This project implements a recommendation system that employs two main algorithms: the Apriori algorithm for frequent itemset mining and association rule generation, and collaborative filtering for personalized recommendations. By combining these approaches, the system aims to provide accurate and relevant product recommendations to users based on their past interactions and preferences.
The goal of this recommendation system is to enhance user experience and engagement by suggesting products that align with their interests and past purchase history.
- Collaborative Filtering: Utilized to provide personalized recommendations by analyzing user-item interactions and finding similarities between users or items.
- Apriori Algorithm: Used for mining frequent itemsets from transaction data and generating association rules. This helps identify patterns in user behavior and product co-occurrences.
The dataset used in this project was generated manually. It's essential to note that manual generation may introduce biases into the data, potentially affecting the recommendation results. Careful consideration should be given to the interpretation and application of the recommendations generated by this system.
To set up the project, follow these steps:
-
Clone the repository:
git clone https://github.com/Ubeydkhoiri/recommender-system-using-apriori.git # move to the directory cd recommender-system-using-apriori
-
Create your virtual environtment:
python -m venv venv_name # activate your virtual envirotnment venv_name\Scripts\activate
-
Install dependencies:
pip install -r requirements.txt
-
Run the recommendation system:
streamlit run app.py
Below is an example of the recommendation system's output after running app.py
: