I broadly agree. Technology is the easy part. I would think in terms
of "architecture of coordination". Let me comment on one related
aspect.
"Finally, there is the question of what incentive there is for
third-party developers to build such an API?"
Surprisingly enough, it doesnt have to be very high. The process has to be
efficient. If it requires too much of mental and process context switch, it
gets difficult. People like me are not in school anymore.
-Venkata
On Mon, Jul 1, 2013 at 9:50 AM, Gora Mohanty <[email protected]> wrote:
> On 1 July 2013 09:28, Nagarajan M <[email protected]> wrote:
> > Hi All,
> >
> > The Open Data movement is evolving rapidly with the publication of
> datasets
> > on www.data.gov.in
> >
> > The problem with government data is that its more statistics than
> actionable
> > data. However the open data movement will evolve to demand for more
> granular
> > and near real time data.
> >
> > The Government system will not be able to provide it because the systems
> of
> > data collection and management are not standardized. There are not enough
> > tools that are built to enable the Government machinery to integrate data
> > collection through work flow.
>
> Actually, I think that it will *have* to be done through the
> Government of India, and the people running data.gov.in
> It is just too difficult for third-party developers to do this
> in a consistent manner that will keep working in the future:
> Please see below.
>
> > I am interested in creating a Data Repository that can act as an API
> where
> > different workflows of government can link up and operate. Use existing
> > metadata and also add their own.
> >
> > I request your suggestions with regard to architecture, technologies and
> > tools.
> [...]
>
> Architecture, tools, etc., are not that complicated. One
> could start with a system that parses the current spreadsheets,
> digests the data, and provides an API. However, there are
> several issues with third parties doing this:
> 1. While several people have done specific case studies by
> manually extracting the data from the spreadsheet files,
> this is obviously not something that can scale.
> 2. It is possible to build a system that enlists the help of people
> interested in specific datasets. For example, a prototype of
> such a system would parse the first few lines of any given
> spreadsheet file, make an educated guess about data types,
> and allow the user to modify the guesses.
> 3. Such crowd-sourcing might work at one level, but a big
> problem is that:
> (a) There is no consistent standard followed by the spreadsheets.
> E.g., at least names of states, and date formats could
> be standardised.
> (b) Several of the spreadsheets that we looked at have
> internally inconsistent data. Please see the issues
> brought up by Supreet at
>
> http://www.mail-archive.com/[email protected]/msg29943.html
> 4. Finally, there is the question of what incentive there is for
> third-party developers to build such an API?
> For sustainability, this either has to be done by a publicly-
> funded institution or there has to be a commercial basis for it.
> If it is done through public funds, why should data.gov.in not
> handle it themselves?
> There might well be a business case in selling access to such
> an API, but the cost of developing the API and hosting the
> services will be quite high. However, I cannot seem to find any
> indication on the data.gov.in sites as to whether such third-
> party, commercial access is allowed. In fact, I could not find
> any information on the terms of usage of the data published
> there.
>
> Regards,
> Gora
>
> --
> For more details about this list
> http://datameet.org/discussions/
> ---
> You received this message because you are subscribed to the Google Groups
> "datameet" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
>
--
For more details about this list
http://datameet.org/discussions/
---
You received this message because you are subscribed to the Google Groups
"datameet" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.
