Go to content

Sherlock, a Watson Cognitive Computing Use Case by Sandhya Kapoor, James Weaver and Daniel De Luca

You know when a problem has you stuck, you Google’d it, checked Stack Overflow and still can’t find what you need? Earlier this year Voxxed connected with IBM, asking if Watson could help find technical answers differently. We wanted to look at Cognitive Search, hook it up to a library of videos and articles we’ve published over the years, and try to make something better than what is already around. Recently we started hacking and want to share what we did - hopefully some of you will help take it further. We’ll walk you through the project so far, using Watson Speech to text and Concept Insights APIs to build a cognitive engine to find technical videos and articles in a corpus we’ve created. And we’ll ask whether Concept Insights helped us make better decisions when analyzing quantities of unstructured data. We’ve only scratched the surface. More tech, sources and ideas need to be plugged in, and we’ll iterate between Devoxx conferences in different locations. Come and hear what we could do, what we’ve managed so far, and how to get your hands on the code. Some of the Watson API’s we’ve been playing with are at: https://console.ng.bluemix.net/docs/services/watson.html [RMC-9947]

November 7, 2016