File size: 1,463 Bytes
a30026f | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 | ---
title: PLRS Logic Engine
emoji: 🧠
colorFrom: blue
colorTo: indigo
sdk: streamlit
sdk_version: 1.33.0
app_file: app.py
pinned: true
license: mit
tags:
- education
- knowledge-tracing
- recommendation-system
- pytorch
- transformers
---
# PLRS — Personalized Learning Recommendation System
> Constraint-aware personalized learning recommendations powered by Self-Attentive Knowledge Tracing (SAKT) and DAG prerequisite constraints.
## What it does
PLRS combines a SAKT transformer model with a curriculum knowledge graph to generate recommendations that are both **personalized** and **pedagogically sound**. Topics are classified into three tiers:
- ✅ **Approved** — prerequisites met, ready to learn
- ⚠️ **Challenging** — prerequisites partially met
- ❌ **Vetoed** — prerequisites not met, blocked
## Key results
| Metric | PLRS | Collaborative Filtering |
|--------|------|------------------------|
| Val AUC | **0.7692** | — |
| Prerequisite Violation Rate | **0.0%** | 81.3% |
## Bundled curricula
- **Nigerian Secondary School Mathematics** (38 topics, 45 edges, JSS3–SS2)
- **CS Fundamentals / Digital Technologies** (31 topics, 39 edges)
## Architecture
```
Student History → SAKT → Mastery Vector → DAG Constraint Layer → Ranker → Recommendations
```
## Links
- 📦 GitHub: [clementina-tom/plrs](https://github.com/clementina-tom/plrs)
- 📄 Paper/Report: Final Year Project, Computer Science
|