Table 1 A brief overview of related work

Early prediction of undergraduate Student’s academic performance in completely online learning: a 5-year study [15]

Proposed a collection of AI models to predict student academic progress from LMS interaction data and student academic data like GPA and enrolment test data. The data consists of LMS log files, demographics, and academic achievement. No research methodology is identified

Predicting Students’ Academic Performance Through Supervised Machine Learning [61]

Developed an AI based system to predict student performance from their demographical and LMS interaction data. The dataset comprises of demographical characteristics and LMS interaction data including gender, country, birthplace, view of the LMS content, quiz attempts, and assessment submissions. The nature of the dataset does not allow early prediction. The research methodology is not clear

Predicting Students’ Academic Procrastination in Blended Learning Course Using Homework Submission Data [62])

Develop an algorithm to enhance students’ academic progress by detecting struggling students through their homework submission behaviours e.g., no submission or late submission. The nature of the dataset does not allow enough time to offer timely interventions and support to enhance student academic performance. No research methodology is identified to construct the predictive model e.g., DSR or DBR

An Efficient Approach for Multiclass Student Performance Prediction based upon Machine Learning [52]

Predicted the students’ performance by using four classification algorithms

The same dataset is used in other studies as well but with different ML classifiers [63, 64]. The study used secondary school students, not HE and did not use of LMS data

Used socio-economic attributes of students which do not allow timely identification of the at-risk student. The research approach is not based on the similarities of DSR and DBR principles

Design, development, and evaluation of a mobile learning application for computing education [65]

Applied DSR approach to developing mobile learning application for HE for better student learning. The research approach is only based on the DSR approach and not on DBR principles or similarities between DSR and DBR. No AI (DL or ML) models are used to predict student academic performance

Predicting Student Performance in Higher Educational Institutions Using Video Learning Analytics and Data Mining Techniques [66]

Created a model to predict student overall performance at the end of the semester by analysing student academic information and video interactions data. The model is trained and tested using was tested with eight classification algorithms. The research approach used is quantitative prediction methodology which is not based on the similarities of DSR and DBR principles. The study mentioned early stages, but it does not state a definitive timeframe within the semester to show whether there is enough time to offer support to enhance student performance