Vorträge auf der European Society for Empirical Legal Studies (ESELS) Annual Conference Link to heading
Noch mehr rechtsempirische Vorträge! Ich spreche am 19. und 20. Juni auf der Jahreskonferenz der European Society for Empirical Legal Studies (ESELS) in Toulouse — zweimal.
Am 19. Juni spreche ich über unsere Bemühungen (mit Tilko Swalve) Korpora aller Urteile und Beschlüsse des Bundesgerichtshofs (BGH) in Deutschland zu erstellen. Ich freue mich besonders auf den Vortrag über den UN-Sicherheitsrat am 20. Juni, bei dem ich unser neues Open Access Pre-Print “Words of Power: Introducing a Comprehensive Corpus of UN Security Council Resolutions”, geschrieben mit Lorenzo Gasbarri and Niccolò Ridi, vorstellen darf.
Sehen Sie sich unbedingt auch die Konferenz-Webseite und das Konferenzprogramm an, um andere spannende Vorträge zu entdecken!
Zeitplan Link to heading
Panel | Datum | Zeit | Titel des Vortrags |
---|---|---|---|
2.2 | 19 June | 14:00 – 15:30 | Introducing Corpora of Judgments for the German Federal Court of Justice (Bundesgerichtshof) in Criminal Matters (1950-1999, 2020–2024) and Civil Matters (2000-2024) |
6.1 | 20 June | 16:00 – 17:30 | Words of Power: Introducing a Comprehensive Corpus of UN Security Council Resolutions |
Vortrag zum UN-Sicherheitsrat (UNSC) Link to heading
Words of Power: Introducing a Comprehensive Corpus of UN Security Council Resolutions
The United Nations Security Council is the most influential and powerful of the principal UN organs. Article 24 of the UN Charter assigns the UNSC “primary responsibility for the maintenance of international peace and security”. Chapter VII of the Charter grants the Council unique binding powers, but the UNSC plays a prominent role in all affairs of the UN.
We introduce a novel international legal dataset containing texts and metadata for all resolutions of the United Nations Security Council (UNSC) from resolution 1 (1946) through resolution 2722 (2024). The dataset provides resolution texts in all six official UN languages, draft texts and meeting records in English, as well as extensive metadata, for a grand total of 82 variables.
The computational workflow is engineered end-to-end as a fully automated extract-transform-load (ETL) data pipeline with citation analysis and NLP components, unit tests and complex reporting. The declarative workflow is fault- tolerant, resumable and stores intermediate results in over one hundred individual checkpoints. The underlying code is published open source under the GPL-3.
In an international community founded on the rule of law the activities of the United Nations must be public, transparent and defensible. In these troubled times where the rule-based international order is under attack it is more important than ever to document, understand and promote systematic approaches to international law.
Keywords: United Nations, UN Security Council, Data Engineering, Network Analysis, Natural Language Processing
Vortrag zum Bundesgerichtshof (BGH) Link to heading
Introducing Corpora of Judgments for the German Federal Court of Justice (Bundesgerichtshof) in Criminal Matters (1950-1999, 2020–2024) and Civil Matters (2000-2024)
The Bundesgerichtshof (BGH), the German Federal Court of Justice, is the highest court of appeal in civil and criminal matters in Germany. BGH judgments receive regular and intense scrutiny, but German doctrinal scholars routinely focus only on a single decision or a small set of decisions. Quantitative approaches are often infeasible because of lack of data.
We present two original datasets on the German Federal Court of Justice: 1) a corpus of all 77,892 judgments in civil and criminal matters for the period 2000-2024 with 36 variables plus the internal citation network and 2) a corpus of 36,316 judgments in criminal matters for the period 1950-1999 with 31 variables.
We use state-of-the-art data engineering techniques to construct reproducible extract-transform-load (ETL) pipelines that automatically acquire, clean, test, analyze and document German Federal Court of Justice data from the official BGH database and an older criminal justice dataset provided to the authors. Citations are extracted from judgment full texts, analyzed and provided in GraphML format.
Despite centuries of intense academic engagement with the law, quantitative and empirical approaches remain exceedingly rare in Germany. Lack of data, lack of training, lack of awareness and many other reasons have caused this situation. Solving the data problem is a first step to empowering German lawyers in the use of quantitative methods and may help with an empirical turn.
Keywords: Bundesgerichtshof, Criminal Law, Civil Law, Data Engineering, Network Analysis