Employee Emotion Recognition and Job Satisfaction Enhancement in Hotel Management: A Convolutional Neural Network-Based Solution

Abstract

As the impact of employees' emotions on job satisfaction and performance has been increasingly emphasized, how to accurately predict employees' job satisfaction has become an important topic in the management field. However, existing satisfaction prediction methods usually rely on a single data source and fail to fully consider the dynamic changes and diversity of emotions, resulting in less accurate prediction results. To address this issue, this study proposes the EmoFusionNet model, which combines emotion recognition with job satisfaction prediction by fusing facial expression images and speech emotion signals, taking full advantage of the complementary nature of multimodal data. The model uses feature-weighted fusion, emotion classification and regression analysis to effectively improve the accuracy of employee job satisfaction prediction. Experimental results show that EmoFusionNet significantly outperforms traditional unimodal methods in multimodal emotion recognition tasks, especially in employee emotion recognition and real-time satisfaction prediction, exhibiting high accuracy and robustness. This study provides an intelligent emotion management tool for industries such as hotel management, and also provides a useful reference for future work environment optimization based on multimodal data.