Skip to content
  • Sign In
  • Try Now
View all events
Large Language Models (LLMs)

Knowledge Graphs & Large Language Models Bootcamp

Published by O'Reilly Media, Inc.

Intermediate content levelIntermediate

Give meaning to your data and power your data science applications

Course outcomes:

  • Understand what a knowledge graph is and how it can benefit your data science applications when paired with Large Language Models (LLMs).
  • Learn how to set up a knowledge graph development project
  • Understand the project steps, techniques, and technologies that can support them, and the factors that are critical for success
  • Learn how to use knowledge graphs to make LLMs more accurate and reliable

Join expert Panos Alexopoulos in an engaging course that explores knowledge graphs and their partnership with Large Language Models (LLMs). You’ll learn the main stages of knowledge graph development, including kickstarting the project, developing the schema, populating the graph, and controlling its quality. You’ll understand the skills necessary for building real-world knowledge graphs, such as acquiring requirements from stakeholders, designing schemas in Neo4j, discovering entities and relations from text with NLP tools, and detecting quality issues. You will learn how LLMs can effectively help you in building your knowledge graph but also how your knowledge graph can enhance the accuracy of LLMs and elevate your data science capabilities. The first week focuses on knowledge graph basics, including how to create and manipulate data in Neo4j, while the second week’s focus is populating and controlling the quality of the knowledge graph, with practical exercises in mining entities and relations, detecting knowledge graph quality problems, and using knowledge graphs to eliminate LLM hallucinations and improve their accuracy, reliability and explainability.

What you’ll learn and how you can apply it

  • Decide whether a knowledge graph is a proper solution for your data challenges, and specify its desired characteristics
  • Design a knowledge graph’s schema in a way that makes the rest of its development much easier
  • Populate a knowledge graph in an automatic way
  • Implement mechanisms to assess and improve the quality of a knowledge graph
  • Use LLMs to enhance all the steps of developing a knowledge graph
  • Use knowledge graphs to ground LLMs and make them more accurate and reliable

This live event is for you because...

  • You’re an aspiring or practicing data scientist or data engineer.
  • You work with unstructured and messy data and want to learn how to develop knowledge graphs to give this data structure and meaning.
  • You work with LLMs and want to learn how to eliminate their hallucinations and improve their accuracy, reliability, and explainability

Prerequisites

  • Familiarity with Python and the Jupyter Notebook
  • Familiarity with entity-relationship diagrams
  • Familiarity with basic NLP and machine learning concepts

Recommended preparation:

Recommended follow-up:

Schedule

The time frames are only estimates and may vary according to how the class is progressing.

Week 1

Understanding knowledge graphs and their relation with LLMs (60 minutes)

  • Presentation: Knowledge graphs and why would you want to build one; stages of the knowledge graph development lifecycle
  • Presentation: Large Language Models (LLMs) and their interplay with knowledge graphs
  • Q&A
  • Break

Understanding and working with Neo4j(45 minutes)

  • Presentation: Representing and storing data in Neo4j according to the property graph model; manipulating and querying data with Cypher
  • Jupyter notebook: Create a Neo4j sandbox and showcase basic data creation, manipulation, and access queries (
  • Q&A
  • Break

Developing a knowledge graph schema in Neo4j (75 minutes)

  • Presentation: Elements of a knowledge graph and their representation in Neo4j; acquiring schema requirements for a given domain and mapping to graph elements
  • Presentation: Using LLMs in schema design: what works and what doesn’t
  • Jupyter notebook: Create a knowledge graph schema in Neo4j about films and TV shows
  • Q&A

Week 2

Populating the knowledge graph (60 minutes)

  • Presentation: Tasks involved in populating a knowledge graph; knowledge sources and approaches
  • Jupyter notebook: Automatically mine and add entities and relations to the knowledge graph with Large Language Models
  • Q&A
  • Break

Controlling knowledge graph quality (60 minutes)

  • Presentation: Knowledge graph quality dimensions, metrics, and trade-offs; typical quality problems and debugging methods
  • Jupyter notebook: Detect and correct quality problems in the knowledge graph using Large Language Models
  • Q&A
  • Break

Grounding LLMs with knowledge graphs (60 minutes)

  • Presentation: How knowledge graphs can be combined with LLMs to eliminate hallucinations and improve accuracy, reliability and explainability
  • Jupyter notebook:
  • Q&A

Your Instructor

  • Panagiotis Alexopoulos

    Panos Alexopoulos has been working since 2006 at the intersection of data, semantics, and software, contributing to building intelligent systems that deliver value to business and society. Born and raised in Athens, Greece, he currently works as Head of Ontology at Textkernel, in Amsterdam, Netherlands, where he leads a team of Data Professionals in developing and delivering a large cross-lingual Knowledge Graph in the HR and Recruitment domain.

    Panos holds a PhD in Knowledge Engineering and Management from National Technical University of Athens, and has published more than 60 papers at international conferences, journals and books. He is the author of the book "Semantic Modeling for Data - Avoiding Pitfalls and Breaking Dilemmas" (O'Reilly, 2020), and a regular speaker and trainer in both academic and industry venues.

    linkedinXlinksearch