2D Interactive Visualization: Twitter Social Network Analysis to tackle disinformation

1. Interactive Visualizations for Twitter Social Network (Twitter SNA)

I’m sharing several demonstrations of the interactive visualizations for Twitter SNA I developed during my 6-month internship at AFP. To experience these interactive visualizations, you can download the plugin and register an account on their website Verification Plugin.

To help journalists and fact-checkers detect disinformation on Twitter, we used social network analysis to identify abnormal accounts, unusual patterns, and more from complex data using graphics. Several useful factors for detecting abnormalities include posting tweets at odd times, frequent and regular tweeting, repeated content, and active sharing of URLs leading to fake news websites. Our proposal involved combining multiple interactive charts and graphs that complement each other, allowing for in-depth exploration, new insights, intriguing questions, and capturing the full value of the data.

**Bubble chart of top 100 most active accounts**, aimed to detect coordinated behaviors and bots.

**Associated hashtags graph** reveals key discussion topics (hashtags) and their co-occurrence (relationships).

These interactive charts and graphs enable in-depth exploration of insights. By clicking on suspicious spots, specific nodes or connections of interest, zooming in/out of the visualization, etc. …, it can reveal coordinated behavior or bot patterns, spreading fake news strategies, URLs that redirect to fake news websites, and so on.

2. Use cases

Case 1: Analyzing the hashtag #Plandemic

Related to the hashtag #plandemic: Who/What were at the center of discussions on Twitter in May 2020?

Clicking on the most mentioned accounts:

Donald Trump was called for actions (i.e. end the lockdown, stop listening to Gates foundation agenda, etc.), for responses to #plandemic, #qanon, #wwg1wga conspiracies.
YouTube was mentioned to invite other accounts to watch the uploaded videos, called for removing harmful videos,…
Many accounts falsely alleged that Bill Gates made plans for the pandemic and vaccination.
Many accounts thanked @DrJudyAMikovits for her falsely claims and accusations in the 26-minute conspiracy video. Some accounts blamed her for the falsely accusations.

Related to the hashtag #plandemic: Who actively posted/shared tweets on Twitter in May 2020? Which topics/hashtags were they trying to spread? Based on their activities on Twitter, did they seem like bots?

Which hashtags/domains were associated with/attached to the hashtag #plandemic in May 2020?

Socio-semantic graph reveals connections between the entities in the graph with the searched hashtag #plandemics:

Top domain names: twitter.com, youtube, bitchute.com, plandemicmovie.com, banned.video, bit.ly, vimeo.com, facebook.com, instagram.com, vimeo.com…
Top associated hashtags: #covid19, #coronavirus, #qanon, #scamdemic, #wwg1wga, #plandemicdocumentary…

Case 2: Analyzing the hashtag #QAnon

In May 2020, regarding the hashtag #QAnon, who were mentioned, and which hashtags/domains were associated with the tweets?

Clicking on the most mentioned accounts:

@realDonaldTrump and @POTUS usually were mentioned by QAnon supporters. They expressed the Pro-Trump attitude and support him for the re-election.
@YouTube, where variety of spreading falsehood videos were uploaded.

In the socio-semantic graph, the popular URLs refer to plenty of websites which often publish falsehoods, hoaxes, conspiracy theories such as thegatewaypundit.com, breitbart.com, foxnews, OneNewsNow.com, pscp.tv, threadreaderapp.com,…

Regarding the hashtag #QAnon in May 2020, who were actively posting/spreading tweets? What were the contents of those tweets?

My sharing regarding interactive visualizations for Twitter Social Network Analysis comes to a close. It’s essential to supplement the analyses with the official plugin disclaimer:

Disclaimer: this software is provided “as is”, without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose and non infringement. In no event shall the authors or copyright holders be liable for any claim, damages or other liability, whether in an action of contract, tort or otherwise, arising from, out of or in connection with the software or the use or other dealings in the software.
https://weverify.eu/verification-plugin/