With the whole of IBM now marching to its cognitive computing message, the company’s Insight conference in Las Vegas was always going to be heavily influenced by analytics, Watson Analytics in particular. After all, it is the combination of advanced analytics – including text analytics and machine learning – in that product that is likely to lead the cognitive computing charge for the company. Having announced its decision to launch a cognitive computing consulting practice earlier in October, IBM is now pushing cognitive computing the way it pursued smarter cities, Linux and other strategic initiatives in the past. It also announced a new cloud-based analytics service.
With a portfolio as large as IBM’s, the company has to choose to focus on certain products over others each year. At this year’s conference, the focus was more on analytics than on data or content management, at least as far as we could tell. And the star of the show was Watson – specifically, Watson Analytics and the company’s cognitive computing strategic direction. After all, without analytics – and a specific type of analytics – cognitive computing isn’t possible. As Mike Rhodin, senior vice president IBM Watson put it, where IBM has previously scaled computing, the company is now helping businesses scale knowledge.
In its seventh annual IBM Institute for Business Value analytics global study, IBM found that out of a survey of 1,226 people, 71% said they were using analytics in three or more functional areas of their business, whereas in 2014, that percentage was only 10%, which would seem to indicate some sort rapidly accelerating momentum. And in other more tangible signs of momentum, IBM has managed to attract more than 500,000 registered users of Watson Analytics in less than a year, working at – we understand – more than 5,000 companies. Obviously, a relatively small percentage of those are paying customers at this point, but that is still a rapid adoption rate by any measure.
A key part of the technology stack driving the new Insight Cloud Services – as well as Watson Analytics – is text analytics, specifically entity extraction and recognition, where IBM reckons it is ahead of the game. Entity extraction – identifying people, places and things in unstructured text – is not new; we’ve been covering it as part of the text analytics market for more than a decade. IBM has built and bought its text analytics stack, including key acquisitions such as Systems Research & Development (SRD) way back in 2005, which gave it identity resolution software and the expertise of SRD founder Jeff Jonas, who now is IBM’s chief scientist of context computing. IBM has technology that can disambiguate entities, not just identity them, so it can tell that this person called John is this specific John and not another one; it’s been using it in anti-fraud technology and services for years.
There may be some ‘Watson washing’ going on here, whereby IBM-built and acquired products are getting a Watson makeover, but Watson Analytics is largely new and is blazing a trail for self-service prescriptive and predictive cloud-based analytics.