[LINK] Big Data - bad or otherwise

Bernard Robertson-Dunn brd at iimetro.com.au
Wed Jan 15 13:55:26 AEDT 2014


On 15/01/2014 12:27 PM, Janet Hawtin wrote:
> On 15 January 2014 07:37, Jan Whitaker <jwhit at janwhitaker.com> wrote:
>
> This issue of Cutter's newsletter has a couple articles on this subject:
>> http://www.cutter.com/research.html
> If big data is generated/collected on behalf money because that's who can
> afford it then the way that data will be phrased, framed, focused will suit
> those interests
>
> If newspapers hire big data journalists to mine big data for stories to me
> that seems like a risk they could be 'embedded' in those interests. Would
> they then be able to ask people face to face to confirm themes in data? How
> can you be objective about non objective data or data which
> critiques/defines citizens but can not see back up the telescope to
> entities which have capacity to define the information playing field.
> So some entities are opaque. How does that contrast shape things?

Big data is mined for two reasons

1. To find out what happened. This covers Janet's point but government's 
and others may also be interested in history/the past.

This is relatively benign.

However, bad conclusions can be drawn (see True Fact: The Lack of 
Pirates Is Causing Global Warming,
http://www.forbes.com/sites/erikaandersen/2012/03/23/true-fact-the-lack-of-pirates-is-causing-global-warming/ 
)

As long as this "information" is not acted on, no harm is done.

Which leads to the second reason:
2. To predict outcomes and make decisions.

This is where my previous observations on models really comes into play. 
As with the pirate data it is easy to see correlations and coincidences, 
it is much harder to identify cause and effect. This requires an 
explanation, usually (always?) based upon verifiable models.

IMHO, apart from the privacy issues, this is one of the greatest dangers 
of analysing Big Data - people will see things in the data that just 
aren't there.

-- 

Regards
brd

Bernard Robertson-Dunn
Sydney Australia
email: brd at iimetro.com.au
web:   www.drbrd.com
web:   www.problemsfirst.com
Blog:  www.problemsfirst.com/blog




More information about the Link mailing list