Development of artificial intelligence software that instantly measures students' attention to the lesson

The purpose of this study is to develop and implement an artificial intelligence software for the instantaneous measurement of students' attention levels in the classroom. Additionally, the opinions of instructors regarding the measurement of students' attention levels using artificial intelligence software in a classroom setting were examined. The study employed a mixed research method encompassing both qualitative and quantitative approaches. The research was conducted according to the explanatory sequential design, which is one of the mixed research methods. The quantitative dimension of the research was conducted with 396 students studying at the Forensic Computing department of Fırat University Faculty of Technology during the fall semester of 2021-2022. The qualitative dimension of the research, on the other hand, was carried out with 5 instructors involved in the implementation phase of the research. In the quantitative dimension of the research, the developed artificial intelligence software within the scope of the study, the attention control scale and the attention level labeling forum filled out by three field experts were used as the data collection tool, while a semi-structured interview form was employed in the qualitative dimension. In the initial stage of the research, an artificial intelligence algorithm was developed to measure students' attention levels using readily available and freely shared photos on the internet. The results obtained from the application indicated that the artificial intelligence software achieved a 97.2% accuracy in classifying attention levels. Furthermore, strong evidence for the measurement of students' attention levels using the developed artificial intelligence software was obtained through a pilot study conducted with 10 students. The final implementation of the research conducted with 396 students demonstrates that the artificial intelligence software can collect information about the position of students' heads and visually report their instantaneous attention status. In addition, a positive significant correlation was detected between the students' average attention level measured by artificial intelligence and the attention level measured by the educational sciences expert, instructor and body language expert. In addition, a strong positive significant correlation was determined between the students' average scores from the attention control scale and the average attention level obtained with artificial intelligence. The findings indicate that students' attention levels can be measured and communicated to educators in real-time using artificial intelligence software in a classroom environment. However, in line with the opinions of instructors, it was determined that the artificial intelligence-based attention measurement method has both positive aspects and potential drawbacks. Recommendations for future research are also provided.