ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] So you want to be a Data Scientist?

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Edmon Begoli <ebegoli@xxxxxxxxx>
Date: Fri, 28 Dec 2012 18:49:52 -0500
Message-id: <CAGj+Ysf9pfh3=o8ufaxEK=2RMjqivxRDUZ2B=D3dFJV+WkZgHw@xxxxxxxxxxxxxx>
This is an excellent discussion. 
As someone who is by nature of my work [1],[2] 
very close to the "Big Data" movement, and my personal research interests 
are deeply within KR field (I might also be just a KR cheerleader :-) ) let me offer few hopeful observations.

First, while I consider "Big Data" term mostly a buzzword and a brand with huge sales potential
there are some very curious developments that might help bring familiar fields into a mainstream.

For the first time since I have been active in IT industry (~18 years) I am witnessing introduction of sub-fields that 
are more sophisticated and challenging than your typical IT application development
- we are seeing unfulfillable demand for mathematical (statistics) and artificial intelligence (machine learning)
methods that are the center of the "Data Science" hype.

Contemporary problems and availability of sophisticated technology (Hadoop, ...) 
are starting to demand application of sophisticated methods.
There is someone willing to pay for these and wanting to profit from the findings.

I am absolutely hopeful and confident that this trend of sophistication of demands and methods through 
IT will continue. I think that "Big Data" is blazing a trail for mainstreaming of Computer Science and AI. 

Second, "Big Data" today is dumb. It is a pile of sand, mound of dirt, a hay stack. We are producing 
mounds of this raw material without putting any organized effort to make it into a 
more of a prefabricated material that one can easily assemble into new sources of information or even knowledge. 
This is where I see opportunity for businesses, if not a human kind, to make our lives easier and business 
more profitable.

Finally and as John Sowa observed, if we are to make this field "popular" and valuable to the broader, non-academic
audiences we should try to focus on "killer apps" and to ride this wave by introducing creative KR 
solutions. I am relative beginner in the field of KR and ontologies, and yet I observe innumerable 
opportunities how to bring decades of research into relatively under-informed mainstream.  


On Fri, Dec 28, 2012 at 6:03 PM, John F Sowa <sowa@xxxxxxxxxxx> wrote:
John B and Kingsley,

As I have said many times, my primary goal since the 1970s has been
to bring AI technology into the mainstream.  That was the theme of
my first publication about conceptual graphs in 1976:

    http://www.jfsowa.com/pubs/cg1976.pdf
    Conceptual graphs for a database interface

My greatest frustration is that the AI community keeps shooting
themselves in the foot every time they have a chance to crawl out
of their niche.  In 2000, I thought that the Semantic Web was
finally going to make a breakthrough into the mainstream.

Tim B-L's proposal of 2000 was very promising, but the academic
community found a way to destroy that opportunity.

JB
> Their [MSFT's] projections for 2020 is about 1/2 $T for Internet embedded
> systems.  The database system for MS Embedded 8 products  is MS SQL.
>
> There was no mention of taxonomies or ontologies.

KI
> I agree with your core insights and key message. That said, I would like
> to add that "Big Data" and "Data Scientists" are buzzwords *primarily*
> conjured up by marketeers who lack profound understanding of the fundamental
> challenges that have dogged data since the inception of modern computing.

I agree that their web pages have a high percentage of hype.  But it
is about the same percentage as the hype generated by the SemWebbers.

The main difference is that they Big Data people know how to talk to
IT professionals and their managers.  Meanwhile, the SW hype machine
focuses on academic pursuits that have no connection with reality.
Their biggest blunder was to make decidability their primary slogan.

KI
> I think we might as well add the moniker "Data Artist" to the mix.

That term will not attract the people who have money to spend, and
without funds, you can't build industrial-strength systems.

John

_________________________________________________________________
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



_________________________________________________________________
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    (01)

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