Creating Knowledge Graphs from Documents with Whyis
KGC 2025
•
32m
Jamie McCusker, Rensselaer Polytechnic Institute, Research Director
In this talk, we will explore how the Whyis Knowledge Graph Framework can be used to build knowledge graphs by extracting insights from business documents using in-context learning. Knowledge graphs transform unstructured data into structured, actionable knowledge, and Whyis allows users to track the origins and evolution of data, ensuring its integrity.
We will introduce in-context learning, an AI technique that enables models to extract meaning from text by utilizing existing context, without needing retraining. This method is particularly useful for extracting valuable information from business documents like contracts and reports, allowing organizations to handle vast amounts of unstructured data efficiently.
The session will include demonstrations showing how Whyis integrates with in-context learning to extract key entities and relationships from documents, turning them into structured data that can be visualized and interacted with in a knowledge graph. This combination enhances decision-making, improves data transparency, and supports more effective business analytics.
By the end of the talk, attendees will understand how to use Whyis and in-context learning to build scalable, reliable knowledge graphs from business documents.
Up Next in KGC 2025
-
Driving Business Decisions through On...
Renu Rana
Ontologies are largely in the domain of the Ontologist or Developer, locked up in a modeling tool that few can understand. How can we make the most of the huge investment in developing an ontology? Lets see how it can be used for knowledge sharing, canonical models for data exchange,...
-
Agentic GraphRAG: AI’s Logical Edge
Stephen Chin, Neo4j, VP of Developer Relations
AI models are getting tasked to do increasingly complex and industry specific tasks where different retrieval approaches provide distinct advantages in accuracy, explainability, and cost to execute. GraphRAG retrieval models have become a powerful t...
-
Next Version of RDF, or: Reification ...
Ora Lassila, Amazon Web Services, Principal Technologist
Adrian Gschwend, QLeverize, Zazuko, CEOIn the RDF specification, reification is probably the most disliked but also the most misunderstood feature. There has been a long-running discussion in the RDF community about how to “fix” reificati...
