Academic grades in assessments are predicted to determine if a student is at risk of failing a course. Sequential models or graph neural networks that have been employed for grade prediction do not consider relationships between course descriptions. We propose the use of text mining to extract semantic, syntactic, and frequency-based features from…