Nadejda Komendantova participates at several sessions dedicated to artificial intelligence tools at the European Geophysical Union (EGU) Annual Meeting which is the largest and the most significant annual event in the area of disaster risk reduction. Prof. Komendantova is the convener of the session NH9.15 ~ Positive and negative impacts of societies on natural hazards and risks: from shaping their occurrence to improving risk governance via citizen-centered approaches https://meetingorganizer.copernicus.org/EGU24/session/48761 and provides several talks and presentations including:
1. Cyber-Echoes of Climate Crisis: Unraveling Anthropogenic Climate Change Narratives on Social Media
Social media platforms have a key role in spreading narratives about climate change, and therefore it is crucial to understand the discussion about climate change in social media. The discussion on anthropogenic climate change in general, and on social media specifically, has multiple different narratives. Understanding of the discourses can assist efforts of mitigation, adaptation, and policy measures development. In this work, we collected 333,635 tweets in English about anthropogenic climate change. We used Natural Language Processing (NLP) and machine learning methods to embed the semantic meaning of the tweets into vectors, cluster the tweets, and analyze the results. We clustered the tweets into four clusters that correspond to four narratives in the discussion. Analyzing the behavioral dynamics of each cluster revealed that the clusters focus on the discussion of whether climate change is caused by humans or not, scientific arguments, policy, and conspiracy. The research results can serve as input for media policy and awareness-raising measures on climate change mitigation and adaptation policies, and facilitating future communications related to climate change.
How to cite: Yosipof, A., Elroy, O., and Komendantova, N.: Cyber-Echoes of Climate Crisis: Unraveling Anthropogenic Climate Change Narratives on Social Media, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16779, https://doi.org/10.5194/egusphere-egu24-16779, 2024.
2. Intention to apply Artificial Intelligence using fact checking tools in disaster management
The daily dissemination of a substantial amount of information concerning to disasters and crises on social media platforms, including Facebook, Instagram and Twitter in one side, and the sensitivity of this information, on the other hand, underscores the importance of evaluating the credibility of online information in this area. Fact-checking tools employing artificial intelligence represent a novel approach to verifying the validity of online information across various fields, including disaster management. The inclination of individuals to utilize fact-checking tools in such circumstances is influenced by their perceptions. Although there is a limited studies on the impact of perceptions and information processing on the intention to employ fact-checking tools in disaster-related contexts, it is anticipated that factors like critical thinking, as a concept that involves meticulous assessment of unclear or requiring careful consideration, heuristic processing, a concept indicating acceptance of news content without filtering, and the new-source tracking a concept demonstrating openness and positivity towards social media information, play pivotal roles in predicting this intention. Consequently, a conceptual framework was formulated wherein critical thinking, aside from its direct impact on the intention to use fact-checking tools, also exerts influence through two mediators of information processing and the new source tracking variables. This study's framework was examined using data from 202 respondents across various European countries, collected through an online survey. The conceptual framework analysised utilizing AMOS software. Descriptive findings indicate a moderate level of familiarity with misinformation detection tools among respondents (M=2.65; sd=1.04). Respondents exhibited close knowledge levels regarding fact-checking tools such as Rbutr, Foller, me and Botometer, Fakespot, NewsGuard, and Greek Hoaxes Detector, ranging between approximately (1.57-1.70). Contrary to initial expectations, the study's results reveal that critical thinking, was unable to directly predict the intention to use fact-checking tools. However, the indirect effect of critical thinking was confirmed through the two mediators of new source tracking and information processing (heuristic processing). Critical thinking significantly influenced the new source tracking (β=.49; p <0.0001) and heuristic processing (β=.41; p <0.0001). Both new source tracking (β=.19; p=0.043) and heuristic information processing (β=.31; p=0.001) emerged as direct predictors of the intention to use fact-checking tools. The evidence examined in this study provides empirical support that the conceptual framework has been able to predict 22% of the changes in the intention to use fact checking tools and still a significant amount of it needs to be researched.
How to cite: Zobeidi, T. and Komendantova, N.: Intention to apply Artificial Intelligence using fact checking tools in disaster management, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19362, https://doi.org/10.5194/egusphere-egu24-19362, 2024.
3. Participation in climate change adaptation
This paper presents a comprehensive analysis of participatory elements in climate change adaptation policies at both the EU level and within national contexts, focusing on Germany and Spain. The study delves into the crucial role of co-production and citizen engagement in shaping effective climate adaptation strategies. The research methodology involves policy identification and selection, analysis of participatory elements, and the application of Arnstein's ladder of citizen participation to evaluate the level of citizen engagement in the identified policies. The analysis reveals the diverse mechanisms and approaches employed to foster inclusive and participatory processes in climate adaptation policies. The study highlights the significance of stakeholder involvement, consultation mechanisms, transparency, capacity building, and feedback mechanisms in shaping robust climate adaptation strategies. Furthermore, it underscores the importance of citizen participation in driving transformative climate adaptation initiatives, emphasizing the need for broad geographical representation, inclusive approaches, and the integration of diverse knowledge systems. The study identifies gaps and areas for improvement in the participatory elements of the analyzed policies, emphasizing the need for more comprehensive mechanisms to engage the general public and vulnerable communities in the adaptation planning process. It also underscores the importance of systematic studies of gaps and barriers to stakeholder participation and the representation of marginalized communities in adaptation planning and decision-making processes. The paper offers valuable insights into the participatory elements of climate change adaptation policies, providing a nuanced understanding of the approaches employed at both the EU and national levels. The findings contribute to the ongoing discourse on inclusive and effective climate adaptation strategies, emphasizing the need for continuous improvement and the meaningful involvement of diverse stakeholders in shaping resilient climate futures.
How to cite: Komendantova, N. and Erokhin, D.: Participation in climate change adaptation , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10811, https://doi.org/10.5194/egusphere-egu24-10811, 2024.
4. Persistance of Rumours and Hate Speech Over Years: the Manchester Arena Bombing
Following the 2017 Manchester Arena bombing, the ensuing discussions in the media and on social platforms highlighted the potential of terrorism to deepen societal divisions. This study investigates the dynamics of rumors on social media and in the press after the attack, as well as the subsequent discourse on migration policies. We compiled a dataset comprising 3,184 press articles and 89,148 tweets pertaining to the Manchester Arena bombing. The research aims to identify prevalent rumors, assess their short- and long-term effects on user engagement, analyze the sentiment in tweets related to each rumor, and scrutinize perceptions of terrorism threats and migration policies among both the press and Twitter users.
The findings reveal that Twitter acted as an echo chamber for misinformation, amplifying specific rumors related to the attack, while the press demonstrated fact-checking practices and offered nuanced perspectives. Notably, one rumor suggesting the attacker was a refugee gained traction over time, reflecting a surge in anti-immigrant sentiments. Emotional responses on Twitter varied from a neutral tone to heightened distress and anger, underscoring the significant impact of social media narratives on public sentiment. The research highlights the polarization of views on social media, influenced by the concise format of tweets and the rapid production cycle, with Twitter users predominantly expressing very negative attitudes toward immigration. This study emphasizes the crucial role of the media in dispelling misinformation and cultivating a nuanced public understanding in complex socio-political contexts.
How to cite: Vicari, R., Elroy, O., Komendantova, N., and Yosipof, A.: Persistence of Rumours and Hate Speech Over the Years: the Manchester Arena Bombing , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19300, https://doi.org/10.5194/egusphere-egu24-19300, 2024.