<![CDATA[The development of artificial intelligence technology, particularly deep learning, has opened up opportunities for analyzing students' non-cognitive aspects, such as emotions and focus levels during the learning process. This study aims to analyze the development of the use of deep learning-based facial expression recognition (FER) in identifying students' emotions and focus, and to examine the relationship between these two aspects in Indonesian language learning for fourth-grade elementary school students. The scope of the study focuses on a review of various previous studies related to the application of FER in educational contexts, particularly those related to the analysis of student engagement. The method used is a qualitative approach with a literature review. Data were obtained from various scientific sources in the form of relevant research journal articles. The data analysis process was carried out using qualitative descriptive techniques through the stages of data reduction, data presentation, and drawing conclusions by comparing and interpreting findings from various previous studies. The results of the study indicate that deep learning-based FER technology, especially with the Convolutional Neural Network (CNN) model, is able to classify various student emotional expressions such as happy, sad, angry, neutral, confused, and bored with a relatively high level of accuracy. In addition, it was found that students' emotional states have a significant relationship with the level of learning focus, where positive emotions tend to increase concentration, while negative emotions contribute to decreased attention. Thus, it can be concluded that the integration of FER-based emotion and focus analysis has the potential to become a more objective, adaptive, and technology-based learning evaluation tool. However, its implementation still requires further development and attention to technical aspects, infrastructure readiness, and ethical use of student data.]]>
