Classroom - Learning


Enhancing Student Performance Assessment: A Case Study on ELO-Based Rating System with Recurrent Neural Network Integration

Szymon Fonau, CTO @ OctoShrew


This case study explores the development of an innovative project for an education company that leverages an ELO (Elo Rating System)-based system to rate student performance on exercises. To enhance the accuracy and effectiveness of the rating system, a recurrent neural network (RNN) is integrated to adjust the ELO ratings based on various factors such as exercise completion time, time intervals between exercises, and the performance of other students on the same exercises. The study delves into the development process, the functionality and features of the system, and the potential benefits and impact on educational assessment and student learning.


The assessment of student performance is a critical aspect of education. Traditional grading systems often fall short in providing a comprehensive understanding of a student's abilities and growth. This case study focuses on a project developed by an education company, aimed at revolutionizing student performance assessment. The project combines an ELO-based rating system, originally designed for competitive games, with the integration of a recurrent neural network (RNN) to adjust ratings dynamically based on additional factors.


The ELO rating system, known for its use in chess and other competitive games, provides a framework for assessing relative skill levels of players. However, applying this system to educational exercises requires customization and adaptation. Traditional grading methods often struggle to capture the nuances of student performance and growth. This project addresses these challenges by integrating a recurrent neural network to enhance the ELO-based rating system.

Development Process:

The development process of the project involved collecting and preparing student exercise performance data. This data was then used to create an ELO-based rating system, mapping student performance to a rating scale. The recurrent neural network was designed and trained to analyze various factors, such as exercise sequence, time taken, and peer performance. The system underwent iterative adjustments and fine-tuning to ensure accuracy and reliability.

Functionality and Features:

The ELO-based rating system assigns initial ratings to students and dynamically adjusts them based on exercise performance. The integration of the recurrent neural network allows for more accurate rating adjustments by considering factors like exercise sequence and time intervals between exercises. Peer performance and class-wide data are also incorporated to influence ratings, fostering healthy competition and motivation.

Benefits and Impact:

The project brings several benefits to student assessment and learning. The ELO-based rating system provides a personalized assessment, offering insights into individual strengths, weaknesses, and areas for improvement. It promotes healthy competition and motivation among students, while the tailored exercises based on individual performance and ratings enhance learning experiences. Educators benefit from effective feedback and intervention strategies, enabling timely and personalized support for struggling students.


The integration of an ELO-based rating system with a recurrent neural network presents an innovative approach to student performance assessment. By combining the inherent strengths of the ELO system with the adaptive capabilities of the RNN, the project enhances the accuracy and effectiveness of student rating. The case study has highlighted the development process, system functionality, and the potential benefits and impact on educational assessment and student learning. Future advancements and applications of this project hold the potential to reshape assessment practices and foster optimal learning outcomes for students.


Related blogs