Possible to whitelist or 'train' the AI?
We had a customer post 'I liked your service so much I cancelled my appointments with other providers' and this was scored as Negative sentiment. Is there a way to flag it as positive and have the service learn from it?
Anonymous
shared this idea
1 comment
-
Hello,
Thank you for your feedback. Can you provide the API version that produced the negative sentiment? Our v3 API returns a positive sentiment for the text you posted. Here is the output JSON:
{
"id": "4",
"sentiment": "positive",
"confidenceScores": {
"positive": 0.65,
"neutral": 0.0,
"negative": 0.35
},
"sentences": [
{
"sentiment": "positive",
"confidenceScores": {
"positive": 0.65,
"neutral": 0.0,
"negative": 0.35
},
"offset": 0,
"length": 78,
"text": "I liked your service so much I cancelled my appointments with other providers."
}
],
"warnings": []
}
