Software

Python script

To retrieve tweet data from the Twitter API, we used the snsscrape script, below. In this script, we inserted the keywords for hashtags associated with the user-driven auditng cases, and the date-ranges that news articles and users were known to be talking about the cases.

snscrape twitter-search "[INSERT KEYWORD] since:2021-01-01 until:2021-04-09" > name_file.txt

After using the snsscrape script, we then wrote a python script to request data from the Twitter API. The python script we used to retrieve data from the Twitter API is in this Colab Notebook

Classifier Development

Once we retrieved data from the Twitter API, we used the open-source software LightSide to train and test the classifiers we used to label tweets by their: (1) Relevance, and (2) Division of Labor.

The software can be downloaded at this link: LightSide