Research Projects

Research Project Leader: Mgr. Miroslav Kubát, Ph.D.

Research Project 1 focuses on the linguistic analysis of online media texts. We are analyzing and interpreting the mechanisms behind the creation of disinformation on various levels (linguistic, emotional and socio-semantic), both in Czech fake news (RP1.1) and in selected foreign languages (RP1.2). Building on the current state of linguistic research, we are applying a diverse range of indicators from stylometry and qualitative discourse analysis.

Research Objectives:

• To gain an understanding of how linguistic resources are used in the creation of disinformation in both Czech and other languages.
• To compile corpora of texts presenting information and disinformation.
• To analyze the compiled corpora using methodological tools from quantitative and qualitative discourse analysis.
• To identify quantitative indices capable of distinguishing between Czech texts presenting information and disinformation.
• To identify characteristic verbal and nonverbal means of disinformation texts in various languages.
  
  

Miroslav Kubát
is an Assistant Professor at the Department of Czech Language, Faculty of Arts, University of Ostrava. His research focuses on quantitative linguistics, especially stylometry. He has led several grant-funded research projects and is a co-author of the  QUITA software. He has published the results of his work in a number of prestigious international journals (e.g. DSH, JQL, CLLT). He has completed research stays at the universities of Athens, Trier, and Padua. He is a member of the international scientific associations IQLA and AELINCO, and he serves on the editorial board of the journal Glottometrics.

Research Project Leader: Mgr. Josef Kundrát, Ph.D.

Research Project 2 examines trust in multimedia content generated by large language models (LLMs) from the perspectives of psychology, sociology, and media studies. Through experimental research, we will observe how respondents react to artificial intelligence (AI)-generated materials and the level of trust they attribute to them. Simultaneously, we will explore the strategies used by multimedia content creators when handling AI-generated content in relation to its credibility. The research combines a questionnaire survey on a representative sample of the population, thematic analysis of factual content in the Czech and Slovak online environments, and in-depth interviews with content creators. The innovation of this research lies in linking AI-generated content with the intent of its creator.

Josef Kundrát
is an educational psychologist at the Department of Psychology, Faculty of Arts, University of Ostrava. His research interests include trust in algorithms and artificial intelligence, the use of physiological signals to measure trust, metaphorical mapping, and embodied cognition. Recently, Dr. Kundrát was awarded a Fulbright-Masaryk Scholarship at the University of Connecticut, where he established collaboration with international experts in the development of objective biomarkers of trust.

Research Project Leader: Ing. Mgr. Jaroslav Hlinka, Ph.D.

As part of this research, we explore modern ways to detect disinformation, especially on social media. We aim to place individual misleading posts into the broader picture of ongoing disinformation campaigns, understand how such content spreads, and develop strategies for prevention and early response. In addition to traditional methods like fact-checking and language analysis, advanced computer science and AI techniques -particularly large language models (LLMs)- are becoming increasingly important due to their strong ability to understand text meaning.

The project (VZ3) is divided into three connected parts:

1. Developing more efficient ways to use LLMs, such as simplifying their structure or breaking them into smaller parts. This will allow us to analyze large amounts of text more affordably, compared to using very large models like ChatGPT (VZ3.1);
2. Studying how disinformation spreads by looking at the structure and dynamics of the networks it moves through, and using methods that help uncover possible cause-effect relationships (VZ3.2);
3. Creating tools to analyze social media content using LLMs, with a focus on Central Europe. These tools will be combined with techniques that track how messages spread and where they come from, resulting in a complete system for detecting disinformation. A public web app will also be developed (VZ3.3).
   
   

Jaroslav Hlinka
is Deputy Director and Chair of the Board at the Institute of Computer Science, Czech Academy of Sciences. He leads the Department of Complex Systems, focusing on data-driven analysis of complex systems, especially the human brain. His work includes the development and application of methods for modeling and predicting the behavior of real-world systems. He has lead multiple interdisciplinary projects and has authored over 150 scientific publications.

Research Project Leader: doc. RNDr. Michal Holčapek, Ph.D.

As part of our research, we focus on the analysis of disinformation from the perspectives of logic, argumentation theory, and epistemology. Our goal is to develop new approaches and methodologies within these disciplines and apply them to strategies for combating disinformation. The research concentrates on the creation of innovative logical systems and methods that allow us to describe and represent the properties of (dis)information exhibiting various types of imperfections—such as inconsistency, vagueness, or incompleteness—and to integrate them within a unified formal framework. Furthermore, we will focus on the analysis of truth values in problematic statements containing natural language quantitative determiners, using fuzzy quantifiers and related methods of logical reasoning. An important part of our research also involves examining the mechanisms and argumentative strategies employed by disinformers in spreading manipulative messages, and exploring possibilities for limiting them through artificial intelligence and machine learning (AI). The research on logical systems and the modeling of linguistic determiners will utilize the analysis of the linguistic features of disinformation texts. At the same time, insights into trust in technology and artificial intelligence, as well as research into the information behavior of people living in segregated communities, will be incorporated into the study of disinformation mechanisms. The methods developed will provide tools for working with language models and enable detailed study of the spread of disinformation in the media space.

Michal Holčapek
is an associate professor at the University of Ostrava and the director of the Institute for Research and Applications of Fuzzy Modeling (IRAFM). He specializes in the field of fuzzy (intermediate) quantifiers, aggregation operators, mathematical modeling, and generalized measure theory. He has participated in a number of research grants, including GAČR-funded projects focused on time series analysis and modeling, as well as the application of aggregation operators in data analysis. He is an area editor of the journal Fuzzy Sets and Systems and the author of more than 80 publications indexed in the Web of Science.

Research Project Leader: doc. Mgr. et Mgr. Kateřina Mikulcová, Ph.D.

Segregated neighborhoods are a very specific context for the spread of (dis)information, because information marginalization and information poverty resulting from the structural inequalities of contemporary society are very strongly manifested here. As part of the research, we will focus on the development of social work interventions through understanding and describing the strategies of information behavior of people living in segregated neighborhoods in the Czech Republic. The research is based on the hypothesis that the social environment in the form of neighborhood contexts has an impact on the life chances and strategies of individuals. In order to completely eliminate or at least mitigate the mechanisms of information poverty in segregated neighborhoods, it is necessary to act on a dynamic set of factors at various levels. By combining quantitative and qualitative research based primarily on the co-creation of knowledge, the research will therefore focus on two specific sub-aims: 1) to reveal the mechanisms of information behavior in segregated neighborhoods; 2) to determine the role of social work in the cultural policy of the neoliberal state in relation to people living in segregated neighborhoods; and what impacts this policy has on their information behavior. The application result of the research project will be methodological material and professional texts freely available to the (professional) public.

Kateřina Mikulcová
works at the University of Ostrava's Faculty of Social Studies. She is the Director of the European Research Institute for Social Work (ERIS). Her long-term research focus centres on the life situations of people living in social exclusion and segregated neighborhoods. In her research, she collaborates with a range of experts from other countries, particularly Finland, the UK and Poland. She has been (and is currently) the main investigator or co-investigator in several grant projects funded by the Technology Agency of the Czech Republic and the Czech Science Foundation. She publishes regularly in leading international journals.