Patient Emotion Recognition in Human Computer Interaction System Based on Machine Learning Method and Interactive Design Theory

There are more and more human computer interaction systems (HCIS) in the medical field. Improving the service quality of HCIS and making them more intelligent is an inevitable trend in the future. Emotion recognition is of great significance for patients using HCIS. Some excellent HCIS not only satisfies the needs of patients, but also judges the emotional state of patients based on the results of emotional recognition, thereby providing more intimate medical services. Therefore, emotion recognition is crucial for HCIS. To effectively optimize the correct rate of emotion recognition, a novel emotion recognition framework based on machine learning is proposed. The core of the framework is to select the optimal classifier for different emotional data, and fuse the classification results of each classifier to get the global classification result. Experiments demonstrate that the proposed framework not only improves the accuracy of emotion recognition, but also improves the stability and reliability of the recognition results. The emotion recognition function based on the framework is applied to the HCIS design, so that the HCIS of the medical institution can better serve the patient during use, keep the patient happy, and improve the patient's happiness index and rehabilitation rate.