{
  "name": "Octoparse Docs RAG-Ready Web Data Sample",
  "version": "1.0",
  "description": "A public evaluation sample showing how Octoparse-owned documentation can be delivered as cleaned, deduplicated, provenance-preserving source records for a customer-managed RAG pipeline.",
  "generated_at": "2026-07-16T23:20:37.574Z",
  "source_owner": "Octoparse",
  "copyright_holder": "Octoparse",
  "usage": "evaluation_only",
  "rights": "Copyright Octoparse. Provided for evaluation only; no redistribution license is granted.",
  "source_scope": "Selected public English Octoparse Docs pages from the platform and academy collections.",
  "record_count": 24,
  "duplicate_source_count": 0,
  "formats": ["JSONL", "CSV"],
  "fields": [
    "document_id",
    "source_url",
    "domain",
    "title",
    "clean_text",
    "content_type",
    "language",
    "published_at",
    "updated_at",
    "collected_at",
    "content_hash",
    "dedup_status"
  ],
  "methodology": [
    "Parse the semantic article container from each public source page.",
    "Remove interface text and normalize whitespace while preserving readable document structure.",
    "Validate required metadata and minimum clean-text length.",
    "Create a SHA-256 content hash and remove exact duplicate source documents.",
    "Deliver source-level documents without chunking, embeddings, vector indexes, or retrieval logic."
  ],
  "intended_use": "Evaluate source-document structure, provenance, and cleanliness before scoping a custom web dataset.",
  "limitations": [
    "This is a point-in-time sample, not a continuously refreshed production feed.",
    "The sample contains Octoparse-owned public documentation only.",
    "Published and updated timestamps remain null when the source page does not expose them.",
    "Customers remain responsible for chunking, embeddings, indexing, retrieval, and application behavior."
  ],
  "service_url": "https://www.octoparse.com/data-service/web-data-for-ai",
  "guide_url": "https://www.octoparse.com/data-service/web-data-for-ai/how-to-build-a-rag-ready-dataset"
}