Sign In

Communications of the ACM

ACM TechNews

Watson and the Future of Cognitive Computing

View as: Print Mobile App Share:
The logo of the Watson supercomputer

IBM Research-Australia is implementing cognitive computing on the Watson supercomputer.

Credit: IBM

In an interview, IBM Research-Australia director Glenn Wightwick discussed the technologies behind the Watson supercomputer and what differentiates cognitive computing from past approaches to machine learning and natural-language processing.

Wightwick says that much of the technology behind cognitive computing has existed for many years, but many applications now being explored involve processing a tremendous volume of data extremely rapidly to enable cognitive systems to engage with people in a natural way.

"What is dramatically different about how we are approaching cognitive computing based on the Watson technology is we approached natural language as a stochastic problem," he says. "Watson learns through a combination of training via machine learning, adapting for features of the language that are new to a particular domain, and ingesting all the information it can find on the domain."

Cognitive computers such as Watson have the potential to transform a wide range of industries, including healthcare, finance, education, law, government services, and commerce. Wightwick says the next cognitive computing advances for IBM will include expanding recall, learning, judgment, reasoning, and inference. He says cognitive computers will recognize emotions, increase expressiveness in generating speech, and add perception and creativity.

From Computerworld Australia
View Full Article


Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account