How to run

1. Download image dataset

• This repo has already downloaded Anime Recommendations Database¹ from Kaggle
• But still need to download images dataset for running
  • Download link : anime_images.
  • Create a new folder named "anime_images" below ./client/public/.
  • Move anime_images.zip to ./client/public/anime_images and unzip.

2. Back-end server (Flask)

# Anaconda Powershell Prompt
$ cd server
$ conda create --prefix ./.conda python=3.11 --file requirements.txt
$ conda activate ./.conda  
(env)$ conda install -c conda-forge scikit-surprise
(env)$ flask run --port=5001 --debug

If you use pip to install scikit-surprise, you may encounter an error that requires Microsoft Visual C++ build tools 14.0 or higher version.

This hybrid algorithm needs at least 2GB RAM to run, otherwise Flask would crash.

3. Front-end server (Vue3.js)

$ cd client
$ npm install
$ npm run dev

Please ensure that your Node.js version is higher than v18.20.2(LTS).

4. Deployment

• Front-end: localhost:3000
• Back-end: localhost:5001

Reference

[1]“Anime Recommendations Database,” Kaggle, Dec. 21, 2016. https://www.kaggle.com/datasets/CooperUnion/anime-recommendations-database