Miller, Blake Andrew Phillip 
ORCID: 0000-0002-4707-0984
(2016)
Automated detection of Chinese Government astroturfers using network and social metadata.
.
University of Michigan, Michigan, USA.
Abstract
Astroturfing is the practice of an organization communicating a message using fake “grass-roots” sources. Because these messages attempt to mimic ordinary individuals, distinguishing them from real grass-roots messages is a difficult task. In this paper, I present a method for automatically detecting pro-government astroturfers in China (colloquially referred to as the Fifty Cent Party), using comment metadata from a dataset of 70 million news media comments posted on 6 million news articles from 19 popular news websites in China. I estimate that approximately 15% of all comments made on these 19 news websites are made by government astroturfers. This method of comment propaganda detection is automated, and does not require manual human labeling. Instead, data are labeled according to metadata characteristic of the work procedures and behavioral patterns of government astroturfers. Models trained on these metadata predict posts from a leaked dataset of government astroturfers with as high as 94.1% accuracy. This method allows researchers timely access to government astroturfer commentary from China. Additionally, this method allows for prediction of astroturfers’ bureaucratic affiliation using social network data, and can allow researchers to explore variance in how this information control tactic is deployed in different bureaucracies and localities in China. It also suggests a forensic method for detecting astroturfers in different countries and online platforms.
| Item Type: | Monograph (Working Paper) |
|---|---|
| Additional Information: | © 2016 The Author |
| Divisions: | Methodology |
| Subjects: | H Social Sciences > HN Social history and conditions. Social problems. Social reform |
| Date Deposited: | 20 Aug 2019 16:00 |
| Last Modified: | 20 Dec 2024 00:21 |
| URI: | http://eprints.lse.ac.uk/id/eprint/101418 |
Actions (login required)
 |
View Item |
