Andrea Di MaioWith all this discussions on pros and cons of open government data, it occurred to me that I have not seen any reference to how time-sensitive data may be.
The US Open Government Directive recites:
Timely publication of information is an essential component of transparency. Delays should not be viewed as an inevitable and insurmountable consequence of high demand.
and this is pretty much the only place where I have seen a reference to “time” in connection with data.
But what happens when data becomes old, either because there is a new set of data or because it is deemed to be no longer valid as a consequence of one of the many controls that the Directive suggests to put in place?
Not all open data are real-time feeds, and I would argue that most won’t be, as the demand for quality will urge more agencies to filter data and add extra controls.
Two scenarios here.
In the first one government does not regularly update data: this could happen for various reasons, including process delays, changing priorities, new regulations that change data formats. This risk can be reduced with tighter controls and a stronger enforcement of the Directive (although implementation guidelines mentioning timeliness and validity of data would be very welcome).
The second scenario is trickier. Government regularly updates data but data aggregators (“mashers”) copy that data elsewhere and mash up the latter. When government data is updated, its copy may not be, so in both scenarios citizens may be indirectly using obsolete information.
The question is: how do citizens and aggregators know what is the expiry date of the basic data they are using? What should their expectations be about how usable data is?
My impression from looking at this and other aspects of open government data is that it is a great, inspirational idea that may create as many problems as it solves.
What am I missing?
Category: open government data Tags:
3 responses so far ↓
1 Tweets that mention Does Open Government Data Have an Expiry Date? -- Topsy.com December 17, 2009 at 5:19 am
[...] This post was mentioned on Twitter by Andrea DiMaio, Thomas Langkabel. Thomas Langkabel said: RT @AndreaDiMaio: Does Open Government Data Have an Expiry Date? – http://bit.ly/6G7uHV #opengv #gov20 [...]
2 Greg December 17, 2009 at 12:07 pm
As you implied, Scenario 1 is something that can be “fixed” for certain values of “fixed” (ie: incentives to update more often or disincentives for not updating often) but Scenario 2 is just the way things happen when you give people data. Someone will use it (OMG!) and produce something that is inherently temporal (not forever valid) and I think people know how to deal with that; they look at the date it was produced.
You ask if there is an expiry date. I don’t believe there is ever an expiration date on data as long as you have provenance information for that data (when it was created, by whom, from where, etc). It is useful to have the data sets over time so you can analyze trends, for instance.
3 uberVU - social comments December 17, 2009 at 5:41 pm
Social comments and analytics for this post…
This post was mentioned on Twitter by AndreaDiMaio: Does Open Government Data Have an Expiry Date? – http://bit.ly/6G7uHV #opengv #gov20…