In 2011, IBM's Watson made headlines when it competed on and won Jeopardy!. Afterwards, IBM researchers put the technology to work, creating WatsonPaths, a cognitive system that uses natural language processing and machine learning to solve medical problems.
My challenge was to design an interface that helps medical students at Cleveland Clinic learn how to make better decisions—and as WatsonPaths interacts with users, it also learns and becomes more accurate by incorporating their responses into its immense repository of medical knowledge.
At the core of WatsonPaths is a diagram that visualizes the system's decision making process. In it, users can see how Watson creates multiple levels of inferences based on an initial problem statement, ending with the answers it has the most confidence in on the right.