Publikationsdatenbank

Mild adverse events of Sputnik V vaccine extracted from Russian language telegram posts via BERT deep learning model (2021)

Art

Vortrag

Autoren

Semenov, Alexander

Jarynowski, Andrzej (WE 16)

Kaminski, Mikołaj

Belik, Vitaly (WE 16)

Kongress

2021 Informs Annual Meeting

online, 24. – 27.10.2021

Quelle

Informs annual meeting 2021: Anaheim, CA : October 24-27, 2021 — The Institute for Operations Research and the Management Sciences (Hrsg.)

— S. Session VTD53

Sprache

Englisch

Verweise

URL (Volltext): https://www.abstractsonline.com/pp8/?utm_campaign=INFORMS%20eNews&utm_medium=email&_hsmi=154433787&_hsenc=p2ANqtz-9g52i1K8dm49Afz9sReuYNsKjUaOCfCDoj-0oWOCzeJTJ1IXO1Vq5l9t7LTtd8_OjYon_UQGcNCac6LzZuSq7Fpjru7UwZF-S9jffGNwsnn30Vj4E&utm_content=154433787&utm_source=hs_email#!/10390/presentation/7884)

Kontakt

Institut für Veterinär-Epidemiologie und Biometrie

Königsweg 67
14163 Berlin
+49 30 838 56034
epi@vetmed.fu-berlin.de

Abstract / Zusammenfassung

Anti-COVID-19 vaccines effectively preventing severe outcomes are the most promising remedy for the current COVID-19 global health crisis. For many vaccines extensive fact sheets on possible adverse reactions are publicly available, which is hardly the case for Sputnik V (Gam-COVID-Vac) vaccine. We used social media data to fill this gap. To this end we investigated the discourse in a Telegram group with more than 10,000 posts devoted to Sputnik V mild adverse events. We utilized multi-label classification of messages by pre-trained BERT deep neural language model for Russian language, and tuned on the Telegram data to calculate frequencies of mild adverse events. Moreover, we perform stratification of results according to sex, age and for the first and second vaccination dose separately. We also compare the results with published data on other anti-Covid vaccines.