Computational techniques to analyse court decisions
Court decisions are commonly studied in legal education by means of manually reading, interpreting and comparing cases. In legal academia and legal practice, computational techniques have emerged that assist scholars and practitioners in analysing court decisions. This development is nonexistent, barely visible at best, in legal education in the Netherlands. Consequently, students (current and future) are not provided with the 21st-century skills they need to enter the (legal) labour market.
In order to close the gap, this project has developed software that analyses and visualizes how court progress decisions relate and evolve over time, providing students with an opportunity to navigate through the body of case law, to identify clusters of decisions and to find patterns that remain unknown when applying traditional methodologies. An educational software platform, based on a prototype that was developed for research purposes, will be built that can be integrated into existing legal courses.
Collective analysis and visualisation of content
The status quo of legal education presents at least two problems. Firstly, most law students will not be exposed to technology other than traditional legal resource databases such as Kluwer Navigator (https://www.navigator.nl), Legal Intelligence (https://www.legalintelligence.com) and Rechtspraak (https://www.rechtspraak.nl), which allow browsing and simple keyword search to find and filter legal documents. However, they do not support collective analysis and visualisation of the content of these documents that would further assist students in legal problem solving. Secondly, the current legal curriculum does not ensure that law students learn essential knowledge about the computational techniques that are applied in new legal technologies.
The platform is based on network analysis, which has been used to identify relevant precedents. Precedents are (rules in) prior court decisions that are used to decide subsequent cases. Treating court decisions as nodes, and citations to other decisions as edges, in a graph, one can construct a precedent network from this information that allows quantitative measurement of how central precedents are, and secondly what kinds of sub-communities (clusters) in the network can be discerned.
Further development of the software
The results will remain available when the project is completed, and will be implemented in one or more courses at Maastricht University, Faculty of Law. Code and docker images will be further made available on GitHub as much as possible in order to allow others educational institutions to be able to use the technology. The software will be further developed, and more data (e.g. European court decisions) will be added to the data over time. Research activities are funded under the Sectorplan – Digital Legal Studies project, which will run until at least 2026, and possibly beyond.
| Ontwikkelperiode: | 2019-2021 |
|---|
Producten of presentaties over dit project
Comenius Teaching Fellowship 2019
