Knowledge workspace / Import

Make your research searchable.

Bring your language data into one trusted library, with its sources, meanings, and examples intact.

Add research data

Choose a file or paste a structured JSON document.

Choose your research fileJSON, CSV, or plain text
Use UTF-8 encoding for Kashmiri and Urdu text.Browse files

Imported records become part of your research corpus. Review your source before importing.

Administrator access

An admin key is required to import data or build the search index. It stays in memory until you refresh or close this page.

A better corpus starts with context

Prepare your data

01 / Preserve the source: Include a title, author, and year wherever available. Your source information stays linked to each annotation.
02 / Choose your format: JSON supports nested examples. CSV uses a header row with originalText, language, annotationType, and meaning. TXT creates one Kashmiri sentence per non-empty line.
03 / Add meaningful context: Translations, usage notes, and examples help distinguish meanings. Importing adds retrieval data; it does not retrain model weights.

JSON format example

{
  "source": {
    "title": "Source title",
    "author": "Author",
    "year": 2026,
    "notes": "Short source note"
  },
  "annotations": [
    {
      "originalText": "گھر",
      "language": "kashmiri",
      "annotationType": "word",
      "meaning": "house; home",
      "contextualMeaning": "Depends on usage context.",
      "translation": "home",
      "examples": [
        {
          "exampleText": "یہ چھُ میون گھر۔",
          "translation": "This is my home.",
          "language": "kashmiri",
          "context": "Everyday speech"
        }
      ]
    }
  ]
}