Imagine walking into a dark, abandoned library, armed only with a flashlight, searching for an answer somewhere in the building. With the narrow beam of the flashlight, you can read the titles on the spines of the books, but little else. Now imagine you find a light switch and the old card catalog. Suddenly you have an index of every book, searchable by subject, author or title. Now you can find the most obscure piece of information, no matter where it is.
In many ways, standard search methodologies, such as Google and other text-based search engines, are like that flashlight — a narrow band of results from a single perspective. In fact, the basic function of a search engine has remained the same since the early 1980s — a user enters a search term, which returns links to pages that include the word or words found in that search string. But there are now search technologies that surpass the limitations of keyword searching.
So how do you ditch the narrow search results of traditional keywords and move to a world where you have all the information you need at your disposal? We at the Gordon Flesch Company believe that the answer to this shortfall is cognitive computing.
Why do we believe in cognitive computing and natural language processing? Sometimes, being able to find a document you need is not enough. What is more useful and more powerful is to be able to find actionable information. A business doesn’t just need to find records like invoices or orders from past years. You need to be able to find insights in those records, like what factors might be driving orders or new business. Those kinds of insights can make a business more efficient, more profitable, and more nimble than the competition.
Consider IBM Watson, one of the most mature and advanced cognitive systems. Watson is a system that is not simply programmed; it is trained to learn based on interactions and outcomes. It’s a cognitive system that rivals a human’s ability to answer questions posed in natural language with speed, accuracy and confidence. It’s the most successful example of a system to navigate the complexities of human language and to analyze massive amounts of data exceptionally quickly (over 200 million pages in three seconds when it beat Ken Jennings on Jeopardy!).
Unlike a search engine, Watson’s cognitive system can be trained to consider the particular demands of your inquiry and will learn over time what results are most relevant for your organization. For example, if you are researching a legal issue, Watson understands that your search results should not just feature the most relevant link, but only information that is relevant for your jurisdiction, the most recent precedent, and considers pending regulatory changes.
Of course, it has taken time and years of research for cognitive systems to reach current levels of maturity. In fact, the first time Watson appeared on Jeopardy, the machine lost to a human. Back in 2010, Watson answered the Final Jeopardy! category “U.S. Cities,” with the answer, “What is Toronto?”
Not only did Watson fail to get the right answer, but it failed to recognize that it was looking in the wrong country. Given the same question today, Watson would not make the same mistake.
In attacking the problem of the ambiguity of human language, computer science is now closing in on what researchers refer to as the “Paris Hilton problem.” That is, the ability, for example, to determine whether a query is being made by someone who is trying to reserve a hotel in France or is simply to pass time surfing the internet looking up has-been celebutants.
The next time Watson appeared on the show, the machine was able to parse language better than an English professor. It answered, “A recent best seller by Muriel Barbery is called ‘This of the Hedgehog,’” with “What is Elegance?”
It even showed its facility with medical diagnosis. With the answer: “You just need a nap. You don’t have this sleep disorder that can make sufferers nod off while standing up,” Watson replied, “What is narcolepsy?”
The coup de grâce came with the answer, “William Wilkenson’s ‘An Account of the Principalities of Wallachia and Moldavia’ inspired this author’s most famous novel.” His human opponent, Ken Jennings got the right answer as well (Bram Stoker) but realized that he could not catch up with Watson’s winnings and wrote out his surrender, adding “I, for one, welcome our new computer overlords.”
With the use of Natural Language Processing (NLP), a user can simply ask a question in plain English and receive answers and insights, not just a document that happens to include a particular keyword. Ideally, cognitive systems don’t require significant intervention by business intelligence experts with a Ph.D. to continually maintain and update a knowledge base. It is artificial intelligence that works for businesses of any size or sophistication.
Mike Adams is manager of development with GFConsulting for the Gordon Flesch Company, where he identifies, develops, and deploys intelligent content management solutions for customers. To learn more, visit Gordon Flesch Company here.