Skip to content

HondaPL/Recommending-Systems-Project

Repository files navigation

Recommending Systems Project

Content-Based Recommender - Adam Hącia 2022

Table of contents

Achieved results compared to Amazon Recommender

Example screenshot

Project Description

Content-based recommender system that recommends products based on the content of the product description.

Technologies

  • Python 3.9.7
  • Anaconda + Jupyter

Setup

  1. Install Anaconda with Python >= 3.8.
  2. Prepare your conda environment
    • Run the following command: conda create -n recommender python=3.8
    • Activate the conda environment by running: conda activate recommender
  3. Install the required packages
    • Run the following commands: conda install numpy conda install pandas conda install matplotlib conda install seaborn conda install sklearn conda install hyperopt
  4. In Bash type: jupyter notebook The notebook will be opened in your browser.
  5. In Jupyter Notebook open these files and run all cells: project_1_data_preparation project_1_recommender_and_evaluation
  6. Last cell in project_1_recommender_and_evaluation.ipynb should contain the results of the evaluation with HR@10 metric for
    • LinearRegressionCBUIRecommender
    • AmazonRecommender
    • RandomForestCBUIRecommender
    • XGBoostCBUIRecommender

Status

Project is: completed

Contact

Created by @HondaPL