Mapping the network of anti-vaxxer and pro-vaxxer supporters’ – Two discourses, opposite worlds? - Kmetty Zoltán előadása

   2021. szeptember 28.

A Társadalomtudományi Kutatóközpont Mesterséges Intelligencia Nemzeti Laboratórium TK MILAB Speaker Series néven indította el online rendezvénysorozatát 2021. tavaszán, melynek során online kutatásbeszámolókat, beszélgetéseket tart a mesterséges intelligencia társadalmi hatásairól. Az előadássorozat 2021 őszén-telén is folytatódik.

A sorozat következő állomása 2021. szeptember 28-án 10:00 órától „Mapping the network of anti-vaxxer and pro-vaxxer supporters’ – Two discourses, opposite worlds?” címmel kerül megrendezésre. 

Előadó:

Kmetty Zoltán (TK CSS-RECENS)

Moderátor:

Ságvári Bence (TK CSS-RECENS)

 

Absztrakt:

Mapping the network of anti-vaxxer and pro-vaxxer supporters’ – Two discourses, opposite worlds?
 
Zoltán Kmetty12 – Eszter Katona12, Krisztián Boros1, Anna Molnár1, Júlia Koltai123, Bence Ságvári1
1. ELKH TK, Budapest, Hungary 2. ELTE TÁTK, Budapest, Hungary 3. CEU, Vienna, Austria
 
The debate about anti-vaccination was renewed in the last year as a consequence of Covid-19 pandemic. For the effective community level immunity against Covid-19, 60-70 percent of the population needs to be vaccined. But the acceptance of Covid-19 vaccines is lower in many countries (Lazarus et al 2020), and the anti-vaxer narratives are louder than ever (Johnson et al 2020).  Standard methods (survey, focus groups) can give an overview about the anti-vaccination trends, but they can’t map the whole dynamic of the process, the main narratives behind anti-vaccination and the communication links and/or barriers between pro- and anti-vaxxer supporters. Social media posts, and comments in the online space could give a more detailed and more valid result about how people think and feel about vaccination. 
In our project (https://milab.tk.hu/en) we use online textual data and advanced text mining methods to analyze the anti-vaxxer discourse in Hungary under the second wave of Covid-19. Using a corpus of more than 1 million posts and comments we classify the texts into neutral/partial/anti-vaxxer groups. For this we use a state-of-the-art NLP language model – BERT – which achieved high performance in many NLP related classification tasks. Within the texts we identify key actors (people, institutions), and in many cases we could also identify who commented on whom. This NLP based approach makes it possible to create the discourse network of pro- and anti-vaxer supporters. In the analyses part of the paper, we focus on this discourse network. Our research question is about the existence of links and barriers between the pro- anti-vaxxer supporters’ network.
 
Reference
Johnson, N. F., Velásquez, N., Restrepo, N. J., Leahy, R., Gabriel, N., El Oud, S., ... & Lupu, Y. (2020). The online competition between pro-and anti-vaccination views. Nature, 1-4.
 
Lazarus, J. V., Ratzan, S. C., Palayew, A., Gostin, L. O., Larson, H. J., Rabin, K., ... & El-Mohandes, A. (2020). A global survey of potential acceptance of a COVID-19 vaccine. Nature medicine, 1-4.
 

Az előadás online, Zoom alkalmazáson kerül megrendezésre, az eseményen való részvétel regisztrációhoz kötött. Kérjük részvételi szándékát jelezze az alábbi űrlap kitöltésével: https://cutt.ly/hEgcM0f

Az esemény linkjét külön e-mailben küldjük a regisztrált résztvevőknek.