SANDS

This is an annonymous repository containing code and data necessary to reproduce the results published in "Semi-supervised Stance Detection of Tweets Via Distant Network Supervision" for the proposed method SANDS and compared baselines.

Steps to run SANDS

1. Install the required packages mentioned here.
2. Download and extract the data and place under the working directory
3. Change directory to SANDS/SANDS/codes and run 'python3 run_model.py $dataname $splitsize $numclasses' where $dataname can be either INDIA or USA, $splitsize among 500, 1000, and, 1500. $numclasses currently support 5 and 7 for USA and INDIA arguments, respectively.