Wide & Deep Learning for Judging Student Performance in Online One-on-one Math Classes
Project Overview
The document explores the integration of generative AI in education, particularly through the development of a Wide & Deep learning framework designed to enhance the assessment of student performance in online one-on-one math classes. It highlights the significance of precise evaluations to inform teaching strategies and details the application of deep learning techniques to analyze interactions during classroom sessions. By combining handcrafted features with deep learning methodologies, the study reveals improvements in performance predictions, showcasing the potential of AI to automate and refine the judgment process in educational environments. The findings suggest that leveraging advanced AI technologies can lead to more effective teaching outcomes by providing educators with deeper insights into student engagement and understanding. Overall, the document underscores the transformative role of generative AI in personalizing education and optimizing learning experiences.
Key Applications
Wide & Deep learning framework for predicting student performance
Context: Online one-on-one math classes for grade 8 students
Implementation: The framework combines handcrafted features to capture teacher-student interactions with deep learning models that analyze conversational data.
Outcomes: Achieved improved accuracy in predicting student mastery levels with a reported accuracy of 0.728 and better F1 scores compared to existing methods.
Challenges: Dependence on the quality of conversational data and the need for effective feature engineering.
Implementation Barriers
Data Quality
The effectiveness of the AI model relies on the quality and quantity of the classroom conversation data.
Proposed Solutions: Utilizing robust data collection methods and improving the transcription accuracy of conversational data.
Project Team
Jiahao Chen
Researcher
Zitao Liu
Researcher
Weiqi Luo
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Jiahao Chen, Zitao Liu, Weiqi Luo
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai