Knowledge Graphs for AI Engineers: Designing KG embeddings, SPARQL and Cypher patterns, and integrating graphs (Applied LLM Systems: Production ... Context, and Knowledge Graphs, Band 4) - Softcover

Inhaltsangabe

Knowledge Graphs for AI Engineers is an end-to-end technical guide for building and operating knowledge graphs that power modern AI systems. Starting with ontology and schema design, this book provides concrete blueprints for modeling domain knowledge, representing triples, and handling taxonomies and hierarchical relationships. You’ll find pragmatic guidance on selecting graph stores (Neo4j, JanusGraph, ArangoDB, RDF triple stores), designing ETL and ingestion processes, and implementing entity resolution and canonicalization pipelines.
A central theme is converting graph semantics into numerical representations: learn KG embedding algorithms, vectorization strategies, and how to integrate KG embeddings into hybrid retrieval and RAG flows. The book covers query languages (SPARQL and Cypher), rule-based reasoning, and integrating inference engines with LLMs to create grounded, explainable responses. Operational topics include governance, schema evolution, privacy, security, visualization, and scaling patterns. Case studies illustrate clinical knowledge graphs, enterprise product catalogs, and customer-support knowledge bases.
What’s inside:

• Ontology and schema blueprints for common domains.
• ETL patterns and pipeline templates for reliable KG ingestion.
• Entity linking, canonicalization, and de-duplication best practices.
• KG embeddings, graph neural net overviews, and conversion to vector indexes.
• Query examples in SPARQL and Cypher tuned for retrieval and analytics.
• Integration patterns for combining graph queries with vector search in RAG stacks.
• Reasoning and rule-engine examples for inference and business rules.
• Visualization and analytics workflows for KG insights and reporting.
• Governance: access control, data provenance, compliance, and lifecycle management.
• Scaling, replication, and operational playbooks for production KGs.

• Knowledge engineers, data engineers, ML engineers, and technical product teams building graph-backed AI.
• Enterprises needing explainability, versioning, and governance in their AI pipelines.
• Research teams and tool-builders interested in KG embeddings and hybrid retrieval.

Die Inhaltsangabe kann sich auf eine andere Ausgabe dieses Titels beziehen.