Watson User Experience - overview
The amount of information available to professionals is exploding—in the medical domain, it's doubling about every five years, and most is available in natural language. How can a decision support tool provide access to contextually relevant information in a given workflow? The Social Computing Group took on the challenge of imagining how users might engage with IBM Watson, the natural language processing system that successfully competed on Jeopardy! in 2011. In partnership with the Watson team and Columbia Medical School, we designed and developed a series of demos and functional prototypes for Watson's human interface in both general and healthcare-specific applications. HCI challenges included the formation of accurate and relevant questions, the provision of easily accessible and relevant information for users, and the adaptation of the question-answer technology to users' workflows. Healthcare demos have been featured in Technology Review and the Associated Press.