One of my favorite topics for discussion on this blog is the subject of Open Data. In following this interest, I worked with Erik Wilde and Raymond Yee in developing a site to help guide implementation of transparency measures. The site is located at:

The site has demonstrations and an accompanying report (all under a Creative Commons attribution license). We’ve developed a set of simulated data that conforms to the Office of Management and Budget’s (OMB) February 18th specifications for disclosure. These data are offered in a variety of human and machine-readable RESTful web services. We hope that this simulated data will help act as a guide for implementation federal agencies.

We machine-readable XML data, it was pretty simple to do a variety of “mashup”-things:

However, one topic that needs more attention is the issue about what kind of information is required for “transparency”. To help answer this question, we’re seeking feedback from the wider community. Do these data really help in offering a more meaningful level of transparency? What additional information would be required to make this even more useful for community oversight?

Information architectures, services, and machine-readable data are all essential requirements for making data open and encouraging transparency in both research and policy. However, in some ways, these are the easy questions. What’s harder is knowing the specifics about what information is required to make open data actually meaningful for wider communities, whether its for research, instruction, or public oversight of government.

Any feed back and help on these questions would be most welcome!

PS. See Erik Wilde’s blog post for more.