Text Annotation JSON

This page outlines the JSON format used for both input and output in the Text Classification Studio. It defines how NLP (Natural Language Processing) data is structured using JavaScript Object Notation (JSON). The format includes the entity's content along with its associated metadata, making it compatible with DDOE's platform.

Read here about the NLP Annotation Studio.

Input JSON format

{
  "shebang": "dataloop", // Indicates the data is intended for Dataloop platform
  "metadata": {
    "dltype": "text-classification", // Type of task: text classification
    "callId": "3154919731610621952", // Unique call identifier
    "companyId": "262834820328732"   // Company/project ID within the platform
  },
  "formatting": "html", // Specifies that the content is HTML-formatted
  "elements": [
    {
      "key": "task-a", // Unique ID for this content block
      "annotatable": true, // Set to true so annotators can label this
      "content": "<p><strong>[A_17]</strong> Hello, this is XYZ from ABC.<br><br><strong>[T_1881]</strong> Everything went well during the call.<br><br><strong>[SELLER_1017]</strong> Yes, I think we're focused on the right approach.<br><br><strong>[PROSPECT_1881]</strong> I’ll definitely follow up on that.</p>" // HTML content for annotation
    },
    {
      "key": "text-b", // Contextual or reference text
      "annotatable": false, // Not for annotation
      "content": "<p>Intent: not_decision_maker<br>Does the customer indicate that they are not the decision-maker?<br><br><br><br> <a href=\"https://gong.app.gong.io/call?id=3154919731610621952&amp;highlights=%5B%7B%22type%22%3A%22SHARE%22%2C%22from%22%3A48%2C%22to%22%3A78%7D%5D\">URL to Call 3154919731610621952</a></p>" // Instruction and external reference
    },
    {
      "key": "task-c",
      "annotatable": true,
      "content": "<p><strong>Hello</strong>, this is XYZ from ABC. Everything went well during the call.</p>"
    },
    {
      "key": "text-d",
      "annotatable": false,
      "content": "<p>Your html texts</p>"
    },
    {
      "key": "text-e",
      "annotatable": false,
      "content": "<p>Your html texts</p>"
    },
    {
      "key": "task-e",
      "annotatable": true,
      "content": "<p>Your html texts</p>"
    },
    {
      "key": "task-g",
      "annotatable": true,
      "content": "<p>Your html texts</p>"
    }
  ]
}

Output JSON format

{
	"annotations": [
		{
  "id": "6569a9bb64a270e5cc8ae821",
  "datasetId": "64b4d464bdf1fccfe1ba5790",
  "url": "https://gate.dataloop.ai/api/v1/items/6569a9bb64a270e5cc8ae821",
  "dataset": "https://gate.dataloop.ai/api/v1/datasets/64b4d464bdf1fccfe1ba5790",
  "createdAt": "2023-12-01T09:39:07.264Z",
  "dir": "/",
  "filename": "/Architecture.txt",
  "type": "file",
  "hidden": false,
  "metadata": {
    "system": {
      "encoding": "7bit",
      "isBinary": false,
      "mimetype": "text/plain", // JSON: "application/json", HTML: "text/html", CSS: "text/css", and Text: "text/plain".
      "originalname": "Architecture.txt",
      "refs": [
        {
          "id": "65941368b9bcb6a3fb4a598d",
          "type": "task"
        },
        {
          "id": "65941368b9bcb6ff2b4a598f",
          "type": "assignment"
        },
        {
          "id": "67b5b294dcaf2107243df699",
          "type": "assignment"
        },
        {
          "id": "67b5b294dcaf213f133df697",
          "type": "task"
        }
      ],
      "size": 465,
      "taskStatusLog": []
    }
  },
  "name": "Architecture.txt",
  "creator": "user@dataloop.ai",
  "updatedAt": "2023-12-01T09:39:07.304Z",
  "updatedBy": "user@dataloop.ai",
  "stream": "https://gate.dataloop.ai/api/v1/items/6569a9bb64a270e5cc8ae821/stream",
  "thumbnail": "https://gate.dataloop.ai/api/v1/items/6569a9bb64a270e5cc8ae821/thumbnail",
  "annotations": [
    {
      "id": "67b5b27511348cc12361627e",
      "datasetId": "64b4d464bdf1fccfe1ba5790",
      "itemId": "6569a9bb64a270e5cc8ae821",
      "url": "https://gate.dataloop.ai/api/v1/annotations/67b5b27511348cc12361627e",
      "item": "https://gate.dataloop.ai/api/v1/items/6569a9bb64a270e5cc8ae821",
      "dataset": "https://gate.dataloop.ai/api/v1/datasets/64b4d464bdf1fccfe1ba5790",
      "type": "text_mark",
      "label": "sentence-2",
      "attributes": [],
      "coordinates": {
        "end": 465,
        "label": "sentence-2",
        "start": 282,
        "type": "block"
      },
      "metadata": {
        "system": {
          "attributes": {},
          "automated": false,
          "clientId": "9df88b6b-44f5-4946-a6e8-ce32e2a8553b",
          "endFrame": 1,
          "endTime": 0,
          "frame": 0,
          "itemLinks": [],
          "objectId": "4",
          "openAnnotationVersion": "1.90.2-prod.0",
          "recipeId": "6569aa2b27d7bdd0605b8fc9",
          "snapshots_": [],
          "startTime": 0,
          "system": false
        },
        "user": {}
      },
      "creator": "user@gmail.com",
      "createdAt": "2025-02-19T10:29:09.919Z",
      "updatedBy": "user@gmail.com",
      "updatedAt": "2025-02-19T10:29:09.919Z",
      "hash": "83855a80fca7f5ef19fc57f67fcafe7af1fbd666",
      "source": "ui"
    }
  ],
  "annotationsCount": 1,
  "annotated": true
}
	]
}

Dictionary Table