Personalized prediction of student progress in moral education using multi modal learning.

Increased interest has been observed in recent years to find out what are the most influencing factors for student’s " performance in higher education, especially by utilizing data mining methods and techniques. Research in this area is commonly known as Educational Data Mining (EDM). The reason behind this interest is the applicability of such studies to help decide in advance about the low performing students so that their learning difficulties can be addressed and their learning performances can be improved, eventually meeting the institutional goals of providing high-quality learning environments. Also, EDM is rapidly becoming an important research area because of its ability to generate new knowledge from vast amounts of student’s data. Student’s academic performance is one of the success indicators in higher learning universities. The good academic record of achievement increases the rank of the institutions as one of the indicators of a quality university. For the student, having good academic performance will increase the chances of getting jobs where good academic achievement is one of the determining factors by employers. Nowadays, there are full-fledged, mature, and advanced technologies whereby an individual from any background, even with minimal programming experience, can forecast their future data. machine learning is now a normal technology to forecast data from grocery stores to space. Data is used by scholars and administrative personnel to predict a student’s performance upon admission, project the employment possibility for a student upon course completion or the dropout based on the combined numbers of the whole set of students, or predict an individual student’s success or failure rate in the subsequent grades. Student success at academics in ethics studies is a multidimensional and multifaceted phenomenon, and there are many variables influencing it. Contemporary literature in the field of forecasting academic success in this discipline is greatly reliant on the traditional models and single-dimensional data. All of these approaches lack in seamlessly combining heterogeneous information (such as attributes of individuals, cognition, personality, and learning environments), retaining the intricate and nonlinear relationships among these factors, and lastly, producing personalized and accurate predictions. Hence, a comprehensive and advanced framework of multimodal learning that can overcome this shortfall is acutely needed. This research seeks to supply an answer to this requirement and makes a solid step towards doing this by creating a smart model that, based on sophisticated artificial intelligence techniques, can analyze various data simultaneously and holistically and make individual predictions of student’s ethical attainment.