PindoDocs
Web AppWeb

Named Entity Recognition

Identify and classify named entities (people, organizations, locations, dates, and more) in text. NER supports Kinyarwanda and English.

POST/ai/ner/{lang}

Headers: Content-Type: application/json

Request

FieldTypeDescription
textrequiredstringThe text to extract entities from.
langrequiredstringLanguage of the text: rw or en.
labelsstring[]Entity types to extract. Defaults to ["person", "location", "organisation", "date"].

Pass any keywords you like in labels, for example person, location, organisation, date, time, money, percent, quantity, ordinal, or cardinal.

Example request

# Public access
curl -X POST "https://api.pindo.io/ai/ner/rw/public" \
     -H "Content-Type: application/json" \
     -d '{
       "text": "Yohani ukorera minisante atuye i musanze.",
       "lang": "rw",
       "labels": ["person", "location", "organisation"]
     }'

# Authenticated access
curl -X POST "https://api.pindo.io/ai/ner/rw" \
     -H "Content-Type: application/json" \
     -H "Authorization: Bearer YOUR_ACCESS_TOKEN" \
     -d '{ "text": "Yohani ukorera minisante atuye i musanze.",
           "lang": "rw",
           "labels": ["person", "location", "organisation"] }'

Response

Returns a map of each requested label to the entities found.

{
  "person": ["Yohani"],
  "location": ["musanze"],
  "organisation": ["minisante"]
}

Try it

Interactive PlaygroundExtract named entities from Kinyarwanda text
Access mode
Labels
personlocationorganisation
cURL
curl -X POST "https://api.pindo.io/ai/ner/rw/public" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Yohani ukorera minisante atuye i musanze.",
    "lang": "rw",
    "labels": ["person", "location", "organisation"]
  }'