Becky Smethurst - “Tricking” social media algorithms to engage the “unreachable” with astrophysics

Описание: Rebecca Smethurst (University of Oxford) - “Tricking” social media recommendation algorithms to engage the “unreachable” with astrophysics

Abstract:
Producing informative science content for an online audience on social media is an excellent way of reaching people worldwide with complex topics. However, the majority of social media platforms, including YouTube, TikTok, Instagram and Twitter, all employ neural networks in order to make content recommendations to their users. These machine learning algorithms use inputs such as age, gender and location in order to make recommendations based on other user’s activity. This results in an insular network of users who engage with science content online. For example, YouTube channels with male presenters are more likely to be watched by other men, and therefore recommended to more men. However, of the top 25 science & technology channels by subscribers on YouTube in 2024, 23 are run by men. This leads to a situation where subsets of the public are being excluded from the scientific conversation online unless they specifically search for that content, e.g. young girls or countries without government space agencies where engagement in traditional media with astronomy content is low. I shall share my experiences producing content for social media, in particular for the astronomy YouTube channel “Dr Becky” with over 700,000 subscribers and 77 million total views, that specifically aims to overcome the constraints set by recommendation algorithms in order to engage the “unreachable” online with astrophysical research in an accessible way.

-----------------------------------------------------

MIST Online Seminar Series: https://www.mist.ac.uk/meetings/mist-...

-----------------------------------------------------

Image credit:
ESA/Hubble