Rules and Reasoning with Knowledge Graphs: Specifying Inference Rules for Knowledge Graphs
  Irene Polikoff   Irene Polikoff
  Ralph Hodgson   Ralph Hodgson


Tuesday, October 15, 2019
01:45 PM - 05:00 PM

Level:  Intermediate

In this talk, we will discuss how knowledge graphs can support reasoning (also known as inference), how to describe information required by reasoning in a form of rules, and how to manage the inferred facts across the data lifecycle. This capability is important for many applications such regulatory compliance. For example, if we captured information about our data sources in a knowledge graph and knew that one of the sources stored data about EU residents, we could infer that the data source and a company that owns it are subject to GDPR regulations.

Topics we will cover in this talk are:

  • Syntax for defining rules that will support inferences
  • Differences between rules that check for data quality/validity and rules that infer new facts
  • Testing and debugging rules
  • Static (batch) inferencing vs. dynamic, query time inferencing
  • Performance considerations for using inferencing
  • Uncertainty and inferencing 

Irene Polikoff, the CEO and co-founder of TopQuadrant, was a co-chair of the W3C SHACL Working Group and is now a co-chair of the W3C SHACL Community Group. She has more than two decades of experience in software development, management, consulting, and strategic planning. Since co-founding TopQuadrant in 2001, Irene has been involved in more than a dozen projects in government and commercial sectors. She has written strategy papers, trained customers on the use of the Semantic Web standards, developed ontology models, designed solution architectures, and defined deployment processes and guidance. Before co-founding TopQuadrant, Irene was a Principal in the national Knowledge, Content Management, and Portals Practice at IBM Global Services. Prior to that, she was a Senior Development Manager and a Project Executive for IBM Worldwide Consultant's tools and methods. Irene has a master's degree in Operations Research from Columbia University.

Ralph Hodgson is a co-founder and the CTO of TopQuadrant, Inc., a U.S.-based company that specializes in semantic technology solutions, tools, consulting, and training, focusing on data governance and lineage. He combines expertise in semantic technologies with over 30 years of experience in business application development and deployment, consulting, software development, and strategic planning. Prior to starting TopQuadrant in 2001, he held executive consulting positions at IBM Global Services where he was a founding member of Portal Practice and Object Technology Practice. Prior to IBM, he was European Technology Director, founder, and Managing Director of Interactive Development Environments, which was an international CASE tools vendor. Ralph is a published author and a frequent speaker at conferences. He has co-authored the books "Adaptive Information," published by John Wiley in 2004, and "Capability Cases: A Solution Envisioning Approach," published by Addison-Wesley in July 2005.