Text-Processing
Por
Jacob |
Actualizada 20日前 |
Text Analysis
Popularidad
9.6 / 10
Latencia
319ms
Nivel de servicio
100%
Health Check
N/A
Volver a todas las conversaciones
Obama is an ORG?
jdimm
10年前
Input to NER tagger: Barack Obama and Clinton
Output:
{
“text”: “(S\n (PERSON Barack/NNP)\n (ORGANIZATION Obama/NNP)\n and/CC\n (PERSON Clinton/NNP))”
}
I’ll try some more, but this is a pretty easy test case. You must have seen “Barack Obama” frequently in training data.
Únase a la conversación, añada un comentario a continuación:
Inicie sesió/Regístrese para publicar nuevos comentarios
Another test: Barack Obama and Hilary Clinton and Mitt Romney and Michele Obama and Barack Obama
That gets all but the first “Obama”. The “Barack Obama” at the end of the string is correctly identified.
NER tends to work better in a real context instead of just lists of names. But my models are also trained on somewhat old data, so I’m not sure they’ve actually encountered Obama specifically.