Video Recommender System

This project implements multiple recommender systems using the KuaiRec dataset.

All scripts must be run from the project root directory.

---

KuaiRec Dataset (Required)

The KuaiRec dataset must be downloaded manually and placed into the KuaiRec/ folder at the project root.

Download the dataset from Google Drive:
https://drive.google.com/drive/folders/1_vhNlfe5G6N1oI1RLb1zQPZujDWd7yx0?usp=sharing

⚠️ Important:
The folder name must be exactly KuaiRec/. Do not rename it.

---

Project Root Structure (Required)

Ensure the project root is structured exactly as follows:

video-recommender/ ├── collaborative_filtering/ ├── content_based_filtering/ ├── dataPreparation/ ├── deepfm_recbole/ ├── sasrec_recbole/ ├── svdpp/ ├── models/ ├── KuaiRec/ ├── advanced_rs.py ├── conventional_rs.py ├── hybrid_recommender.py ├── evaluation.py ├── data_loader.py ├── recommender_ui.py └── README.md

⚠️ Important:
All scripts assume relative paths from the project root.
Do not move or rename directories.

---

KuaiRec Dataset Directory Structure

Inside the KuaiRec/ directory, the following files must be present:

KuaiRec/ ├── item_categories.csv ├── item_daily_features.csv ├── kuairec_caption_category.csv ├── kuairec_caption_category_translated_sample.csv ├── small_matrix.csv ├── small_matrix_100k.csv ├── social_network.csv ├── translation_cache_zh_en.json └── user_features.csv

Missing or renamed files will cause data loading errors.

---

Install Dependencies

Install all required Python packages using:

pip install -r requirements.txt

---

Running the System Run the interactive recommender systems:

python conventional_rs.py

python advanced_rs.py

Run evaluation:

python evaluation.py