Hi, I was trying to do some reporting around sentiment analysis scores, but don't understand why there are 2 sentiment scores - oSentimentScore and oCustomerSentiment. What do each of these mean and how are they different? I have tried searching for their definitions, but I have not found them documented anywhere.
but I have not found them documented anywhere.
Have you seen the Metrics page?
| Metric | Description | |
|---|---|---|
| oCustomerSentiment | Score of the sentiment detected for a participant in a conversation | |
| oSentimentScore | Overall sentiment score detected for a conversation. |
Yes, but these do not explain the difference or how they are calculated. From what I understand, the sentiment score is only calculated for the customer participant as I see from
/api/v2/speechandtextanalytics/conversations/{conversationId}/communications/{communicationId}/transcripturl
So why are there 2 different scores if only the customer sentiment is evaluated?
sentiment score is only calculated for the customer participant
If I understand correctly, then if there is a single customer, then there should be one oCustomerSentiment and oSentimentScore, and they should be the same? That does not appear to be the case. When I query
/api/v2/analytics/transcripts/aggregates/query then on the same conversation, I get two differing scores when there is only a single one for both.
I have no issue understanding math, but that does not explain why the two metrics are different. Is the oSentimentScore a different equation than the oCustomerSentiment?
The article is also worded unclearly, as it switches between "overall customer sentiment score" and "overall sentiment score". Which of those two metrics is it referring to?