This is a pytorch implementation of our prediction model. The method of processing dataset is borrowed from DeepMove:https://github.com/vonfeng/DeepMove.git.
This project contains three parts including: codes,data,results
- python 3.6.6
- torch 0.4.1
- numpy
- matplotlib
main.py and main_batch.py are non-batch(batch=1) and batch version of project; train.py contains the method of processing training dataset; model.py contains the prediction model; sparse_traces.py is the method of preprocessing.
three preprocessed datasets
- foursquare.pk (https://github.com/vonfeng/DeepMove.git.)
- foursquare_2012.pk (original dataset[1])
- gowalla.pk (original dataset[2]).
It contains results and two pretrain models of two datasets as foursquare and foursquare_2012. The result of gowalla will come soon.
Note that two pretrain models are non-batch version
You can directly run Python main.py. You can also directly run Python main_batch.py, which is the batch_version of our project.
the paramameters are listed in the results.
[1]:Yang, D.; Zhang, D.; Zheng, V. W.; Yu, Z. (2014): Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns. IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 45, no. 1, pp. 129–142.
[2]:Cho, E.; Myers, S. A.; Leskovec, J. (2011): Friendship and mobility: user movement in location-based social networks,” Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining, ACM, pp. 1082–1090.