API Specs

POST /api/v1/predict

Since predicting may take some time depending on the model we choose, etc, the consumer will POST the paper details to this endpoint, which will return a 202 Accepted with a header "Location" with the URI where results will be once they are ready.

Param	Description	Required	Default	Example
abstract	Text of the abstract	Yes	n/a	"In this paper we discuss a thing and find some results"

Response: 202 Accepted

Location: /api/v1/predict/ID

GET /api/v1/predict/ID

Response: json

{
    "policy": {
        "prediction": bool,
        "confidence": float,
    }, 
    "public understanding": {
        "similarity_score": float,
    }
}