Research on Emotional Generative Chatting Method for Preschool Education Chat Robots
In order to address the issue of traditional chat robots being unable to meet the emotional needs of children and par-ents,a dialogue emotion recognition model based on sequential contextual context and user information was first proposed in the stud-y.Then,an emotion generative chat method was designed,and finally,an intelligent interaction model for preschool education chat robots was constructed.The research results show that the average accuracy and F1 values of the dialogue emotion recognition model in the recognition of positive and negative emotions are 98.3％and 98.7％,98.5％and 98.7％,respectively.Moreover,the genera-tive emotion chat model has increased by 1.78 compared to the extraction emotion chat model in bilingual evaluation substitution.The final user's comprehensive score is 4.9,and 93.6％of users have a five-star rating.In summary,the method proposed in the study has good performance and has achieved emotional recognition and empathy recovery functions for children and parents,with excellent application results.
万方数据
维普
