Izv. prof. dr. sc. Ana Meštrović sudjelovat će na konferenciji Women in Data Science CEE s kratkim predavanjem pod nazivom „Data Science and Infodemic: COVID-19 related information spreading in social media“ u okviru kojeg će prezentirati istraživanje na HRZZ projektu InfoCoV.

Konferencija će se održati 8. ožujka 2021. na platformi WEBEX.

Sve informacije o WiDS CEE dostupne su na službenim stranicama, a registrirati se možete putem sljedeće poveznice.

Sažetak predavanja:

Social media platforms such as Twitter, YouTube, Reddit and Facebook accelerate information diffusion and amplify misinformation spreading. This trend is reflected in the infodemic, which makes it difficult for individuals to find reliable sources of information when they need it. Infodemic and misinformation are especially dangerous during the crisis, such as COVID-19 pandemic. The recent trends in NLP, deep neural networks, combined with social networks analysis are promising approaches for automatic detection of the fake news and for automatic classification of positive, neutral and negative attitudes in social media. Thus, data science may help in alleviating the infodemic and improve various aspects of crisis communication.

The aim of the project “Multilayer Framework for the Information Spreading Characterization in Social Media during the COVID-19 Crisis - InfoCoV”, funded by the Croatian Science Foundation is the characterisation of the information spreading in social media. We propose a multilayer framework that integrates three aspects of text message in social media: (i) content, (ii) context and (iii) dynamic. In the first phase of the project, we collected tweets in the Croatian language related to COVID-19 posted on Twitter in 2020 and performed quantitative analysis. In this talk, we will also present how we studied patterns of retweeting and analysed how trends of positive, neutral and negative attitudes in messages were changing over time.