Design of Robot Interactive Emotion Model Based on Artificial Intelligence Technology
Rongchang Liu · 2023
At present, in AI (artificial intelligence) and machine learning, most learning and decision-making algorithms only consider the external motivation rewards from external goals or costs, but ignore the motivation rewards from the internal cognition and emotion of the machine itself. This paper mainly constructs a set of robot interactive emotion model based on AI technology, and uses the end-to-end generation model to generate a reasonable reply that best fits the situation by combining emotional factors. The Chinese text is classified by using the character-level CNN (Convolutional Neural Network). In the actual dialogue scene classification, the word vector similarity classification method is integrated to improve the scene distribution accuracy of the model in this paper and increase the user experience. The emotional evaluation results show that the average value of ECM (Emotional Chat Machine) model is 1.213, and the average value of the model proposed in this paper is 1.618, with a difference of 0.405. The results show that the model proposed in this paper is superior to the ECM model in emotional content expression, and the sentences generated by the model proposed in this paper have content information close to the topic.