I am proposing a social bookmarking mechanism that allows tagging statistical data and building a data search engine. This is the main idea:
"By simply putting a structured tag, here called “Simple Statistical Signature” on web pages that hold statistical data, we can considerably reduce the cost of searching statistical data on African countries ,while providing a fully decentralized mechanism for the collaboration of statistical communities to assess data quality, and even build a reputation mechanism for data producers covering both estimated and forecasted data . This article proposes such a decentralized social bookmarking mechanism for statistical data that can be a very useful collaboration tool for data communities in the context of the Data Revolution."
Here is a link to the full proposal:
Any thoughts about it?
Thanks in advance,
Web pages aren't static anymore, so a tag does very little to identify information.
Secondly, statistical data is so very experiment-specific that there really needs to be a rich meta-data layer to capture what the data is and what it is not.
I would think a better architecture for trying to solve your requirement is to build a data API and try to socialize that API by helping folks publish their data through that API. This is the tack that Socrata and other OpenData efforts take.
Fundamentally, though, the meta-data layer is the problem, and there are a hundred different attempts in play right now with no clear winners.