Semantic Search and Information Retrieval using GenAI

Course details

Generative AI is a win for traditional search technologies, especially when you use its power in tandem. This technical course designed for data professionals shows you how to design, implement, and evaluate sophisticated search solutions that combine knowledge graphs, AI, and traditional search technologies. Learn how to make architectural decisions, implement security measures, and establish effective KPIs for enterprise-scale search systems. Along the way, instructor Ashleigh Faith emphasizes practical implementation, security considerations, and performance optimization while providing hands-on experience with modern search architectures.

Skills covered

Neo4jOrchestrating AI SystemsSearch Engine Marketing (SEM)Search Engine Optimization (SEO)AI for Data Engineers and ScientistsRole-Based AI ApplicationsAI Development Tools and PlatformsData VisualizationBuilding with AIData AnalysisMarketingData ScienceBusiness Analysis and StrategyBusiness Software and ToolsOpen SourceOne-Off

Concepts

Introduction

The power of semantic search and information retrieval
What you need to know for the course

Selecting Search

Use case - Two Trees Olive Oil
What kind of search do you need
Deciding if you need a vector database, a knowledge graph, or both
What about non-LLM AI for semantic search

Building Blocks of Semantic Search

What is information retrieval
What is semantic search
What are knowledge graphs, and how are they used in semantic search
What are vector embeddings

Enhancing Semantic Search

What is semantic similarity
How is semantic similarity calculated
Weights, boosts, and blocks
User and localization data
Walking the graph
Other factors for semantic similarity
Search performance in semantic search
Automated and human search relevancy assessment

Optimizing Data and Entities for Semantic Search

Text and document quality
Entity recognition and disambiguation
Alternative search methods

Mastering Semantic Search with AI, Human Feedback, and Key Data

Where does search fit in an AI or data pipeline
How can AI use both vector and graph data for semantic search
How can key performance indicators (KPIs) be used in semantic search

Bridging Text and Graphs for Smarter Semantic Search

Loading our sample model and data into Stardog
Building a graph query
Query options
Using graph data to ground an AI response
Comparison GraphRAG grounding alone vs. RAG with semantic knowledge graphs

Conclusion

Continuing your learning journey