Genesys Cloud - Main

I am working on incorporating sentiment trend as a KPI in our call center, but we are struggling to match the data we are receiving from pure insights to match what is shown in Genesys. I cannot tell if there are multiple versions/calculations of sentiment trend, or if they are the same thing. 

For example, when under the Agent Performance View in Genesys, I can see "Avg Sentiment" and "Average Sentiment Score."

Avg Sentiment - Phrase-level customer sentiment; simple average, regardless of events' relative location in the interaction.
Average Sentiment Score - Conversation-level sentiment score; weighted average, relative to events' location int he interaction. 

My understanding is that "Avg Sentiment" is describing sentiment score and that "Average Sentiment Score" here is describing Sentiment Trend. 

Sentiment Trend according to the Resource Center: The sentiment trend is determined by comparing the sentiment events found in the first half of the interaction to the sentiment found in the second half of the interaction. The calculation is: sentimentTrend = (sentiment score of end-events – sentiment score of start-events) / 2. 

When pulling from Pureinsights, we are able to see both a sentiment score and a sentiment trend. However, the sentiment trend here does not match the "Average Sentiment Score" that is shown in Genesys despite having the same filters applied. 

My question is, are there more than 2 types of sentiment scores? Or is it just sentiment score and sentiment trend, and these are all just different naming conventions for the same result? If so, does anyone have an idea as to why the sentiment score that is pulled out of Genesys would not match up with either of the scores in Genesys? 

Yes, this is all covered in the official Resource Center article:
https://help.genesys.cloud/articles/understand-sentiment-analysis/

According to the documentation, Genesys defines distinct sentiment metrics, not just different naming conventions:

• Phrase-level sentiment (used for averages like Avg Sentiment), which is a simple average of sentiment markers across phrases.
• Overall / conversation-level sentiment score, which is a weighted average, giving more importance to sentiment events toward the end of the interaction.
• Sentiment Trend, which is calculated separately by comparing sentiment in the first half versus the second half of the interaction.