Using the customer side of an interaction, analysis is conducted on the text to identify if the overall sentiment of the interaction is positive, neutral or negative.
Scoring is from -1 to +1 where:
-1 represents very negative sentiment
0 represents neutral sentiment
+1 represents very positive sentiment
The way we calculate sentiment is using a standard NLP (Natural Language Processing) approach. A dictionary containing word weights from negative, to neutral to positive is used.
The model runs through all the words in the text and applies the score for each word. These scores are aggregated and divided by the length of the text to provide a consistent comparison.