Given the heavy workload faced by teachers, automatic speaking scoring systems provide essential support. This study aims to consolidate technological configurations of automatic scoring systems for spontaneous L2 English, drawing from literature published between 2014 and 2024. The focus will be on the architecture of the automatic speech recognition model and the scoring model, as well as on features used to evaluate phonological competence, linguistic proficiency, and task completion. By synthesizing these elements, the study seeks to identify potential research areas, as well as provide a foundation for future research and practical applications in software engineering.

This paper conducts a comparative analysis of mentoring and metacognition in education, unveiling their intricate connections. Both concepts, though seemingly disparate, prove to be interdependent within the educational landscape. The analysis showcases the dynamic interplay between mentoring and metacognition, emphasizing their reciprocal influence. Metacognition, often perceived as self-awareness and introspection, is found to complement the relational and supportive nature of mentoring. Within this context, metacognitive education within mentoring emerges as a vital component. Practical recommendations are offered for effective metacognitive training, highlighting its role in enhancing cognitive and metacognitive skills. Moreover, the paper introduces the concept of a “mentoring scaffolding system.” This system emphasizes mentor-led gradual independence for mentees, facilitating their professional and personal growth. The necessity of fostering a metacognition culture in education is a central theme. Such a culture promotes improved performance and lifelong learning. The paper suggests integrating metacognition into curricula and empowering learners as essential steps toward achieving this culture. In conclusion, this paper advocates for the integration of metacognition into mentoring and education, fostering self-awareness, independence, and adaptability. These attributes are deemed crucial for individuals navigating the challenges of the information age.

In this study, the authors propose a method that combines CNN and LSTM networks to recognize facial expressions. To handle illumination changes and preserve edge information in the image, the method uses two different preprocessing techniques. The preprocessed image is then fed into two independent CNN layers for feature extraction. The extracted features are then fused with an LSTM layer to capture the temporal dynamics of facial expressions. To evaluate the method's performance, the authors use the FER2013 dataset, which contains over 35,000 facial images with seven different expressions. To ensure a balanced distribution of the expressions in the training and testing sets, a mixing matrix is generated. The models in FER on the FER2013 dataset with an accuracy of 73.72%. The use of Focal loss, a variant of cross-entropy loss, improves the model's performance, especially in handling class imbalance. Overall, the proposed method demonstrates strong generalization ability and robustness to variations in illumination and facial expressions. It has the potential to be applied in various real-world applications such as emotion recognition in virtual assistants, driver monitoring systems, and mental health diagnosis.