Topic hub
Product graphs & semantics
Tables store products. Relationships are what a distribution business actually runs on: which record is the same part, what replaces what, what is only compatible, which document proves a value. A product graph stores those relationships as typed, sourced, scored data so software and AI agents can traverse and trust them.
This hub groups the concepts in the order they build on each other: where the graph sits in your stack, how it is modelled, how relationships are defined, and how evidence keeps every claim defensible.
Start here
The model and where it sits.
-
Guides
Product Identity & Evidence Graph: Products Aren't Rows
Products aren't rows. They are entities connected to suppliers, variants, documents, attributes and evidence. How Claro resolves fragmented records into a graph your ERP, PIM and AI agents can trust.
-
Glossary
What Is a Semantic Product Layer?
A semantic product layer connects ERP, PIM and PLM references to resolved product entities, with source, confidence and relationships machines can reason over. Definition and how it differs from a PIM.
-
Articles
Your Catalog Is Already a Graph. You Store It in Tables.
A spreadsheet shows rows; your business sees relationships. Why matching, classification, document matching and supplier onboarding are all edges, and what you gain by storing them that way.
Modelling meaning
Ontology, taxonomy and the relationships that matter.
-
Glossary
What Is a Product Ontology?
A taxonomy tells you where a product belongs; a product ontology tells machines what it is and how it relates to other things. A practical definition for industrial catalogs, without the academic baggage.
-
Guides
Product Relationship Types: SAME_AS, EQUIVALENT_TO, COMPATIBLE_WITH
SAME_AS, VARIANT_OF, EQUIVALENT_TO, COMPATIBLE_WITH, REPLACED_BY and the rest: what each product relationship means, where the lines are, and why the distinctions have real consequences in distribution.
-
Articles
Catalog, Taxonomy, and Attribute Schema Are Not the Same Thing
A precise mental model for separating product records, categories, schemas, source documents, enrichment jobs, and review workflows.
-
Glossary
Product Knowledge Graph: What It Is and Why It Powers AI Search
A product knowledge graph connects products, attributes, and relationships so machines can reason across a catalog — not just store flat rows.
-
Glossary
What Is a Product Identity Graph?
A product identity graph turns one-off record matching into persistent identity knowledge: a canonical product linked to SKUs, MPNs, suppliers and documents with scored edges, including NOT_SAME_AS, that survive re-ingestion.
Evidence and provenance
Where a value came from, and how sure we are.
-
Glossary
What Is Data Provenance?
Data provenance is the recorded origin and history of every product attribute. Learn why source lineage is essential for trusted catalogs, AI enrichment, and write-back.
-
Articles
Product Data Provenance: Why the Source Matters More Than Another Extracted Attribute
Extracting product attributes is becoming easy. Trusting them is not. Learn how provenance turns values into auditable product records.
-
Articles
From Product Record to Product Memory
A record says what is true now. Product memory says why, since when and on what evidence. The case for canonical records that keep history, sources and decisions, and what that unlocks.
-
Glossary
Confidence Score in Data Matching: A Practical Guide
Learn what a confidence score in data matching is, how thresholds work, and why calibration determines whether automated matching is trustworthy.
Build it
Practical walkthroughs.
-
Guides
How to Build a Product Evidence Graph from PDFs, ERP and Supplier Files
A practical workflow for connecting products, variants, supplier files, attributes and evidence into one reviewable product record.
-
Guides
Why Product Data Belongs in a Knowledge Graph, Not a Document Folder
Folders store product files. Knowledge graphs connect products, variants, suppliers, evidence and decisions. Learn why the relationship matters.
-
Articles
Product Matching Should Produce Knowledge, Not a Score
A 94% match leaves nothing behind for the next ingestion. A SAME_AS relationship with evidence, a confidence score and a reviewer compounds. Why matching output should be durable, and near-misses kept.
Related topic hubs
Claro
Stop maintaining this by hand
Claro keeps product and supplier data trusted as catalogs change — matching, deduplication, enrichment, and validated write-back into the systems you already run.
Book a demo