// By Jane Weber Brubaker //
IBM’s cognitive computer Watson became a household name and overnight sensation in 2011 when it beat two of Jeopardy’s top contestants, smoothly handling the show’s lineup of cleverly named categories and obscure facts:
“Dialing for Dialects for $800,” says Watson, choosing a category in the opening round.
Host Alex Trebek reads the clue: “While Maltese borrows many words from Italian, it developed from a dialect of this Semitic language.”
Watson beats its opponents to the buzzer. “What is Arabic?” asks Watson.
“You are right,” says Trebek.
“$1,000, same category,” Watson requests, and continues answering questions correctly, ultimately winning the $1 million challenge.
Watson was specifically developed and trained to compete on Jeopardy, as a high-profile test of human vs. machine thinking and cognitive technology. To compete, Watson ingested vast amounts of information from a wide range of sources, as well as thousands of Jeopardy answers and their corresponding correct questions. It then had to learn the nuances of clues that use puns and idioms, recognizable by humans, but complex for computers. The technology has the capacity to learn and improve its accuracy if it has enough examples. In one phase of training, featured in a PBS Nova documentary, it took Watson several examples of correct answers in one Jeopardy category to understand that the answer had to be the name of a month.
Physicians Take Note
Soon after Watson’s victory, IBM began to get inquiries from physicians, particularly oncologists, interested in knowing if cognitive technology could be applied in a clinical context. These discussions spawned IBM Watson’s earliest partnership, with Memorial Sloan Kettering Cancer Center. The organization refers to itself as “Watson’s oncology teacher.” As new information is added, Watson for Oncology learns and improves its accuracy.
Suzanne Sawyer is Vice President, Portfolio Marketing at IBM Watson Health. “An oncologist can query Watson to say, ‘What are the most appropriate treatment options and clinical trials for this cancer and this patient?” Sawyer says. “Watson brings back the knowledge of MSK oncologists and a vast repository of clinical research that’s been done around those types of cancers, or that patient profile.” The results are evidence-based recommendations, ranked and scored for relevance, and available in seconds for a physician’s consideration.
IBM Watson Health partners with other leading healthcare organizations, including Cleveland Clinic, Mayo Clinic, MD Anderson, Baylor College of Medicine, CVS, Apple, Medtronic, Welltok, Johnson & Johnson, the American Heart Association, the American Cancer Society, and the American Diabetes Association.
Only members can access this content.
So what does cognitive computing have to do with healthcare marketing? The same cognitive technologies that help oncologists zero in on the most promising treatments at an individual level may help healthcare marketers aggregate all their data for patients and prospects, and use it to make the best decisions about who to market to, how to market to them, and even how to plan their media buying. From the perspective of population health, Watson is enabling clinicians to help patients manage chronic conditions, a key concern for all health systems. Continue reading to learn how a medical device manufacturer is using Watson to improve patient care, how one innovative startup helps marketers streamline analytics and use data from disparate sources, and how IBM is using Watson to make it’s online media buying more efficient. Please log in.
Not a member yet?
Sign up for a trial membership and get instant access.
Renew your membership now and get instant access.