LegalTech

Automatic Segmentation of Czech Court Decisions into Multi-Paragraph Parts

Jakub Harašta

Jaromír Šavelka

František Kasl

Jakub Míšek

Category:

Articles

Region:

Czech Republic

Field of law:

LegalTech, E-Justice

Citation: Jakub Harašta / Jaromír Šavelka / František Kasl / Jakub Míšek, Automatic Segmentation of Czech Court Decisions into Multi-Paragraph Parts, in: Jusletter IT 23. Mai 2019

The authors describe a tool for automatic segmentation of the Czech top-tier court decisions (Supreme Court, Supreme Administrative Court, and Constitutional Court) into multi-paragraph parts. The tool allows segmenting a decision into Header, Party Response, Proceeding Summary, Court Argumentation, Footer, Dissent, and Footnotes. Segmenting text into multi-paragraph parts allows to treat different parts differently even when they contain similar linguistic or other features. Eventually, this is useful in data processing pipelines, as this tool is planned for use in automatic reference recognition purposes.

Table of contents

1. Introduction
2. Related Work
3. Task
4. Dataset
5. Results and Discussion
- 5.1. Automatic Segmentation
- 5.2. Baselines
- 5.3. Results
6. Future Work and Conclusion
- 6.1. Amount of Annotated Documents
- 6.2. Reference Recognition Pipeline
7. Contribution Statement & Acknowledgment
8. References

Please log in to read the full text.