https://github.com/dgliu/TKDE_KDCRec

Experiments codes for TKDE paper "KDCRec: Knowledge Distillation for Counterfactual Recommendation via Uniform Data"
https://github.com/dgliu/TKDE_KDCRec

Last synced: 19 days ago
JSON representation

Experiments codes for TKDE paper "KDCRec: Knowledge Distillation for Counterfactual Recommendation via Uniform Data"

• Host: GitHub
• URL: https://github.com/dgliu/TKDE_KDCRec
• Owner: dgliu
• License: gpl-3.0
• Created: 2020-12-22T09:04:50.000Z (over 3 years ago)
• Default Branch: master
• Last Pushed: 2020-12-22T09:05:02.000Z (over 3 years ago)
• Last Synced: 2024-02-26T22:33:35.321Z (4 months ago)
• Language: Python
• Size: 3.29 MB
• Stars: 2
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Lists

• Awesome-Knowledge-Distillation - TKDE 2022

README

        
# TKDE_KDCRec

---

Experiments codes for the paper:

Dugang Liu, Pengxiang Cheng, Zinan Lin, Jinwei Luo, Zhenhua Dong, Xiuqiang He, Weike Pan and Zhong Ming. KDCRec: Knowledge Distillation for Counterfactual Recommendation via Uniform Data. Submit to TKDE.

---

## Motivation

This work is an extension of [our SIGIR 2020 paper](https://dgliu.github.io/files/SIGIR20_KDCRec.pdf). The new contents we have added include:

1. We have developed four new strategies to improve and supplement the original strategy.

2. We have included new experimental results and associated analyses, including the evaluation and analysis of new strategies, and the performance trends of the proposed framework on some key factors.

3. We have added more related work and discussions, as well as a formal problem definition.

4. We have made many improvements throughout the whole paper.

---

## Usage

1. You can reproduce the experimental results of the paper by simply running `reproduce_paper_results.py`.

2. If you want to try to perform the complete experiment process, please refer to the `USAGE.txt` file.

---

If you have any issues or ideas, feel free to contact us ().