raihan2108/deep-diffuse

DeepDiffuse:Predicting the ‘Who’ and ‘When’ in Cascades

The twitter dataset is given. For complete version of the datasets please see the following papers

1. The Simple Rules of Social Contagion
2. Meme-tracking and the Dynamics of the News Cycle

If you use the code please cite: DeepDiffuse:Predicting the ‘Who’ and ‘When’ in Cascades ICDM 2018

@inproceedings{islam2018deepdiffuse,
  title={DeepDiffuse: Predicting the `Who'and `When' in Cascades},
  author={Islam, Mohammad Raihanul and Muthiah, Sathappan and Adhikari, Bijaya and Prakash, B Aditya and Ramakrishnan, Naren},
  booktitle={2018 IEEE International Conference on Data Mining (ICDM)},
  pages={1055--1060},
  year={2018},
  organization={IEEE}
}

deep-diffuse's Issues

About the label in test set and validation set

Time loss 'Nan' Issue with `intensity` parameter

When we use intensity as the time loss parameter, we get NaN values in the loss. Any workaround for this issue?

Metrics library missing

Missing references to metrics.portfolio?

Some confusion about data sets

Dear author,Hello!
I'm a little confused about data sets.Can you explain the composition of the training set for me?
The first column of data in the training set is the text number spread by Twitter, followed by the infected user ID and the infected timestamp, right?In addition, what is the meaning of the first column and the second column in the graph.txt file in the data set, and how did this graph file come from, and what is the role of this file?
I hope to get your help. Thank you very much!