/140_to_280

Repository for the paper “How Constraints Affect Content: The Case of Twitter's Switch from 140 to 280” published at ICWSM’18

How Constraints Affect Content: The Case of Twitter's Switch from 140 to 280 Characters

This repository provides information about the data used in our AAAI ICWSM paper.

Content of the repository

  1. The file user_ids.csv contains user IDs of the hundred thousand sampled Twitter users.
  2. The file tweet_ids.json.zip is a compressed file containing tweet IDs of the complete dataset with entire timelines of sampled users. It can be downloaded from: this link
  3. The folder matched_data contains files with filtered and matched tweets, as described in the paper. Matched pairs of tweets are at the same line, comma-separated. The folder contains the following files:
  • study_1.csv (main study, matched for users)
  • study_2.csv (main study, matched for topics)
  • study_3.csv (control study, matched for users)

Reference

Please cite the paper when using the data:

How Constraints Affect Content: The Case of Twitter's Switch from 140 to 280 Characters, Kristina Gligorić, Ashton Anderson and Robert West.
Proceedings of the 12th International AAAI Conference on Weblogs and Social Media (ICWSM), Stanford, California, June 2018. https://arxiv.org/abs/1804.02318