Thursday, February 17, 2011

After Jeopardy what's next for IBM's Watson? Bring about a Healthcare Revolution as "Dr. Watson"?

Not as far-fetched as it sounds. After pummeling previous Jeopardy Super Stars Ken Jennings and Brad Rutter into submission over the last few nights on Jeopardy (with Ken signing off his final answer with " “I, for one, welcome our new computer overlords”), IBM's Watson may take on a new Avatar as a Physician's Assistant (from MobiHealthNews)::

“For IBM, the future will happen very quickly, company executives” told the Times. “[Today] it plans to announce that it will collaborate with Columbia University and the University of Maryland to create a physician’s assistant service that will allow doctors to query a cybernetic assistant. The company also plans to work with Nuance Communications Inc. to add voice recognition to the physician’s assistant, possibly making the service available in as little as 18 months.”
Nuance, of course, offers the very popular Dragon voice recognition software for healthcare providers and others. Imagining a voice-enabled “physician assistant service” that taps into Watson and available as a smartphone app is not at all difficult. A desktop version of the service would be substantially less useful.
“I have been in medical education for 40 years and we’re still a very memory-based curriculum,” Dr. Herbert Chase, a professor of clinical medicine at Columbia University who is working with IBM on the physician’s assistant told the New York Times. “The power of Watson-like tools will cause us to reconsider what it is we want students to do.”
For those who are not Jeopardy Buffs, Watson is a new super computer from IBM that prove itself more than capable of besting and replacing today’s top Jeopardy! players. In a television spectacle reminiscent of the super computer Deep Blue’s win over chess world champion Gary Kasparov in 1997, a room-sized computer developed by IBM managed to beat out a pair of Jeopardy! all-stars over the course of a three night game this week. Watson, ended the run with $77,147 compared to Ken Jennings’ $24,000 and Brad Rutter’s $21,600.

The match included a medical related question — no, a diagnosis question really: “You just need a nap. You don’t have this sleep disorder that can make sufferers nod off while standing up.” Watson beat out the humans with the answer: “What is narcolepsy?” Maybe you don’t need an M.D. for that one, but still the computer got there first.

Wednesday, February 16, 2011

The Business Intelligence Chronicles Part 20: The Battlelines are drawn: Traditional Enterprise BI Platforms vs. Data Discovery Platforms

Gartner's recently released Magic Quadrant for Business Intelligence Platforms highlights a trend which all BI practitioners have observed over the past few years : the divide between Traditional enterprise BI platforms and Data Discovery platforms and the resultant confusion amongst decision makers about which one to opt for.

Each one has its own strengths and weaknesses as Gartner has very nicely captured in this comparison:



Gartner goes on to say 
"The chasm between these segments has deepened because business users find the benefits of using data discovery tools so compelling that they make this choice despite the risk of creating fragmented silos of data, definitions and tools. This has accentuated the need for IT organizations to back away from a single-minded pursuit of standardization on one vendor to a more pragmatic portfolio approach."
Gartner's magic quadrant as it stands today still shows the Traditional enterprise BI Platforms (IBM, Oracle, Microsoft, Microstrategy etc.) in the Leaders quadrant while the Data Discovery Platforms with the exception of Qliktech are in the Challengers (Tableau, Tibco) or in the Niche Players category (Jaspersoft et al)



However, that will be in a state of flux as Traditional BI vendors  add Data Discovery to their portfolio (Microsoft with PowerPivot, SAP with SAP BusinessObjects Explorer, IBM with IBM Cognos Express and Information Builders with WebFocus Visual Discovery ) as they see BI spend dollars slipping to Data Discovery vendors with their ability to model, navigate and visualize data.

Meanwhile, Data discovery tool vendors (like Qliktech) are implementing capabilities to improve their enterprise readiness.
Gartner views "capabilities for integrating departmental and enterprise data models" as one of 10 important capabilities for Quadrant placement.
 
Surprisingly there is no player in the "Visionary" category. So it seems that we are stuck with these two alternate approaches (Traditional Platform and Data Discovery) and their hybrids for the near term horizon. I was hoping that time was ripe for a player to emerge with a Google/Amazon/Facebook mash up kind of approach to BI of the kind I described in an earlier post: The Business Intelligence Chronicles Part 8: Star Trek returns....  :
 
I would focus on the direction it should be headed. And the single word answer to that is "simplification". The current and evolving paradigm of user-technology interface can be defined by a single word : "Google". Never before has the "Keep it Simple, Stupid" paradigm been so successful.
And that is the direction which BI needs to head to also.
A simple, single line search box which delivers context specific search and reporting capabilities. A sales rep types in the name of a city and links to sales report(s) for all customers in that city appear.
Also an Amazon.com kind of feature -" People who viewed this report also viewed...". Gives the user a quick and easy feel of what else s/he should be looking for. Building in "crowd intellect" into the reporting capabilities.
And what about a feature which allows a user to know who else is looking at a particular report in real-time or has viewed the report earlier (a viewing history). Combine this with interactive chat or communication capabilities and you have transformed a "report" into a "collaborative interaction".
Did I miss Geospatial analytics and GPS, how about integrating them with BI capabilities. A manager viewing a report, identifies issues with a particular customer, is able to identify the sales rep closest to that location based on the GPS capabilities of the handheld the rep is carrying and asks him/her to visit the customer.
BI has already ventured beyond Reporting & Analytics to Actionable Intelligence. Now it needs to venture even further to "Intelligent Actions". The gap,distance between the availability of intelligence and resultant action needs to be reduced by incorporating other emerging technologies into the classical BI Framework
A (non) player which I have been bullish about has been Google. I wrote in 2009 ( The Business Intelligence Chronicles Part 12 :BI - "Head in the clouds but feet firmly planted on the ground") :
 

Staying on Google, I think that it may not be long before they enter the BI arena. They know how to manage huge volumes of data, run server farms and manage an analytics front end , currently primarily used to present information related to blogs/website statistics but can easily be modified as a front-end for business related information.
They have confounded me by not showing their hand yet.

Maybe IBM may have something up its sleeve with "Watson". They'll have to put it to work after it is done clobbering Ken and Brad on Jeopardy. BI would be a good space to put is skills to use. I had speculated :
There was a time when we had BI Tools, then vendors started branding them as Applications. These days several vendors are offering BI Appliances. Following the trend to it's logical conclusion, I would think that very soon we may see BI Machines or BI Engines.

Is "Watson" going to be the BI Machine/BI Engine of the future? "Elementary, My dear Watson",  I hear as I stroke my Sherlockian chin.




Search Google

Google

Site Meter