The goal of this research is to develop a facial emotion recognition (FER) system that uses machine learning algorithms to support users in interpreting emotions. This will involve testing various levels of automation to identify which configurations best mitigate automation bias and improve user decision-making. Different automation levels, ranging
from minimal to fully automated recommendations, will be implemented to find the optimal balance. Key steps include configuring automation levels, defining evaluation metrics, conducting user tests, and iterating based on feedback. The findings will contribute to both the field of Human-Computer Interaction (HCI) and emotion recognition by providing insights into how varying automation levels influence user trust, accuracy, and interaction quality.