GovLoop - Knowledge Network for Government

On Tuesday, I'll be speaking on a panel at the first International Data Summit in Washington, D.C. The topic of our panel is the same as the title for this discussion. Before I pull on my suit and head for the microphone, I'd like to hear what's on your mind.

What do you think?

What are the best resources to read?

Which countries, states or cities are leading the way in open government data? What does that mean?

How do you measure success in this space?

What are the barriers to open government data for agencies, municipalities and states?

Tags: ROI, cyber security, open government, opengov, return on investment

Views: 428

Reply to This

Replies to This Discussion

I'm not sure about these questions, but would love it if you could promote the notion of "mashups". The value-add from my perspective is that we are beginning to break down silos between countries, between Agencies, etc.

So for example, at NASA what could we do if we mashed-up the Google patent data with data about NASA's coolest technology innovations over the years? We generally focus on our own data - Google Moon, Google Mars, WWW Telescope, etc, but what new applications could make the whole process of getting a patent easier and more accessible?

Would also like to get the OpenData community to start to think about high-value data sets and how to promote these. Example: patent data and "the law" are two datasets referenced in the Constitution, and should be readily available in every country.
Thank you, Megan. Mashups will definitely be part of the conversation, particularly since the Data.gov team has been prominently featuring several on the site, and I'm grateful for the specific examples.

WIth respect to patent data, it's been interesting to watch the reaction to Google getting a patent database online. There's a long way to go before Malamud's vision for Law.gov is realized - thank you for that example as well.
May be out of date but here's our list of open data cities/state/countreis - http://data.govloop.com/dataset/List-of-City-wide-Data-sites/agy9-zcn8

Here's a good article on measuring value of open data

Socrata has provided some good though leadership in the area.
Also folks like yourself, Andrew Hoppin, Tim O'Reilly, Vivek, and others

I'd measure success in a few ways:
-Quantiative - # of datasets, # of downloads, # of applications built, et
-Qualitative - stories of how improved internal processes, helped citizens, helped create businesses, etc

I'd look at how Twitter, Yelp, Facebook measures the success of API

I also think Chris Dixon's description of why they have a Hunch API makes a lot of sense - one is that it is just business development 2.0. Makes it a lot easier and saves a lot of time. For years, government has given out data - it just comes in the form of FOIA and takes a lot of time and effort.
http://cdixon.org/2010/08/28/good-bizdev-cannabilizies-itself/
Thank you, Steve! Very helpful. Bill Eggers was looking at that exact list this week. I'm hoping you'll have it updated by Monday; safe bet to expect a surge of interest.
Added Catlunya just now and a couple Friday too.

Let me know if now of any off hand missing...will research more manana too
Some ideas to get the conversation flowing:

# of datasets made available by an agency
# of datasets converted to mash-ups or other useful visualizations
Increase in levels of citizen satisfaction/trust in gov't services (using Foresee Survey or similar(
# of datasets made available that lead to revelations by viewers
# of datasets made available that lead to innovative apps, tools and services by developers
# of datasets that invite citizen participation in the process of correcting and/or enhancing accuracy

I could come up with a few more...but want to see what this sparks!

All Open Gov Plans here:
http://data.govloop.com/Government/List-of-Open-Gov-Plans/x46u-4d2e

Please note also that we are working on a project to take this list into even greater detail...more soon!
Great spark, Andy. Thank you!
There are multiple forces at play here, and they need to be broken down so agencies can more effectively address the "open data" problem. Some agencies have, for a long time, been releasing data. The comparatively utility of those releases is diminished given the explosion of analytical capabilities available to the average citizen. Agencies are good stewards of outdated data release practices. Check out things like TaxStats (http://www.irs.gov/taxstats/) or the National Bridge Inventory (http://www.fhwa.dot.gov/bridge/nbi/ascii.cfm). There's a wealth of information waiting to be unlocked in those data sets - but we're releasing them the same way we have been for years - without any sort of acknowledgment of the changes in technology.

For agencies to make their data useful, they're going to have to invest in a "both-and" approach. Some people rely on their regular releases (economists and industry analysts go by information released by the Bureau of Labor Statistics and the Bureau of Transportation Statistics). If we want to empower people, we have to keep those practices alive while nurturing emerging data analysis practices. Then there's what people could learn from some of the more gigantic data sets out there, like the National Household Travel Survey (http://nhts.ornl.gov/download.shtml).

For these reasons, I don't believe in quantitative metrics like how many data sets get released. Agencies that are in the business of releasing data need to baseline their current use before they release them to data.gov or hold some competition around their data. But if their data release practices are less-than-useful to the app developer community, their data is going to get copied to a place and converted into something an app developer can use. We can't measure those things.
I think this is a good point that is not mentioned enough.

Goverment is HUGE with tons of different agencies. A lot of government agencies have tons of experience with making data public for years - think places like Census data which are very good at liasoning with academics/researchers/companies to give out useful Census data in key formats. Also places like Bureau of Labor Statistics with labor data.

There are lots of lessons from agencies there. Specifically, these agencies have spent lots of time and effort figuring out what people want from their data, in what format, and in what frequency. Also, people are usually looking for specific data not data generally.
Very thoughtful take - thank you.
Its great to see all the creative work at http://data.gov including the intuitive navigation for reviewing data sets and data visualization examples. The Value Proposition of Open Government Data should include building a stronger community around this space. Stories about why and how people use Open Government Data may have more impact than measuring number of views or downloads. The meaning of the data and reasons for empty or cryptic fields could be crowdsourced to the community. A small group of community managers could focus on keeping the community healthy.

Adding value to Open Government Data can be done with Application Programming Interfaces (APIs) that extend the value of these data sets. Twitter and Facebook experienced explosive growth after creating and releasing their APIs. These APIs could be used to create "sharable objects" that can be shared, embeded, & mashed up across multiple environments and devices. We can start with simple standard formats before we get into things like SPARQL and RDF. The World Wide Web is already packed with server-side and client-side tools to read and manipulate RSS. We also see data sets in RSS formats at http://data.gov This is one simple way of putting government data into standard sharable formats. Many of the other data sets could be transformed from their raw xml format to standard RSS formats.

What if Open Government Data was available in YQL?
Taking the data to the people seems like a good goal. The Yahoo! Query Language is an expressive SQL-like language that lets you query, filter, and join data across Web services. http://developer.yahoo.com/yql/

What if Data.gov Had a Real App Store?
A Real App Store for Open Government Data that supported the collaborative spirit of adding value to Government Data would be great hosted in a space like http://data.gov This Platform as a Service (PaaS) would support serving apps, mashups, and Software as a Service (SaaS) across the federated space. I believe we would see exponential growth in the value of government data with something like JackBe's Presto powering the Government App Store.
Thank you for the links and analysis.

RSS

© 2014   Created by GovLoop.

Badges  |  Report an Issue  |  Terms of Service