Annotated Gun Control/Rights Tweets
Authors: Adrian Benton, Mark Dredze <mdredze@cs.jhu.edu>
Date: 2019/02/22

The attached data file contains annotated tweets from the paper:

Adrian Benton, Mark Dredze. Using Author Embeddings to Improve Tweet Stance Classification. EMNLP Workshop on Noisy User-generated Text (W-NUT), 2018.

http://www.cs.jhu.edu/~mdredze/publications/2018_benton_wnut.pdf

guns_tweets.tsv is a tab separated file with the following columns:

status_id- the id of the tweet
label_[1=control]_[-1=rights] -- 1 means an automatically tagged gun control tweet, -1 means gun rights
fold_[replication=*]_[num_train=*]

fold_* marks which train/dev/test fold the status ID belongs to for each replication and number of training examples.  We reported results averaged over 5 replications with number of training examples varied in {50, 100, 500, 1000}.  

For further details see the paper. Please cite the paper when using this data.