Knowledge Graph

On this page I wanted to talk about some of my main work in knowledge graphs and discovery. Here are some videos and articles which introduce the work done by my team over the past few years:

Watson Discovery Service

The Knowledge Graph technology which I originally architected for Watson Discovery Advisor, was recently ported to IBM's cloud on top of IBM Watson Discovery Service. Find more information here.

https://console.bluemix.net/docs/services/discovery/building-kg.html#watson-discovery-knowledge-graph

Watson Discovery Advisor

The below features IBM Watson Discovery Advisor, the software I helped envision and design and also eventually develop to a mature technology

Cancer Research and Drug Discovery

An article which describes the work my team was engaged in in the early part of watson in 2013, and 2014 with Baylor Medical College, on finding protein kinases which are relevant in the pathways of a very critical gene in cancer called the TP53 or P53.

An article in IEEE spectrum which talks about some work done by IBM Research in Yorktown, which was leveraging API's from knowledge graph built by my team. (see the picture of the knowledge graph there)

Cyber Security

A video featuring colleague Jeb Linton, who built IBM Cyber security on top of Watson Discovery Advisor and Knowledge Graph, stuff literally built by my team. You also see screen grabs of software which we have made.

Watson Discovery Services on the cloud

This is the product whose GA release my team was behind. This is a lighter weight version of discovery (compared to watson discovery advisor) for the more commodity discovery use cases.

Stuff I can't talk about

While the above is public domain work, one of the most important application of the knowledge graph technology has been in domains and clients I cannot talk about because of confidentiality and the sensitive nature of the analysis we are involved in. The goal is to continue to try to apply the same work on public domain and publish it in academia. Two of my recent papers which are taking some of our work public are below:

Papers and Presentations:

Panel of Knowledge Graph leaders from IBM, Facebook, Google, Microsoft and Ebay
Talk at Stanford Biomedical Knowledge Discovery Lab - Discovery with Knowledge Graphs
Separating Wheat From the Chaff -- A Relationship Ranking Algorithm. Sumit Bhatia and Alok Goel and Elizabeth Bowen and Anshu Jain. The 13th Extended Conference on Semantic Web (ESWC), 2016.
Context Sensitive Entity Linking of Search Queries For Enterprise Knowledge Graphs.Sumit Bhatia and Anshu Jain. The 13th Extended Conference on Semantic Web (ESWC), 2016.
Tutorial on Enterprise Knowledge Graphs for Large Scale Analytics Nidhi Rajshree, Sumit Bhatia, Nitish Aggarwal and Anshu Jain IEEE Big Data Conference, 2017
Tools and Infrastructure for Supporting Enterprise Knowledge Graphs -- Sumit Bhatia, Nidhi Rajshree, Anshu Jain and Nitish Aggarwal Advanced Data Mining and Applications, 2017