Second, the type of directory I'm interested in developing requires the imposition of a bit more structure than Wikipedia imposes. Each organizational profile in the "Impropaganda Review" contains the following sections: * A general description * Personnel * History * Funding * Case studies * Contact information * Related information resources

It isn't possible to obtain all of that information about every organization (for example, some groups don't disclose the identity of their funders), but I would want to strongly encourage contributors to try to follow that framework.

Thirdly, I think it would be ideal if the information I want to develop could be structured as a relational database. For example, there would be a separate article for each ORGANIZATION, but also a separate article for each PERSON, with a many-to-many relationship linking the two categories. In other words, each article on a person would include a list of the organizations with which that person has been affiliated, while the article on an organization would include a list of all its related personnel.

That's my basic concept. Now, here are my thoughts that might be of more general relevance to Wikipedians:

(1) The Wikipedia model could be applied to a number of other uses if the software could be modified to support something that users experience as "relational databases of structured objects." Right now users experience it as a non-relational database of ONE type of object, namely an "article." The ability to insert hyperlinks to other articles provides a quasi-relational capability, but it's not truly relational. If, for example, I put a hyperlink in the "Leon Trotsky" article that points to the "Mensheviks" article, the software doesn't automatically create a corresponding link pointing back from "Bolsheviks" to "Leon Trotsky." For certain types of knowledge databases, it would be nice if the software could be configured to do this automatically. Moreover, the structure of each "article" is entirely free-form, and many knowledge databases try to impose some sort of structure. For example, scientific articles are typically structured to include an abstract and footnotes in addition to the main text. If the Wikipedia software could be configured to impose this sort of structure on individual articles, it could be used to host a number of valuable, specialized knowledge bases in addition to the general-purpose encyclopedia for which it was designed. Moreover, that information could be extracted in a variety of formats for different presentations. A database of "people," for example, could be sorted in different ways according to different fields, such as date of birth or city of residence.

(2) Supporting this sort of relational database structure may also become important as Wikipedia continues to grow. One of the great virtues of Wikipedia is that it supports collaborative synthesizing and summarizing of information. Most of the information in Wikipedia's articles can be found elsewhere on the web, but doing a "Google" search often forces users to wade through mountains of irrelevant information before they can find what they're looking for. As the number of articles on Wikipedia continues to grow, the sheer volume of information may also become daunting for users. Comments are already being made to the effect that there are "too many articles" about certain topics, and suggestions are being made to split off certain topics into separate "sister" wiki projects such as a Wikiteer (gazetteer) or a Wikipediatlas. However, creating separate sister projects also leads to information fragmentation and redundancy. If the Wikipedia software could support structured information objects in addition to simple text articles, it might be possible to encompass a number of "sister" projects as parts of a single whole, while letting users decide what types of objects they are looking for. For one user, it might be a general-purpose "encyclopedia"; for another, a "gazetteer" (an encyclopedia of locations); for another, a "who's who" (an encyclopedia of people).

Anyway, those are my thoughts for the moment. I'd appreciate feedback, and especially advice on how I might go about starting a Wikipedia-style open source information base on PR firms, PR campaigns and front groups.