This is a Pytorch implementation of our model:
A novel dynamic recommendation model that focuses on users who have interactions in the past but turn relatively inactive recently i.e. time-sensitive cold-start users.
-
Clone this repository.
git clone https://github.com/ritmininglab/A-Dynamic-Meta-Learning-Model-for-Time-Sensitive-Cold-Start-Recommendations cd A-Dynamic-Meta-Learning-Model-for-Time-Sensitive-Cold-Start-Recommendations
-
Install the following dependencies. The code should run with Pytorch 1.3.1 and newer.
- Pytorch (1.3.x)
- python 3.5 or newer
- scikit-learn
- scipy
- numpy
- pickle
- Go to each folders of datasets to run the corresponding experiments.
- For example
cd Netflix
and - Run
python proposed_model.py
for the Netflix dataset
This code is based on MeLU