Research on the Application of an Open Domain Conversation Generation Algorithm Integrating Emotional Information in a Chat Robot for Preschool Education
To further improve the educational quality of learning and educational robots,a dialogue generation model based on improved emotional supervision is proposed.Among them,the dialogue generation method based on the bidirectional long-term and short-term memory network Bi-LSTM is combined with the method of emotional supervision to further improve the quality of dialogue responses produced.The experimental results show that compared with traditional end-to-end models and unimproved emotional su-pervision models,the improved emotional supervision model generates significantly higher quality,smoother,more natural,and tai-lored to people's actual conversations.It has significant advantages in evaluating the diversity of bigram indicators.In summary,the designed dialogue generation model based on improved emotional supervision has good performance and can generate high-quality dia-logue responses.When applied in preschool education scenarios,it can help children engage in higher quality dialogue learning.
万方数据
维普
