Knowledge Graphs for LLMs: Harnessing Structured Intelligence for Smarter AI Systems: Graph Construction, Retrieval-Augmented Generation, and Semantic ... 2 (AI Projects and Workflow Systems Series) - Softcover

Miller, Zhao

Softcover

ISBN 13: 9798264455261

Verlag: Independently published, 2025

Inhaltsangabe

A professional, applied manual for designing, building, and operationalizing knowledge graphs that materially improve LLM-driven systems. This book provides the end-to-end technical roadmap data ingestion, entity/relation extraction, schema design, storage, querying, and LLM integration required to create explainable, high-precision hybrid AI systems.
What’s inside

• Fundamentals of graph modeling, schema & ontology design, and graph theory essentials.
• Practical pipelines for extracting structured facts from unstructured text using NLP and embeddings.
• Integration patterns for Neo4j/RDF/graph stores, vector databases, and RAG architectures.
• Querying and analytics: SPARQL, Cypher, and hybrid retrieval approaches.
• Performance optimization, versioning, governance, and visualization techniques.
• Domain case studies (healthcare, finance, enterprise search) demonstrating measurable ROI.

Key topics;
knowledge graphs, graph databases, ontology design, entity extraction, SPARQL, Cypher, RAG, embeddings, semantic search, graph-augmented LLMs, information retrieval, data governance.
Who should read this
Data engineers, knowledge engineers, ML/AI practitioners, and technical product managers tasked with building authoritative retrieval systems or explainable AI features. A working knowledge of databases and basic NLP is helpful.
Deliverables & format

• Reproducible projects that convert raw text into production-ready graph assets.
• Query recipes, integration blueprints, and operational guidelines for graph maintenance and scaling.

Details zur Ausgabe

• Verlag: Independently published
• Erscheinungsdatum: 2025
• Sprache: Englisch
• ISBN 13: 9798264455261
• Einband: Tapa blanda
• Anzahl der Seiten: 194
• Kontakt zum Hersteller: Nicht verfügbar
• Verantwortliche Person: Nicht verfügbar