Think Twice: A Human-like Two-stage Conversational Agent for Emotional Response Generation

Towards human-like dialogue systems, current emotional dialogue In the task of open-domain dialogue generation, emotional dialogue approaches jointly model emotion and semantics with a unified aims to generate responses involving the perception and expression neural network. This strategy tends to generate safe responses of proper emotions. A large number of studies [30, 35, 40] have due to the mutual restriction between emotion and semantics, and demonstrated that emotional dialogue can significantly improve requires the rare large-scale emotion-annotated dialogue corpus. Inspired users' satisfaction in a human-machine conversation. Moreover, by the "think twice" behavior in human intelligent dialogue, building a dialogue system with human emotions is one of the we propose a two-stage conversational agent for the generation of ultimate goals of artificial intelligence.