Large Language Models (LLMs) are powerful but notoriously ungrounded as they generate fluent, plausible text that isn’t always factual, consistent, or explainable. One promising way to address these issues is by connecting LLMs with Knowledge Graphs, namely structured, explicit representations of knowledge that can provide context, constraints, and verifiable facts. The main paradigm for doing this is Graph-based Retrieval-Augmented Generation (GraphRAG), which integrates graph-based reasoning and retrieval into the generation process. In this talk, I’ll introduce the core ideas behind GraphRAG, describe common design patterns, and outline the steps and tools needed to implement such systems in practice.

Graph-based Retrieval-Augmented Generation (GraphRAG) enhances large language models (LLMs) by grounding their responses in structured knowledge graphs, offering more accurate, domain-specific, and explainable outputs. However, many of the graphs used in these pipelines are automatically generated or loosely assembled, and often lack the semantic structure, consistency, and clarity required for reliable grounding. The result is misleading retrieval, vague or incomplete answers, and hallucinations that are difficult to trace or fix.

This hands-on tutorial introduces a practical approach to evaluating and improving knowledge graph quality in GraphRAG applications. We’ll explore common failure patterns, walk through real-world examples, and share a reusable checklist of features that make a graph “AI-ready.” Participants will learn methods for identifying gaps, inconsistencies, and modeling issues that prevent knowledge graphs from effectively supporting LLMs, and apply simple fixes to improve grounding and retrieval performance in their own projects.

Exploration of RDF to LPG transformations and maintaining semantic meaning in knowledge graphs.

What value does semantic data modeling offer? As an information architect or data science professional, let’s say you have an abundance of the right data and the technology to extract business gold—but you still fail. The reason? Bad data semantics. In this practical and comprehensive field guide, author Panos Alexopoulos takes you on an eye-opening journey through semantic data modeling as applied in the real world. You’ll learn how to master this craft to increase the usability and value of your data and applications. You’ll also explore the pitfalls to avoid and dilemmas to overcome for building high-quality and valuable semantic representations of data. Understand the fundamental concepts, phenomena, and processes related to semantic data modeling Examine the quirks and challenges of semantic data modeling and learn how to effectively leverage the available frameworks and tools Avoid mistakes and bad practices that can undermine your efforts to create good data models Learn about model development dilemmas, including representation, expressiveness and content, development, and governance Organize and execute semantic data initiatives in your organization, tackling technical, strategic, and organizational challenges