EduAI – Intelligent Coding Course Recommender

Overview

EduAI is a research-grade, interpretable recommendation system designed to personalize coding education for learners. The system leverages GRU-based Knowledge Tracing to model student learning dynamics, detect performance instability, and generate personalized next-problem recommendations.

The platform provides both predictive analytics and transparent interpretability, enabling educators and researchers to understand learner progress while tailoring interventions for maximum effectiveness.

---

Research Objectives

1. Model Student Knowledge Evolution
   Capture the latent knowledge state of learners over time using sequential interaction data.

2. Detect Learning Instability and Risk Patterns
   Identify sudden drops in learning performance to flag at-risk learners.

3. Personalized Problem Recommendation
   Recommend Top-N coding problems tailored to each student’s mastery profile.

4. Explainable Learning Analytics
   Combine learning curves and knowledge heatmaps to provide interpretable insights.

---

System Features

1. GRU-Based Knowledge Tracing

• Sequential modeling of learner interactions
• Embedding of problem attempts and correctness
• Dynamic latent knowledge state tracking

2. Top-N Personalized Recommendations

• Probability-based ranking of candidate problems
• Normalized Top-N scores for interpretability
• Focus on weak and high-impact learning areas

3. Learning Curve Analytics

• Exponential Moving Average (EMA): short-term performance stability
• Cumulative Accuracy: overall learning trajectory
• Sharp-drop detection to identify conceptual gaps

4. Knowledge Heatmap

• Visual representation of mastery across problems
• Color-coded: green (high mastery), yellow (medium), red (weak)
• Insight into strengths, weaknesses, and intervention points

5. Personalized Learning Plan

• Combines weak mastery items and sudden learning drops
• High-priority problems for immediate focus
• Medium-priority problems for reinforcement

---

Installation and Requirements

Python Dependencies

• streamlit
• pandas
• numpy
• torch
• plotly

Install via:

pip install -r requirements.txt

Hardware Recommendations

• Minimum: CPU, 8 GB RAM
• Recommended: NVIDIA CUDA-enabled GPU, ≥ 16 GB RAM for research-scale experiments

Dataset

• Sequential learner-problem interaction data with columns: student_id, problem_id, timestamp, correctness (0/1)
• Dataset not included due to privacy and licensing restrictions
• Synthetic or licensed datasets are recommended for reproduction

Results Visualization

Example outputs from the system:

Figure 1: Knowledge State Heatmap showing mastery levels across problems.

Figure 2: Student Learning Curve with EMA and cumulative accuracy.

Research Implications

• Provides interpretable insights for educational data mining
• Bridges predictive modeling with actionable interventions
• Supports reproducible, research-grade experiments in personalized education
• Designed to meet academic rigor for top-tier universities

Ethical Considerations

• Compliant with educational data mining ethics
• Transparent and reproducible methods
• No high-stakes automated decision-making; intended for research and teaching purposes

Summary

EduAI demonstrates a scalable, interpretable, and actionable approach to personalized coding education. By combining knowledge tracing, predictive modeling, and explainable visual analytics, it offers both researchers and educators a powerful tool to enhance student learning outcomes.

This project targets research evaluation and adoption in world-leading universities by providing a reproducible, transparent, and high-impact methodology for personalized educational interventions.