Simple Movies Recommendation System

Description

This project is a content-based movie recommendation system that uses natural language processing (NLP) with TFIDF to recommend movies to users based on their comments and preferences. The system utilizes a dataset of movies, including their descriptions and genres, to understand user preferences and suggest similar movies that they might like.

Features

• Load and preprocess movie data from a CSV file.
• Utilize TF-IDF (Term Frequency-Inverse Document Frequency) for processing movie descriptions.
• Filter out common stopwords to improve the quality of NLP analysis.
• Analyze user comments to determine preferences for positive and negative sentiments in movies.
• Recommend movies based on user likes and dislikes using content similarity.

Installation

1. Clone the repository

git clone https://github.com/drissbri/movies-recommender
cd movies-recommender

2. Install required libraries

Ensure you have Python installed on your system. Then, install the required libraries using pip:

pip install -r requirements.txt

3. Prepare Data

• Place your netflix_titles.csv in the movie_data directory (there is sample data there already).
• Ensure you have stopwords.txt, positif_words.txt, and negatif_words.txt in the nlp_data directory (there is sample data there already).

Usage

1. Run the Recommendation System test

python teat.py

2. Interact with the System

• The system will process the dataset and user comments to recommend a movie.
• User interactions are simulated in the code with comments and preferences.

Project Structure

• movies-recommender/nlp.py - Contains the NLP processing functionality.
• movies-recommender/movie.py - Defines the Movie class.
• movies-recommender/user.py - Defines the User class.
• movies-recommender/comment.py - Defines the Comment class.
• test.py - The main test script that runs the recommendation system.
• movie_data/ - Directory containing the movie dataset.
• nlp_data/ - Directory containing NLP-related data, such as stopwords and sentiment words.

Contributing

Feel free to fork the project and submit pull requests.