Analyzing QAnon on Twitter in Context of US Elections 2020: Analysis of User Messages and Profiles Using VADER and BERT Topic modeling

被引：20

作者：

Anwar, Ahmed ^{[1
]}

Ilyas, Sardar Haider Waseem ^{[1
]}

Yaqub, Ussama ^{[1
]}

Zaman, Salma ^{[1
]}

机构：

[1] Lahore Univ Management Sci, Lahore, Pakistan

来源：

PROCEEDINGS OF THE 22ND ANNUAL INTERNATIONAL CONFERENCE ON DIGITAL GOVERNMENT RESEARCH, DGO 2021 | 2021年

关键词：

QAnon; Twitter; sentiment analysis; topic modeling; BERT; Trump; Biden; US elections 2020;

D O I：

10.1145/3463677.3463718

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this paper we analyze Twitter users and their tweets mentioning "QAnon" in context of US Presidential Elections of 2020. We collect over 12 million tweets for 46 consecutive days starting from August 1st - September 15th 2020 containing the keywords "Trump", "Biden" or "Election2020". We identify users mentioning "QAnon" in their messages and perform sentiment analysis using VADER to evaluate their position towards Trump and Biden. Along with this we create word cloud and perform topic modeling using BERT on user profile descriptions. Some of our key findings contradict the popular notion that people discussing QAnon on social media are mostly located in Republican dominated states of the US. We also discover that an over whelming majority of QAnon tweeters are Donald Trump supporters, are conservative and nationalist having terms like "MAGA", "God", "Patriot" and "WWG1WGA" in their Twitter profile descriptions.

引用

页码：82 / 88

页数：7

共 46 条

[41] Framework for Managing the COVID-19 Infodemic: Methods and Results of an Online, Crowdsourced WHO Technical Consultation [J].