Subscribe to DSC Newsletter

LAS VEGAS—IBM has the capability to compete with Google or Microsoft's Bing in the search arena, but the company has no interest in moving in that direction despite the strong showing IBM made in answering a variety of complex questions with itsWatson supercomputer in a "Jeopardy!" quiz show competition.

At IBM's Pulse 2011 systems and servicemanagement show here, Steve Mills, IBM's senior vice president and group executive of software and systems, told a still very Watson-hungry press corps that although Watson demonstrated that it could have a much better success rate at getting to a specifically correct answer to a complex query than a Google search, IBM has other plans for the technology and its natural language processing and DeepQA (Question Answering) technology.

"We learned a lot about what it would take to make it [Watson] commercial," Mills said. "We've capitalized on many, many decades of accumulated computer science."

And Big Blue has some distinct ideas for where to take the Watson technology, most likely toward creating "mini Watsons or baby Watsons," Mills said. IBM announced a partnership with Nuance Communications to incorporate Watson technology into health care applications. But beyond that, IBM has made no commitments, certainly not in search.

"Google's business is about displacing what had been direct mail," Mills said. "IBM is out to make money on its technology. There is a business model behind Google for doing what they do. They're moving money that would have otherwise been spent to them. There's a big expense there, and we're not interested in competing with Google in the business they're in. We're interested in helping businesses get to the bottom of problems and answer questions they run into to improve the quality of their products and services, and enhance loyalty and increase revenue."

With the Google model, companies are paying for relevancy and coming up first, Mills said. "There's always been money spent on that, and Google's moved in to capitalize on that. And people have asked us, 'Why can't IBM bring its technology to bear on that—especially where people are not thrilled with the accuracy [of Google searches]?'"

However, "We know how to do what they do, but it doesn't make sense to spend a massive amount of money to do some sort of super Web crawling system. You've been watching what's going on with Bing," Mills quipped to eWEEK.

Mills said IBM did not build Watson in the Google model. "We built it to come back with THE answer or a relatively few answers and then you apply your judgment on top of that."

In that regard, Mills said he believes Watson will do well in an assistant's role in various industries such as health care, law and call centers.

"I can certainly see this technology as a physician's assistant, not a replacement for the physician. We're trying to come up with practical applications where we can add support for finding the right answers. How much better would the humans have been [on Jeopardy!] if they had their own Watson in their pocket?" Mills asked.

IBM officials said the DeepQA project at IBM shapes a grand challenge in computer science that aims to illustrate how the wide and growing accessibility of natural language content and the integration and advancement of natural language processing, information retrieval, machine learning, knowledge representation and reasoning, and massively parallel computation can drive open-domain automatic question-answering technology to a point where it consistently rivals the best human performance. 


Views: 12

On Data Science Central

© 2020   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service