Neural Personalized Response Generation as Domain Adaptation
This work addresses the problem of creating more human-like and tailored conversational agents, but it appears incremental as it builds on existing encoder-decoder frameworks with a specific adaptation method.
The paper tackles personalized response generation for conversational systems by proposing a two-phase initialization and adaptation approach based on sequence-to-sequence learning, achieving good results in modeling human responding style and generating personalized responses as indicated by lexical divergence metrics.
In this paper, we focus on the personalized response generation for conversational systems. Based on the sequence to sequence learning, especially the encoder-decoder framework, we propose a two-phase approach, namely initialization then adaptation, to model the responding style of human and then generate personalized responses. For evaluation, we propose a novel human aided method to evaluate the performance of the personalized response generation models by online real-time conversation and offline human judgement. Moreover, the lexical divergence of the responses generated by the 5 personalized models indicates that the proposed two-phase approach achieves good results on modeling the responding style of human and generating personalized responses for the conversational systems.