Personalized-Recommendation-System

TrendMatrix

TrendMatrix is a personalized recommendation web app built with Python and Streamlit. It demonstrates how real-world recommendation systems handle personalization, cold-start users, and UI robustness in a deployable application.

---

🚀 Live Demo:

https://personalized-recommendation-system-1.streamlit.app/

---

✨ Features

• Content-based recommendations using item metadata
• Cold-start handling with trending items
• Explainable sections
• Robust image and UI fallback handling
• Modular UI with HTML & CSS
• PostgreSQL-ready backend

---

🧠 Recommendation Logic

• Item similarity computed using TF-IDF + cosine similarity
• User interactions influence ranking over time
• Balanced relevance and diversity to avoid repetition

---

🗂 Project Structure

trendmatrix/
│
├── app.py                  # Streamlit entry point
├── recommender.py          # Recommendation logic
├── db.py                   # Database & user interactions
├── requirements.txt
├── README.md
│
├── data/
│   └── items.csv           # Item metadata
│
├── ui/
│   ├── components.py       # UI helper functions
│   ├── card.html           # HTML card template
│   └── styles.css          # App styling
│
├── assets/
│   ├── logo.png
│   └── images/             # Product images
│
└── .streamlit/
    └── secrets.toml        # App configuration

---

🚀 Run Locally

Clone the repository

git clone https://github.com/Akshu121796/trendmatrix.git
cd trendmatrix

Install dependencies

pip install -r requirements.txt

Run the server

streamlit run app.py

---

🛠 Tech Stack

Python • Streamlit • TF-IDF • PostgreSQL • HTML • CSS

---

If you find this project useful, consider giving it a ⭐ on GitHub.

Thankyou!!!