{
  "@context": "https://w3id.org/croissant/v1",
  "@type": "Dataset",

  "name": "NDL Core RAG Index",
  "description": "A retrieval-augmented generation (RAG) index built over the NDL Core Corpus. The dataset contains pre-computed text chunks with source identifiers and a FAISS vector index to enable semantic retrieval without recomputing embeddings.",

  "creator": {
    "@type": "Organization",
    "name": "The Open Data Institute"
  },

  "publisher": {
    "@type": "Organization",
    "name": "The Open Data Institute"
  },

  "license": {
    "@type": "CreativeWork",
    "name": "Open Government Licence v3.0",
    "url": "https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/"
  },

  "conditionsOfAccess": "Open access. This dataset contains derived representations (text chunks and embeddings) of public sector data. Original source records retain their own licence references in the source dataset.",

  "keywords": [
    "retrieval augmented generation",
    "rag",
    "faiss",
    "vector index",
    "ai-ready data",
    "national data library",
    "semantic search"
  ],

  "distribution": [
    {
      "@type": "FileObject",
      "name": "chunks.parquet",
      "description": "Parquet file containing text chunks and metadata, including identifiers linking each chunk to its original source record.",
      "contentUrl": "chunks.parquet",
      "encodingFormat": "application/parquet"
    },
    {
      "@type": "FileObject",
      "name": "index.faiss",
      "description": "FAISS vector index built over the embeddings of the text chunks to support fast semantic retrieval.",
      "contentUrl": "index.faiss",
      "encodingFormat": "application/octet-stream"
    },
    {
      "@type": "FileObject",
      "name": "rag_config.json",
      "description": "Machine-readable configuration describing the embedding model, chunking strategy, and index construction parameters.",
      "contentUrl": "rag_config.json",
      "encodingFormat": "application/json"
    }
  ],

  "isBasedOn": {
    "@type": "Dataset",
    "name": "NDL Core Corpus",
    "url": "https://huggingface.co/datasets/hkir-dev/ndl-core-corpus"
  },

  "intendedUse": [
    "Retrieval-augmented generation",
    "Semantic search",
    "AI agent knowledge access",
    "Policy and public sector AI experimentation"
  ],

  "limitations": "This RAG index is a static snapshot and reflects the state of the underlying source corpus at the time of index construction. It is not updated in real time.",

  "conformsTo": [
    "https://w3id.org/croissant"
  ]
}