Using the API

Arabic enrichment

Dialect, sentiment, entities and topics for Arabic text.

Add enrich=arabic to a profile, post or comment read and every returned object gets an enrichment block. It costs 3 extra credits per request, refunded if enrichment fails (then enrichment is null).

enrichment

"enrichment": {
  "dialect": "gulf",          // gulf | egyptian | levantine | maghrebi | msa | mixed | not_arabic
  "sentiment": { "label": "positive", "score": 0.72 },
  "entities": [{ "text": "الرياض", "type": "place" }],
  "topics": ["food", "travel"]
}

Enriching your own text

POST /v1/enrich classifies up to 50 texts you send, for 3 credits per request.

shell

curl https://api.crawlfeed.dev/v1/enrich \
  -H "Authorization: Bearer $CRAWLFEED_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": ["والله المطعم هذا يجنن", "الخدمة كانت بطيئة جدا"]}'

Dialect is not_arabic for text that is not Arabic, and mixed when a text switches between dialects or languages.