There's currently a problem with the selection of candidates for French. It appears that the case condition is not applied as it should.
For example, in the following query, common words in lower case raises some candidates corresponding to full upper case terms:
sait - > SAIT -> Southern Alberta Institute of Technology
ou -> OU -> French NER gives institution
aise -> AISE -> Agenzia Informazioni e Sicurezza Esterna
It is not the case for English and it seems related to the French NER.
{
"text": "On ne sait pas si l'autre est à l'aise ou pas, jusqu'où on peut aller. ",
"shortText": "",
"termVector": [],
"language": {
"lang": "en"
},
"entities": [],
"mentions": [
"ner",
"wikipedia"
],
"nbest": false,
"sentence": false
}
There's currently a problem with the selection of candidates for French. It appears that the case condition is not applied as it should.
For example, in the following query, common words in lower case raises some candidates corresponding to full upper case terms: sait - > SAIT -> Southern Alberta Institute of Technology ou -> OU -> French NER gives institution aise -> AISE -> Agenzia Informazioni e Sicurezza Esterna
It is not the case for English and it seems related to the French NER.