The aim of the incubator project is to enhance the application of quantitative text analysis and text mining approaches to political and legal texts. The primary objective of the research network is to provide an interdisciplinary platform for researchers of political, legal and administrative phenomena who use quantitative text analysis as their core methodology. The substantive research agenda of the project consists of both pre-defined work streams and co-operative projects to be discussed and launched at project meetings.

Using artificial intelligence for analyzing of policy topics in political content

The automated coding of political and legal texts for policy topics and preferences is a major development in policy studies. The aim of this stream is foster the machine learning . For more information see the Hungarian branch of the Comparative Agendas Project.

Measuring the quality of government

The topic of good government is a staple of political thought since the antiquity. Studies have attempted to answer this question, international organizations developed indices aiming at particular topics, such as corruption or the easiness of doing business. While useful, they may be easy to manipulate, or are often ridden with ideological biases. Empirical research that are comparable across nations have produced debatable results. In Hungary, there is still a gap in this knowledge. How does the formal quality of government and public policy in post-regime change look like in the country?

Driven by the scarceness of studies covering the quality of government in Hungary since 1990 — and considering that indicators may not take account of all the complexity around what “good government” is—, Miklós Sebők, the principal investigator of the project, laucnhed the “The quality of government and public policy: a Big Data approach”, funded by the Hungarian National Research, Development and Innovation Office.

Dr. Sebők uses Big Data, through quantitative text analysis and text mining, to offset the complexity of measuring legislative quality across time, because the data needed for such an analysis is not readily available. These quantitative methods will bring to light two aspects of policy quality. One is to gauge the stability of pieces of legislation as well executive decrees, or how often they are amended. Another is to ex post evaluate these laws and decrees after entering force.

The comprehensibility of legal texts

Comprehensibility of legal documents, (laws, judgements and other official documents) became an issue with modernity. Since then it is an ongoing complaint, that that legal texts and legal jargon are not understandable for an average citizen, and this must be somehow changed. The research aims to discover the linguistic roots of this complicatedness. Therefore, it takes simple and popular explanatory legal texts from the internet, and compares them with text mining software with “official texts” on the same topic. The initial hypothesis of the research is, that comprehensibility-problem is primarily not a linguistic problem, because the lexicon and the syntactical structure of these texts are not radically different. In the second phase of the research we try to discover, that what are then the differences between the comprehensible and incomprehensible texts.

The use of the law of EU in Hungarian ordinary courts

Hungary is the member of the EU since more than 13 years. There is a database since 2007 which contains a huge sample of the judgements of the ordinary courts, in their original form. The aim of the research is to discover the patterns of “using” the EU law via the citations We will also try to match the changes within the citation patterns with the political turns of the near past. How the change in the overall political environment is reflected in the everyday argumentation of the courts?

Citation patterns of previous decisions in ordinary courts

In 2012 Zsolt Ződi completed a research project on how Hungarian courts are using the previous decisions of the courts in their argumentation. The background assumption was, that each continental system, and also Hungary is slowly converging to a “mild” precedent-system. The final conclusion of the research was, that while quantitatively there is a huge increase in citations to previous (higher) court decisions, the technique of using these decisions is still continental, and not common-law type. The aim of this research is to repeat the research in 2018 and see how the trends are moving, what happened with the Hungarian judicial practice.