Emotional-Aware Virtual Assistants
for superior user engagement
The platform offers sentiment management capabilities. It enables virtual assistants to ‘understand’ the user’s emotions by deciphering verbal and sentence structure clues and parses user utterances for specific words, phrases, and modifiers.
Assistants can handle the happier customers themselves while transferring the unhappier customers to human agents.
Sentiment Types
Emotions influence people’s choices, how they interact with others and their perception of the world around them – including your brand. The platform analyzes user utterances and classifies them based on detected levels of anger, disgust, fear, sadness, joy, and positivity.
However, emotions are not mutually exclusive and usually experience along a gradient. Our tone algorithm and platform intelligence accounts for this reality and uncovers by scoring multiple emotions in a single utterance. For example, an input could yield a high score for joy but a mild score for sadness at the same time.
Sentiment Scoring
The platform’s NL engine scores and ranks sentiment on a scale of -3 to 3, based on the intensity of the emotions detected. This number represents how high or how little the user input corresponds to the six core emotions.
It provides sentiment analysis results as context object variables at two different levels; tone emotions and scores for the current conversation and average tone emotions and scores for the entire session.
Conditional Transitions
Sentiment scores can be stored as context and used to drive the flow of a dialog task via conditional transition statements. It allows you to use these scores to steer user conversations or seamlessly invoke escalation to a live agent.
