mark nottingham

The Semantic Web’s Dirty Little Secret

Tuesday, 30 December 2003

Semantic Web

Browse through the W3C Semantic Web pages and you’ll see this notice in a few different forms:

Additional support for this activity has been provided by DARPA under the DAML program.

Dig a bit further on the DAML site, and you’ll find that it’s more specifically sponsored by the “ Information Exploitation Office,” a part of the US military with a mission that includes;

Develop[ment of] sensor and information system technology and systems with application to battle space awareness, targeting, command and control, and the supporting infrastructure required to address land-based threats in a dynamic, closed-loop process.

Furthermore,

Threats of interest include mobile and fixed surface targets in all environments: open, partially obscured, in hide (e.g. under foliage), in evasive maneuver and in urban settings.

Maybe it’s just me, but finding out that the hard work that people are putting into the dream of the next-generation Web is being funded by those who want to use it in combat is disconcerting. Moreover, it sucks that this technology may very well never have its biggest application seen in public.

Wouldn’t it be cool if we could subvert this? Maybe I’ve just been reading too much Michael Moore*, but it seems to me that if Semantic Web technology can be used to tell the military who the bad guys are and what they’re doing, we should be able to use the very same tools — which they’ve so thoughtfully funded — to keep an eye on those who don’t have our interests at heart.

I’m talking about increasing corporate and government accountability.

There’s lots of information out there that the law requires be available, more in public databases, and still more made available by public-minded research projects, think tanks and universities.

Unfortunately, almost all of it is locked up in proprietary or non-interoperable data formats; you can’t easily make a link between who’s on a company’s Board of Directors and what political contributions that company makes, for example. Neither can you correlate between a congressman’s voting record and their campaign contributions, or the campaign promises.

I’m sure that there are a bunch of Web sites that do this kind of analysis and present you with the results (and I’d love to see some of them pop up in the comments, for my own information if nothing else). However, there’s nothing like the raw data, and making it public and easily machine-readable would save a lot of duplication of effort, make it possible to get more reliable numbers, and allow people to verify and explore the data for themselves.

For example, the contents of the national archives, politician’s voting records, election records, state and federal budgets, census data, general statistics, campaign promises, politician’s quotes, details of election funding and campaign contributions, membership of boards of directors, and so forth.

One of the best tools that someone in power can have is obscurity; it seems like RDF and Semantic Web technologies would be perfect for breaking it down. Tim Bray has offered a prize to the coolest RDF killer app that someone comes up with. I don’t have much to add, but maybe the prize of a better country and world to live it would be enough.

• I’ve asked MichaelMoore.com’s Webmaster for an RSS feed; do the same!