I examine this topic through two research projects in collaboration with co-authors. The first project explores data visualization as a political instrument, analyzing how the representation of data can perpetuate or reshape social relations. The second project aims to quantify the environmental impact of research activities. Our objective is that the outcomes of our work will contribute to the development of sustainable scientific practices by encouraging a more deliberate consideration and reduction of the ecological footprint associated with social science research.

I am advancing my doctoral dissertation using data from Transparency International. Our research examines the appearance of corruption risk (measured here as competition restriction) in public procurement bid documents. Our goal is to identify tenders (and later specific companies) where the language clearly favors a particular bidder, thereby excluding other potentially qualified applicants. Currently, I am primarily working with Hungarian texts; secondarily, this is supplemented internationally with Serbian, Bulgarian, French, and Italian data. 

Our research with the Survey Methods Room research group employs natural language processing (NLP) techniques to identify key topics communicated by political parties on social media and transform these into queryable statements. Based on this analysis, we collect survey data to gauge voters' reactions to these topics. The analysis investigates whether voters who consider a party’s social media topics important are more likely to vote for that party. Moreover, as a result of the analysis, we can examine campaign messages shared on social media, providing a descriptive overview of the messages circulating before the elections.