Blogs

The “one thing”: IBM Watson Analytics

Post Comment
Data Scientist and Professor of Astrophysics & Computational Science, George Mason University

After day two at IBM Insight 2014 in Las Vegas, I am reflecting on the plethora of big data and analytics capabilities, innovations and opportunities that filled the Mandalay Bay Convention Center (which, by the way, is nearly a mile, all indoors, from one end to the other, so there was a lot to see and experience). Following such a wealth of demos and presentations, what does come to mind after day two is the “one thing” (to borrow the phrase from the movie “City Slickers”) that seems to matter the most, that will make the biggest impact and that has captured the essence of big data and analytics technologies for the future, rapidly approaching world of data everywhere, sensors everywhere and the Internet of Things. I believe that “one thing” is IBM Watson Analytics.

One-thing-Blog.jpg

IBM Watson was the first full implementation of a cognitive analytics platform, which still has no comparable competitors. But the lack of major competition has not led to complacency. In fact, IBM has made remarkably deep and wide-ranging investments in Watson Foundations, including a far-reaching vision of where and how it will impact the world: including medicine, finance, cooking, customer service, advertising, retail, security and everywhere else.

So, what exactly is cognitive analytics?

In the simplest terms, cognitive analytics is learning the right question to ask within a given context from the data at hand. As in the game of Jeopardy (which IBM Watson won convincingly against the very best of the best human competition), the goal of cognitive analytics is to look at the presented information (the “answer,” which is embedded in the data stream) and the information’s category (the data’s context) to come up with the correct question to ask. This we see demonstrated in Watson. By accessing, integrating, mining and contextualizing the full set of available data (the 360 view), Watson Analytics becomes your personal discovery advisor, guiding you to ask the right questions of your data, at the right time, at the right place, for the right individual, in the right context.

Watson use cases

Big data and analytics are changing the world: Chef Watson learns from thousands (probably tens of thousands) of recipes about the different types of dishes, styles of cooking and ingredients along with the right mixture of ingredients, the taboo combinations of mixtures to avoid (including user-selected prohibitions, perhaps associated with food allergies), the preparation steps (the actual recipe) and so much more. Simply select the style and type of dish, apply user-selected filters and then Watson comes up with suggestions in the form of a ranked list of 100 new recipes for your culinary exploration. I would love to see the underlying data science scoring algorithm for that ranking—I'm sure that’s a PhD dissertation research topic all by itself!

So Chef Watson could be described as the “new culinary data science,” or perhaps “recipe analytics” or just plain “good tasty fun with science.” Later this year the Watson team is promising to release IBM’s first cookbook—stay tuned for something delicious. 

Chef Watson was just one of many use cases presented at IBM Insight 2014. Other applications included Watson for wealth management and Watson for oncological medicine. After this launch of Chef Watson, food science will never be the same! 

Descriptors of IBM Watson Analytics

  1. Watson Analytics is a personalized data concierge to decision making in this age of big data and information overload
  2. The vast and rich user-created data stream that Watson Analytics taps into (for the 360 view of a single user) includes social, mobile, reviews, web analytics (“bread crumbs”) and digital exhaust
  3. Watson is a cognitive insights advisor (as was powerfully demonstrated by Terry Jones in his WayBlazer demo [ to access the demo, first register for InsightGO, and then go directly to the session recording)
  4. With Watson, every citizen will become a scientist (I love that!)
  5. The big data that Watson currently has access to “ain’t nothing” compared to the coming data deluge from the Internet of Things
  6. Watson will transform the Internet of Things into the Internet of Connected Things, which then becomes the Internet of Insights! 

Watson for everyone

IBM Watson is not just for IBM developers, either. There is an end-user ecosystem already in operation. The Watson Developer Cloud allows anyone to build an API into Watson using IBM Bluemix. I didn’t know much about Bluemix before this conference, but I now know some very important facts about it. Specifically, Bluemix provides the essential tools for building Internet of Things apps: Watson Foundations, DataWorks, Cloudant NoSQL, dashDB, push notifications, embeddable reporting, geospatial analytics, analytics for Hadoop, Time Series Database and the Internet of Things Foundation. Several of these foundational tools were just announced this week!

Regarding the Internet of Things, Jake Porway followed the Watson Analytics presentations at IBM Insight 2014 with the following prediction: a few years from now, when we look back at 2014, we will all declare “Big data—how quaint!” That is definitely the “one thing” that we all can agree upon!

Check out my other IBM Insight posts on my blog: Rocket Data Science