[Top] [All Lists]

Re: [ontolog-forum] FW: Looking to the Future of Data Science - NYTimes.

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: John F Sowa <sowa@xxxxxxxxxxx>
Date: Sun, 31 Aug 2014 12:18:59 -0400
Message-id: <54034AF3.1040509@xxxxxxxxxxx>
This thread has mixed the technical question about defining 'big' with
a huge number of social and political issues.  I agree that all these
issues are important, but it's also important to distinguish them.    (01)

For the purely technical issues, the adjective 'big' can be defined
in a technology independent way as "too big" for polynomial-time
algorithms to process the data with existing technology -- except
at enormous expense.    (02)

> What's prohibitively expensive depends on one's budget, business
> case, business model, etc.    (03)

Yes, but polynomials quickly outrun anything that grows linearly,
such as budgets and server farms.    (04)

John B
> To me Big Data is a successful marketing term with very little
> published rationale.    (05)

I agree.  That's why I prefer to deflate the hype bubble by shifting
the focus to something less sexy, such as (N log N) algorithms.    (06)

John B
> Many systems can deal with large data by indexing dynamically across time.    (07)

That's a good strategy.  But the total time is the same whether you do
the computations piecemeal or in batches.    (08)

Michael B
> In the context of personally identifiable information, Big Data
> is sometimes called "Soylent Green" [1]. It can be seen as the
> opposite of the basic principle of German privacy law...    (09)

The requirements for privacy and security were not anticipated
by the designers of the "personal" computer or the designers
of a system for sharing research papers among academics.  When
you design a sieve, it's hard to plug the gaps.    (010)

Mainframe computers are more secure because they were designed
for bankers.  But when the banks connect to a sieve, they leak.
Today, the worst bank robbers never leave home.    (011)

John    (012)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (013)

<Prev in Thread] Current Thread [Next in Thread>