Having problems with message search?
Thanks for sharing your thoughts on this.
There no doubt that there is demand in the market for "something" that will help them extract knowledge from all their text repositories and social networks. But the question is, are our technologies in a state where it is indispensable to the end users or is it just a "good to have" luxury offering. I feel we are currently positioned as the later in the market and hence my reference to the business outlook being grim in the short term. To make the shift from a luxury product to a necessity product, we will have to solve customer painpoints and that will inturn generate the success stories.
Tell me tangible business problems that have been solved on a consistent basis with text analytics. We still are to reach there. I'm sure we will one day and hats off to all the existing vendors for taking us in that direction. But at this point, its naive to say that text analytics is an all powerful main stream technology. We are still dabbling with early adopters at this stage. And to prematurely assume that all our insights such as sentiments, entity extraction etc. that we generate from terabytes of data will increase our customer's sales by X% or reduce cost by Y% and have definite business impact will only blind us from the real problems in the market.
My $.02
Ravi
On Wed, Feb 25, 2009 at 1:08 AM, Andrew Eichenbaum <andreweichenbaum@...> wrote:
Ravi et al.,
I don't normally respond to posts, but I got a completely different view from the articles. Let me quote two of your comments:
the business outlook for text analytics seems grim...we still are yet to have a success story that can be replicated
As a person who has worked analytics in multiple realms I have to remind you that when most analytical systems are built, they are built with one purpose in mind. That inherently limits the analytical system to one realm/domain, or one set of specifications, or one focus. To build a general (do it all) engine is much harder to build and verify.
I look at the companies looking to cash in on sentiment or popularity of social networks/other online groups as one benchmark, and I see nothing but growth in the requests for "analyst" year over year. This is both in the start-up and big business world.
It does seem that where ever there is a data overload, there is not much insights to be extracted through text analytics...
So you are saying the business side does not understand what their multi-terabyte DB is telling them since they are still limited to Excel. Weclome to one of the most important parts of your job, e.g. relaying your insights to the people who could use them. Business is best done as a dialog, not just requirements being fulfilled.
Just my 2 cents,
Andrew
From: Ravi Shankar <ravi@...>
To: TextAnalytics@yahoogroups.com
Sent: Tuesday, February 24, 2009 11:10:08 AM
Subject: Re: [TextAnalytics] A couple of articles
Seth,
Very insightful article. But putting the pieces together from the article, the business outlook for text analytics seems grim.
As one of them mentioned, we still are yet to have a success story that can be replicated. That indeed is a real issue.
It does seem that where ever there is a data overload, there is not much insights to be extracted through text analytics in a scalable and consistent fashion that will have business impact.And whereever there is a significant business impact possible such as marketing surveys text analytics in its current state seems to be an overkill according to the market.
Thanks
RaviOn Tue, Feb 24, 2009 at 11:17 PM, Seth Grimes <grimes@altaplana. com> wrote:
Perspectives on Text Analytics in 2009http://www.intellig ententerprise. com/blog/ archives/ 2009/02/infonic_ reloade.html
http://www.b- eye-network. com/channels/ 1394/view/ 9720/
Infonic Reloaded, or the Liberation of Lexalytics
--
Seth Grimes Alta Plana Corp, analytical computing & data management
Intelligent Enterprise magazine (CMP), Contributing Editor
grimes@altaplana. com http://altaplana. com +1 301-270-0795
Offline 
Send Email
