Get those machines emotional!
Problem being addressed
While infusing emotion into dialogue systems can make conversational agents more human-like and benefit the interaction between human and machine, customizing and controling the agent’s emotion remains challenging.
A novel framework which extends the emotion-controllable response generation to a dual task to generate emotional responses and emotional queries alternatively. It utilizes two rewards focusing on emotion and content to improve the duality. Additionally, it applies curriculum learning to gradually generate high-quality responses based on the difficulties of expressing various emotions.
Advantages of this solution
The suggested framework significantly outperforms the baselines in terms of coherence, diversity, and relation to emotion factors.
Solution originally applied in these industries
Possible New Application of the Work
Emotional component in dialogue systems is of very high value for healthcare solutions. For example, in psychological counseling the agent is supposed to express sadness to show the sympathy and also convey happiness to cheer the patient up.
Source URL: #############