ChatterNet:

Code and Sample Data for ChatterNet: KDD 2020

Deep Exogenous and Endogenous Influence Combination for Social Chatter Intensity Prediction,
-Subhabrata Dutta, Sarah Masud, Soumen Chakrabarti, Tanmoy Chakraborty
ACM KDD 2020

Paper Link

Dataset:

• Download the Reddit submissions dumps for the months of October, November, and December, 2018, and put them under the directory ChatterNet/Reddit_dumps/Submissions/
• Download the Reddit comments for the months of October, November, and December, 2018, and put them under the directory ChatterNet/Reddit_dumps/Comments/
• Download the news articles, extract, and put them under the directory ChatterNet/News_articles/.

Corpus preprocessing:

• Codes under the directory ChatterNet/Data_processing/’preprocess the raw data into intermediate representations and finally numpy arrays to train the model.

Training and testing ChatterNet:

Codes under the directory ChatterNet/Model_codes/ are used for training and testing the model.

Baselines:

The instructions for running the baselines is in the README of the baseline folder.