A naming standard for Open Government Data

I just watched the Gov 2.0 Expo video (May 27, 2010 in London, UK) featuring Tim Berners-Lee advocating for government data standards.This is my response to Rhiza Labs CEO Josh Knauer's presentation: On open government data, Tim Berners-Lee is almost right.

I am an advocate for data standardization and transparency as a working data quality policy practitioner. I want to express my agreement with your comments.

Tim Berners-Lee is the founder of HTML (Hyper-text Markup Language) and plausibly the entire World Wide Web, yet his premise that linking to URLs is the most effective way to build a framework for data naming consistency reduces to glibness.

URLs are not persistent

Links often break in 10 years, and are far too vulnerable. URLs should be persistent, but they often aren't, not in the real world. I worked as a data governance manager for two managed care programs, one federal government, the other Medicare-funded at the state level. You are correct: uniform identifiers are a superior approach, and UUID’s (universal uniform identifiers, that are not language-specific) would be best.

In the context of my healthcare-related work, I would add the following data classifications used by CMS (Center for Medicare and Medicaid Services, Dept. of Health and Human Services) to the other U.S. government-related data standards you cited:

• HCPCS, ICD-9 and CPT medical codes (maintained by Ingenix);
• National Drug Codes (from the FDA) for pharmaceuticals;
• taxonomy of 11-byte alphanumeric codes for medical specialties

These codes are usable with OR without electronic health records, which isn't possible with Tim Berners-Lee's advocacy of URL-based identifiers.

Tim Berner-Lee’s appeal for standardization, and ability to hold the attention of an audience about data policy, is certainly helpful. (Who is excited about the subject, other than those who do it, and those who suffer from the lack of it?) I hope that the spirit of his message is what is acted upon, rather than being used as a starting point for implementation.