Development of an AI Model for Course Recommendation and Academic Planning in University Curriculums Using Large Language Models

A Retrieval-Augmented Generation (RAG)-based system for intelligent course recommendation and study planning using Large Language Models (LLMs), powered by LangChain.

---

📌 Overview

This project implements a Retrieval-Augmented Generation (RAG) pipeline to support AI-driven course recommendation and study planning. The system connects multiple components—including data retrieval, embeddings, and large language models—using the LangChain framework, with LangSmith for model monitoring and evaluation.

---

🧠 Models Used

Large Language Models (LLMs):

• LLaMA 3.3 70B

• TYPHOON AI 2 70B

Embedding Models:

• Cohere Multilingual 3.0

• BGE M3

---

🔍 Key Findings

Cohere Multilingual 3.0 performs better than BGE M3 in retrieving relevant academic data.

The best-performing combination:

• Embedding: Cohere Multilingual 3.0

• LLM: TYPHOON AI 2 70B

Achieved precision = 0.825, demonstrating high-quality information filtering.

🌐 Data Source

• Course information was collected via web scraping using BeautifulSoup and Selenium.

• The source website: Faculty of Science, Silpakorn University.

• The dataset covers the latest academic year: 2567 (2024).

---

🔧 Tech Stack

• LangChain: Orchestrates retrieval, generation, and overall pipeline.

• LangSmith: Used for monitoring, debugging, and tracking model performance.

• BeautifulSoup + Selenium: For scraping structured course data from university webpages.