What BI, Cloud Computing, and IBM's Watson Have in Common
Watson may give a boost to business intelligence solution developers.
By David S. Linthicum, SVP, Cloud Technology Partners
There's news that IBM's Watson will be offered as a cloud service, and it looks like there are solid BI benefits that can be had from this technology.
Big Blue announced that its supercomputer will be accessible by developers this year. The idea is that developers will be able to invoke the Watson API to drive complex strings of questions, and it will return the answers in real time. This means that those looking for cognitive analytics capabilities, linked with big data analytics capabilities, can leverage this technology.
For years, I've been a huge advocate of the marriage of BI and artificial intelligence-driven systems, considering that we can actually place learning systems next to data. However, the complexity that comes with this kind of integration, as well as the latency and computing power required, meant that there were a few instances of BI and AI working together, but they took a back seat to traditional approaches.
Although IBM Watson is nothing new, the integration with big data and cloud computing means that it's lean and scalable. IBM's Watson was designed to provide an answer to questions, not the meaning of patterns of data. The difference is that Watson's natural language processing capabilities, learning systems, and cognitive processing abstracts the complexity of the data for those who really need to understand the data.
For example, it's not a simple dashboard or report that comes out of a BI system, it's actually questions such as: How has the new product line launch this year done relative to the key economic indicators? and How much fraud has likely occurred within orders placed this year?
This removes the BI analyst from having to assemble a report or graph for an executive. Moreover, it provides more agility and flexibility in leveraging the system. It increases the value of keeping and maintaining large amounts of data. Finally, it's just way cool.
ITBusinessEdge's Loraine Lawson notes that "As IBM points out, it has already used this computing approach to improve patient care and operational efficiency for health care." The technology is not easy to duplicate, considering that IBM has been refining Watson for 14 years, and the system relies on 40 different technologies. However, I've worked on less sophisticated prototypes in the past, and Watson just brings together some core technology to make something similar.
IBM offers a Watson Content Store, which is really just purchased access to key data, such as industry data, economics data, etc. This means that you can bind external data sources with your own data sources to provide better insights into your data.
One example of Watson content is Healthline, a provider of intelligent health information, data, and technology solutions. The company provides a comprehensive, contextually relevant health reference library to enable the promotion of healthy lifestyles, support disease prevention, and offer clinically significant, medically reviewed health information.
Again, I don't view Watson as anything new. We've been working on this problem for years, but not as hard as other aspects of BI. IBM's investment in this technology will pay off, considering that we're clearly looking for new and innovative approaches to BI, and the use of cognitive systems was really the next logical step.
The ability to access Watson-based services using services and APIs is a smart move, as is the integration with big data systems. The ability to leverage pre-integrated and tested information makes this technology more compelling for enterprises seeking critical intelligence to drive their business.
David S. Linthicum is a big data and cloud computing expert and consultant. He is the author or co-author of 13 books on computing, including Enterprise Application Integration (Addison Wesley). You can contact the author at www.davidlinthicum.com.