# A Predictive Model for Student Performance in Classrooms Using Student Interactions With an eTextbook

**Authors:** Ahmed Abd Elrahman, Taysir Hassan A Soliman, Ahmed I. Taloba, Mohammed F. Farghally  
**arXiv:** [2203.03713](https://arxiv.org/abs/2203.03713)  
**Date:** 2022-02-16  
**Summary page:** https://www.summarizepaper.com/en/paper/2203.03713v1/

## Key points

- Online eTextbooks and MOOCs have led to the collection of vast amounts of data related to students' learning
- Educators can gain valuable insights into their students' performance and behavior by analyzing this data
- The authors propose a new model for predicting student performance based on an analysis of how students interact with an interactive online eTextbook
- Predicting students' performance early in the course can help identify those at risk and provide suitable interventions
- The authors considered two main issues: predicting good/bad performance and predicting the final exam grade
- They evaluated the most popular classification and regression algorithms on data from a CS2 course offered in a large public research university
- They applied Random Forest Regression, Multiple Linear Regression, Logistic Regression, decision tree, Random Forest Classifier, K Nearest Neighbors, and Support Vector Machine in their analysis
- This approach enabled them to predict student performance accurately
- This paper presents a promising approach for predicting student performance based on their interactions with an interactive online eTextbook which could help improve educational outcomes for all learners.

## Summary

The rise of online eTextbooks and Massive Open Online Courses (MOOCs) has led to the collection of vast amounts of data related to students' learning. Educators can gain valuable insights into their students' performance and behavior by carefully analyzing this data. In this paper, authors Ahmed Abd Elrahman, Taysir Hassan A Soliman, Ahmed I. Taloba, and Mohammed F. Farghally propose a new model for predicting student performance based on an analysis of how students interact with an interactive online eTextbook. By predicting students' performance early in the course, educators can easily identify those at risk and provide suitable interventions. The authors considered two main issues: predicting good/bad performance and predicting the final exam grade. To build their proposed model, they evaluated the most popular classification and regression algorithms on data from a data structures and algorithms course (CS2) offered in a large public research university. The authors applied Random Forest Regression and Multiple Linear Regression in regression analysis while using Logistic Regression, decision tree, Random Forest Classifier, K Nearest Neighbors, and Support Vector Machine in classification analysis. This approach enabled them to predict student performance accurately. Overall, this paper presents a promising approach for predicting student performance based on their interactions with an interactive online eTextbook which could help improve educational outcomes for all learners by providing educators with early warning signs of at-risk students.