Ongoing scientifical and practical progress in the fields of law, data science and legal tech brought up various needs and issues for different industries and science. German legal courts and research groups involved in legal tech are some of those and the main target group of this thesis. As court rulings are an very important part of the german legal system, this thesis focuses on the tenor of an court ruling. The tenor consists of the legal consequences of an court ruling split up into four particular parts. The goal of this thesis is to analyze and automatically extract those legal consequences out of a given tenor. Another goal is to validate the hypothesis, that the legal consequences from a given tenor can be used to derive the legal consequences from another unseen tenor. Therefore one has to take the architecture of the german legal court system into account (https://www.bmjv.de/SharedDocs/Archiv/Downloads/Schaubild_Gerichtsaufbau_Deutsch.pdf?__blob=publicationFile&v=3).
Research questions
Which teqniques & concepts, from Natural Language Processing in the context of german (semi-) structured text data, can be used to automatically extract legal consequences from a given tenor ?
To what extent can the legal consequences from prior instances be derived by the tenor from the last instance, and is it required to consider other parts of the court rulings as well to make the derivation more accurate?
| Name | Type | Size | Last Modification | Last Editor |
|---|---|---|---|---|
| m turic_bachelor_thesis_final_presentation.pdf | 8,38 MB | 15.03.2021 | ||
| m turic_bachelor_thesis_kickoff_presentation.pdf | 5,43 MB | 15.03.2021 | ||
| Turic_Mario Bachelor Thesis final.pdf | 2,80 MB | 15.03.2021 |