Earlier this month Gartner named semantic technologies to its top ten trends list (see our story here). Recently, we caught up with Gartner vp and distinguished analyst Debra Logan, the lead author on the semantic technologies section of the Top 10 Technology Trends Impacting Information Infrastructure, 2013, to learn more about sem tech’s earning a place on the list.
One interesting point Logan made is that the top ten trends list actually is a reflection of inquiries Gartner sees from its end-user clients. So, semantic technologies’ spot on the list would seem to indicate a bubbling-up of real-world, enterprise interest. As Logan sees it, it’s very much about information overload, about minimizing the risk and maximizing the value of the data on their hands, and about the availability now from providers like Amazon and Google of infrastructures for analyzing Big Data sets.
“If we could get the same meaning from data, we might actually know what is going on, because we sure don’t now,” says Logan, of the quandary facing enterprise IT leaders. “They are struggling with definition issues and reconciliation because of the proliferation of different IT systems.”
When some of those end users inquire about how semantic tech may have a role in solving dilemmas, they’re not always thinking of W3C standards or triple-store databases per se. They may be thinking more about artificial intelligence applications when it comes to cleaning up a large corpus of unstructured information, for instance.
“They are really thinking of classification, based on rules. …They want to sort out the mess and to do that with classification-type technologies that work like rules-based expert systems,” she says. “Expert type systems are about to come back into their own, where you encode rules based on what people know of a domain.”
Data migration activities also are encouraging thinking about classification and categorization. “A lot of clients want to move data on premise from one archiving system to another, and in the process they want to select and further classify and categorize it,” she says. “And, because there is so much more usage of these kinds of colloquial document management systems like Sharepoint, where users go in and go nuts, people want to understand taxonomies and how they are created, and how to automatically assign things to taxonomies. Having that happen with some level of accuracy is something else people are interested in doing.”
The verticals to which semantic tech has mostly been important – financial, health care, oil and gas, and pharma – remain the biggest thinkers about hard-core semantic technology. But other verticals are biting off a bit more here and there. For instance, Logan covers e-discovery which is taking the legal sector along on the semantic journey. Machine learning and its use of predictive coding has a role in sorting out documents that should be reviewed by a lawyer or that are or aren’t responsive to a given meaning tag.
“So the legal industry, that is a Luddite to some extent for technology, is being forced down the path because of the amount of data,” she says. Even judges are starting to talk the predictive coding technology talk, Logan notes. “It’s one of the professions being forced fastest down the road.” Certainly that’s true here in the U.S. but she sees it about to gain more traction in the U.K., too, where she’s currently based, in the wake of judicial reforms set to occur in April.
“They say we are not as litigious and we’re not, but there are the same underlying drivers,” says Logan. “There’s so much more data and it has use as evidence and also in a regulatory context.”