Geotagging API provides a simple interface to interact with Geotagging.ai’s deep learning models on text-based geolocation identification. The models take into account languages, dialects, mentioned entities, events, and numerous location-indicative semantic features of the text input to predict the text author’s location.
The models support over 60 languages (find the full list below) and don’t have any restrictions on the minimum number of words/characters. The accuracy, however, increases with the number of characters (up to 250 characters).
The underlying models represent State-Of-The-Art results in the field of geospatial analysis.
The API endpoints are set up to simplify the process for a user. The only parameter required is a string “text”.
Example request:
{ "text": "上海太大了住在这里" }
Example response:
"geotagging": { "lat": 30.830730438232422, "lon": 116.61170196533203, "confidence": 0.9544845223426819, "version": "2.0" }
Currently, the API takes arbitrary strings of 250 characters max on the input. Future upgrades will account for a bigger text size.
The models analyze the text for features relevant to geolocation detection and will only return results if the received text contains enough unique features to signify the location.
The next version of the API will: