{
  "$schema": "https://km4ai.org/data/schemas/glossary.schema.json",
  "meta": {
    "title": "KM4AI Glossary",
    "description": "Vendor-neutral glossary of Knowledge Management and AI terms used across km4ai.org.",
    "version": "0.1.0",
    "updated": "2026-07-17",
    "license": "https://creativecommons.org/licenses/by/4.0/",
    "license_name": "CC BY 4.0",
    "attribution": "KM4AI (https://km4ai.org/)",
    "canonical": "https://km4ai.org/data/glossary.json",
    "formats": [
      "https://km4ai.org/data/glossary.json",
      "https://km4ai.org/data/glossary.csv"
    ]
  },
  "terms": [
    {
      "id": "rag",
      "term": "RAG",
      "slug": "rag",
      "expanded": "Retrieval-Augmented Generation",
      "definition": "An architecture that retrieves relevant documents or knowledge chunks before an LLM generates an answer, grounding outputs in organizational sources rather than model memory alone.",
      "category": "ai-architecture",
      "related": [
        "chunking",
        "embedding",
        "grounding",
        "hallucination",
        "vector-database"
      ],
      "guides": [
        "rag-readiness",
        "governance-provenance-permissions"
      ],
      "url": "https://km4ai.org/glossary/rag.md",
      "html_planned": "https://km4ai.org/glossary/rag/"
    },
    {
      "id": "embedding",
      "term": "Embedding",
      "slug": "embedding",
      "definition": "A numerical vector representation of text (or other content) that captures semantic similarity, enabling search and retrieval beyond keyword matching.",
      "category": "ai-fundamentals",
      "related": [
        "rag",
        "vector-database",
        "semantic-search",
        "chunking"
      ],
      "guides": [
        "rag-readiness",
        "knowledge-graphs-101"
      ],
      "url": "https://km4ai.org/glossary/embedding.md",
      "html_planned": "https://km4ai.org/glossary/embedding/"
    },
    {
      "id": "ontology",
      "term": "Ontology",
      "slug": "ontology",
      "definition": "A formal model of concepts and relationships in a domain, used to constrain meaning, support reasoning, and align human taxonomies with machine-readable knowledge.",
      "category": "knowledge-structure",
      "related": [
        "taxonomy",
        "knowledge-graph",
        "metadata-schema",
        "folksonomy"
      ],
      "guides": [
        "taxonomies-ontologies-llm",
        "knowledge-graphs-101"
      ],
      "url": "https://km4ai.org/glossary/ontology.md",
      "html_planned": "https://km4ai.org/glossary/ontology/"
    },
    {
      "id": "folksonomy",
      "term": "Folksonomy",
      "slug": "folksonomy",
      "definition": "An emergent tagging vocabulary created by users rather than controlled by a central taxonomy authority; useful for discovery but often needs governance before AI consumption.",
      "category": "knowledge-structure",
      "related": [
        "taxonomy",
        "ontology",
        "metadata-schema",
        "content-governance"
      ],
      "guides": [
        "taxonomies-ontologies-llm",
        "ai-assisted-curation"
      ],
      "url": "https://km4ai.org/glossary/folksonomy.md",
      "html_planned": "https://km4ai.org/glossary/folksonomy/"
    },
    {
      "id": "tacit-knowledge",
      "term": "Tacit knowledge",
      "slug": "tacit-knowledge",
      "definition": "Know-how held by people that is hard to document\u2014judgment, shortcuts, and context. AI systems underperform when tacit knowledge is never captured or linked to explicit sources.",
      "category": "knowledge-management",
      "related": [
        "knowledge-management",
        "expertise-location",
        "content-governance"
      ],
      "guides": [
        "ai-assisted-curation",
        "measuring-km-impact"
      ],
      "url": "https://km4ai.org/glossary/tacit-knowledge.md",
      "html_planned": "https://km4ai.org/glossary/tacit-knowledge/"
    },
    {
      "id": "knowledge-graph",
      "term": "Knowledge graph",
      "slug": "knowledge-graph",
      "definition": "A network of entities and relationships that makes organizational knowledge queryable as structured links, complementary to document search and RAG chunk stores.",
      "category": "knowledge-structure",
      "related": [
        "ontology",
        "taxonomy",
        "semantic-search",
        "provenance"
      ],
      "guides": [
        "knowledge-graphs-101",
        "taxonomies-ontologies-llm"
      ],
      "url": "https://km4ai.org/glossary/knowledge-graph.md",
      "html_planned": "https://km4ai.org/glossary/knowledge-graph/"
    },
    {
      "id": "grounding",
      "term": "Grounding",
      "slug": "grounding",
      "definition": "The practice of tying AI outputs to verifiable source material\u2014citations, retrieved passages, or structured facts\u2014so answers can be checked and trusted.",
      "category": "ai-quality",
      "related": [
        "rag",
        "hallucination",
        "provenance",
        "content-governance"
      ],
      "guides": [
        "rag-readiness",
        "governance-provenance-permissions"
      ],
      "url": "https://km4ai.org/glossary/grounding.md",
      "html_planned": "https://km4ai.org/glossary/grounding/"
    },
    {
      "id": "hallucination",
      "term": "Hallucination",
      "slug": "hallucination",
      "definition": "Confident but incorrect AI output not supported by source knowledge. Often reduced by better retrieval, metadata, permissions, and content quality\u2014not by model size alone.",
      "category": "ai-quality",
      "related": [
        "grounding",
        "rag",
        "chunking",
        "content-governance"
      ],
      "guides": [
        "rag-readiness",
        "governance-provenance-permissions"
      ],
      "url": "https://km4ai.org/glossary/hallucination.md",
      "html_planned": "https://km4ai.org/glossary/hallucination/"
    },
    {
      "id": "chunking",
      "term": "Chunking",
      "slug": "chunking",
      "definition": "Splitting documents into retrieval units sized and bounded so each chunk is coherent for embedding and citation. Poor chunking is a common cause of weak RAG results.",
      "category": "ai-architecture",
      "related": [
        "rag",
        "embedding",
        "metadata-schema",
        "content-governance"
      ],
      "guides": [
        "rag-readiness"
      ],
      "url": "https://km4ai.org/glossary/chunking.md",
      "html_planned": "https://km4ai.org/glossary/chunking/"
    },
    {
      "id": "metadata-schema",
      "term": "Metadata schema",
      "slug": "metadata-schema",
      "definition": "A defined set of fields and values used to describe content (owner, audience, freshness, sensitivity, topic). Essential for filtering, governance, and AI-safe retrieval.",
      "category": "knowledge-structure",
      "related": [
        "taxonomy",
        "content-governance",
        "provenance",
        "chunking"
      ],
      "guides": [
        "rag-readiness",
        "taxonomies-ontologies-llm",
        "governance-provenance-permissions"
      ],
      "url": "https://km4ai.org/glossary/metadata-schema.md",
      "html_planned": "https://km4ai.org/glossary/metadata-schema/"
    },
    {
      "id": "taxonomy",
      "term": "Taxonomy",
      "slug": "taxonomy",
      "definition": "A controlled hierarchical classification of topics or entities. Taxonomies help humans browse and help AI systems filter and organize retrieval candidates.",
      "category": "knowledge-structure",
      "related": [
        "ontology",
        "folksonomy",
        "metadata-schema",
        "knowledge-graph"
      ],
      "guides": [
        "taxonomies-ontologies-llm",
        "ai-assisted-curation"
      ],
      "url": "https://km4ai.org/glossary/taxonomy.md",
      "html_planned": "https://km4ai.org/glossary/taxonomy/"
    },
    {
      "id": "provenance",
      "term": "Provenance",
      "slug": "provenance",
      "definition": "Traceable history of where a knowledge item came from, who changed it, and when. Required for auditability when AI systems consume and cite organizational content.",
      "category": "governance",
      "related": [
        "content-governance",
        "grounding",
        "knowledge-management"
      ],
      "guides": [
        "governance-provenance-permissions",
        "measuring-km-impact"
      ],
      "url": "https://km4ai.org/glossary/provenance.md",
      "html_planned": "https://km4ai.org/glossary/provenance/"
    },
    {
      "id": "knowledge-management",
      "term": "Knowledge management",
      "slug": "knowledge-management",
      "expanded": "KM",
      "definition": "The discipline of capturing, organizing, governing, sharing, and reusing organizational knowledge so people\u2014and increasingly AI systems\u2014can find and apply what the organization knows.",
      "category": "knowledge-management",
      "related": [
        "tacit-knowledge",
        "content-governance",
        "expertise-location",
        "ai-readiness"
      ],
      "guides": [
        "ai-assisted-curation",
        "measuring-km-impact"
      ],
      "url": "https://km4ai.org/glossary/knowledge-management.md",
      "html_planned": "https://km4ai.org/glossary/knowledge-management/"
    },
    {
      "id": "vector-database",
      "term": "Vector database",
      "slug": "vector-database",
      "definition": "A store optimized for similarity search over embeddings. Commonly used as the retrieval layer in RAG pipelines alongside or instead of traditional keyword search.",
      "category": "tooling",
      "related": [
        "embedding",
        "rag",
        "semantic-search"
      ],
      "guides": [
        "rag-readiness"
      ],
      "tools": [
        "pipeline-engine"
      ],
      "url": "https://km4ai.org/glossary/vector-database.md",
      "html_planned": "https://km4ai.org/glossary/vector-database/"
    },
    {
      "id": "content-governance",
      "term": "Content governance",
      "slug": "content-governance",
      "definition": "Policies and processes for ownership, quality, retention, permissions, and review of knowledge assets\u2014especially important when those assets feed AI systems.",
      "category": "governance",
      "related": [
        "provenance",
        "metadata-schema",
        "ai-readiness",
        "hallucination"
      ],
      "guides": [
        "governance-provenance-permissions",
        "ai-assisted-curation"
      ],
      "tools": [
        "compliance-guardians",
        "ai-metadata-taggers"
      ],
      "url": "https://km4ai.org/glossary/content-governance.md",
      "html_planned": "https://km4ai.org/glossary/content-governance/"
    },
    {
      "id": "semantic-search",
      "term": "Semantic search",
      "slug": "semantic-search",
      "definition": "Search that matches meaning rather than exact keywords, typically via embeddings. Improves findability when users ask questions in natural language.",
      "category": "ai-architecture",
      "related": [
        "embedding",
        "rag",
        "knowledge-graph",
        "vector-database"
      ],
      "guides": [
        "rag-readiness",
        "knowledge-graphs-101"
      ],
      "url": "https://km4ai.org/glossary/semantic-search.md",
      "html_planned": "https://km4ai.org/glossary/semantic-search/"
    },
    {
      "id": "expertise-location",
      "term": "Expertise location",
      "slug": "expertise-location",
      "definition": "Finding who knows what inside an organization. AI can assist by mining contributions, tags, and interactions\u2014but only when people and content relationships are maintained.",
      "category": "knowledge-management",
      "related": [
        "tacit-knowledge",
        "knowledge-management",
        "knowledge-graph"
      ],
      "guides": [
        "ai-assisted-curation",
        "measuring-km-impact"
      ],
      "url": "https://km4ai.org/glossary/expertise-location.md",
      "html_planned": "https://km4ai.org/glossary/expertise-location/"
    },
    {
      "id": "ai-readiness",
      "term": "AI readiness",
      "slug": "ai-readiness",
      "definition": "The degree to which an organization's knowledge assets are structured, governed, permissioned, and current enough for reliable AI use (copilots, RAG, analytics).",
      "category": "knowledge-management",
      "related": [
        "content-governance",
        "rag",
        "metadata-schema",
        "chunking"
      ],
      "guides": [
        "rag-readiness",
        "measuring-km-impact",
        "governance-provenance-permissions"
      ],
      "url": "https://km4ai.org/glossary/ai-readiness.md",
      "html_planned": "https://km4ai.org/glossary/ai-readiness/"
    },
    {
      "id": "datamesh",
      "term": "DataMesh",
      "slug": "datamesh",
      "definition": "In the KM4AI context: a governed access layer that connects platforms so AI and people can reach authorized organizational information without copying everything into a single unmanaged dump.",
      "category": "architecture",
      "related": [
        "content-governance",
        "ai-readiness",
        "provenance"
      ],
      "guides": [
        "governance-provenance-permissions"
      ],
      "tools": [
        "repository-integrators",
        "pipeline-engine"
      ],
      "url": "https://km4ai.org/glossary/datamesh.md",
      "html_planned": "https://km4ai.org/glossary/datamesh/"
    },
    {
      "id": "gigo",
      "term": "GIGO",
      "slug": "gigo",
      "expanded": "Garbage In, Garbage Out",
      "definition": "The principle that AI output quality is bounded by input knowledge quality. Fragmented, outdated, or ungovered sources produce unreliable assistants regardless of model choice.",
      "category": "ai-quality",
      "related": [
        "hallucination",
        "grounding",
        "content-governance",
        "ai-readiness"
      ],
      "guides": [
        "rag-readiness",
        "ai-assisted-curation"
      ],
      "url": "https://km4ai.org/glossary/gigo.md",
      "html_planned": "https://km4ai.org/glossary/gigo/"
    }
  ]
}