PEPLER (PErsonalized Prompt Learning for Explainable Recommendation)

Paper

• Lei Li, Yongfeng Zhang, Li Chen. Personalized Prompt Learning for Explainable Recommendation. 2022.

A small unpretrained Transformer version is available at PETER!

A small ecosystem for Recommender Systems-based Natural Language Generation is available at NLG4RS!

Datasets to download

• TripAdvisor Hong Kong
• Amazon Movies & TV
• Yelp 2019

For those who are interested in how to obtain (feature, opinion, template, sentiment) quadruples, please refer to Sentires-Guide.

Usage

Below are examples of how to run PEPLER (continuous prompt, discrete prompt, MF regularization and MLP regularization).

python -u main.py \
--data_path ../TripAdvisor/reviews.pickle \
--index_dir ../TripAdvisor/1/ \
--cuda \
--checkpoint ./tripadvisor/ >> tripadvisor.log

python -u discrete.py \
--data_path ../TripAdvisor/reviews.pickle \
--index_dir ../TripAdvisor/1/ \
--cuda \
--checkpoint ./tripadvisord/ >> tripadvisord.log

python -u reg.py \
--data_path ../TripAdvisor/reviews.pickle \
--index_dir ../TripAdvisor/1/ \
--cuda \
--use_mf \
--checkpoint ./tripadvisormf/ >> tripadvisormf.log

python -u reg.py \
--data_path ../TripAdvisor/reviews.pickle \
--index_dir ../TripAdvisor/1/ \
--cuda \
--rating_reg 1 \
--checkpoint ./tripadvisormlp/ >> tripadvisormlp.log

Code dependencies

• Python 3.6
• PyTorch 1.6

Code reference

• mkultra: Prompt Tuning Toolkit for GPT-2

Citation

@article{2022-PEPLER,
	title={Personalized Prompt Learning for Explainable Recommendation},
	author={Li, Lei and Zhang, Yongfeng and Chen, Li},
	journal={arXiv preprint arXiv:2202.07371},
	year={2022}
}