BigData - and mashups

Semantically BigData has moved past its roots.  Doug Laney's original definition - Volume, Velocity, and Variety - still holds in the purist sense, but nowadays BigData tends to be used as a grab-bag for all sorts of stuff, including NoSQL, Analytics, Business Intelligence, and, well, pretty much anything involving manipulating and parsing data.  Come to think of it, BigData doesn't even have to be Big, it just has to incorporate the methodologies associated with any/all of the above.  In short, BigData refers as much to methodology as it does to the three Vs.

   On that front, Factual's recent consolidation of their APIs is very much a BigData announcement.  For those who aren't in the know, Factual provides access to datasets for use by 3rd party apps.  

Datasets? Huh?

Well, the classic example is their "Places" datasets, which contain information about "locations" around the world. Think pretty much everything in the Yellow Pages (when those used to exist) for most countries in the world, each with lat-long, categories (e.g. Legal & Financial > Accounting & Bookkeeping), addresses, phone numbers, etc.  They also have health-related databases (including the ever-classic Homeopathic Remedies), Beer, Sports, and a ton more.

What, you may ask, does this have to do with BigData?  Well, it *is* BigData to begin with (d-uh. Its data, and its big!), but much more importantly, they recently consolidated their Places API into a new Global Places API.  Now, you have just one API and one mechanism to use to access information about any/all of their Places regardless of country.

Which is remarkably neat, because it is exactly this type of seamless access (mashups?) across different data sources that actually provides value in the new world of Analytics/Intelligence.  Datafiniti has already started going down this road, providing 'simple' (albeit quite different) access to a remarkable variety of data stores.  There will eventually be a shakeout in this market, but in the meantime I expect to see quite the plethora of innovative applications cross-referencing multiple data stores in an end-user friendly fashion.


Update:  I didn't forget Infochimp, its just that Datafiniti and Factual are so radically different in their access methodologies...