---

Recommendation System

Overview

This project implements a Recommendation System using data ingestion, preprocessing, and recommendation generation workflows.

---

Features

• Data Ingestion: Featch the data from the given link and store the data in raw folder.
• Data Preprocessing: Cleans and merges data into a unified format for modeling.
• Recommendation Engine: Generates recommendations based on user data and behavior.

---

Key Features:

• Content-based filtering: Recommending videos similar to those the user has viewed or liked.
• Collaborative filtering: Leveraging similar user preferences to enhance recommendations.
• Hybrid models: Combining content-based and collaborative filtering for improved accuracy.
• Include a mechanism to recommend videos for new users without prior interaction history (Based on Tranding videos,most viewed videos,Most liked videos and also Based on mood).

Project Structure

data/
├── preprocessed/        # Preprocessed data files
│   └── merged_data.csv  # Cleaned and merged dataset
├── raw/                 # Raw input datasets
│   ├── get_all.csv
│   ├── inspire.csv
│   ├── liked.csv
│   ├── rating.csv
│   └── view.csv
logs/                    # Logs for debugging
Notebook/
├── merge_data.ipynb     # Data merging notebook
├── Recommendation.ipynb # Recommendation engine notebook
data_ingestion.py        # Script for data ingestion
data_preprocessing.py    # Script for data preprocessing
recommendation.py        # Script for generating recommendations
README.md                # Project documentation
requirements.txt         # Dependencies

---

Setup Instructions

Follow these steps to set up and run the project:

Prerequisites

Ensure the following are installed:

• Python 3.8 or higher
• pip (Python package manager)
• Git

Installation

1. Clone the repository:

   git clone https://github.com/your-username/your-repo.git
   cd your-repo

2. Install dependencies:

   pip install -r requirements.txt

---

Running the Project

1. Local Execution: Alternatively, run scripts locally:

   python data_ingestion.py
   python data_preprocessing.py
   python recommendation.py

2. Results:

   recommendations.json

   • It can give the Recommened videos basis of content and user experience.
   • Include a mechanism to recommend videos for new users without prior interaction history (Based on Tranding videos,most viewed videos,Most liked videos and also Based on mood).

---