Optimalizace srozumitelnosti a sémantické přesnosti v právních textech

• Thesis title in Czech: Optimalizace srozumitelnosti a sémantické přesnosti v právních textech
• Thesis title in English: Optimizing comprehensibility and semantic accuracy in legal texts
• Key words: legislativní doména|srozumitelnost textu|významová přesnost
• English key words: legal domain|text comprehensibility|semantic accuracy
• Academic year of topic announcement: 2024/2025
• Thesis type: dissertation
• Thesis language: čeština
• Department: Institute of Formal and Applied Linguistics (32-UFAL)
• Supervisor: doc. Mgr. Barbora Vidová Hladká, Ph.D.
• Author: hidden - assigned and confirmed by the Study Dept.
• Date of registration: 09.08.2024
• Date of assignment: 09.08.2024
• Confirmed by Study dept. on: 01.10.2024
• Advisors: Mgr. et Mgr. Ondřej Dušek, Ph.D.

Guidelines

Comprehensibility of legal text has long been an important topic in legal theory – especially in the context of legal interpretation. It is a cornerstone of proper application of law and, by extension, of the rule of law. However, the increasing complexity of society necessitates increasingly complex laws and derived legal texts (including case law, contracts etc.). It is a difficult task of legislators and other authors of legal texts to find the optimum (sweet spot) in the trade-off between comprehensibility and the level of detail of their texts, so that it is optimally readable by their intended target audience. The aim of this thesis is to develop a robust methodology for measuring these opposing qualities of legal texts using latest Natural Language Processing techniques and thus provide the tools for objective assessment of clarity and precision of legal text.

There is extensive (and ongoing) research in both areas. Semantic accuracy metrics are mostly based on Natural Language Inference, however, the problem is not solved and different contexts demand different approaches. As for comprehensibility; most metrics are based on linguistic and semantic properties of the text, some take into account behavioral and cognitive research. And, again, no single metric prevails.

References

- Celikyilmaz, Asli, Elizabeth Clark, and Jianfeng Gao. 2020. Evaluation of text generation: A survey. CoRR, abs/2006.14799., http://arxiv.org/abs/2006.14799
- Chen, Yanran and Steffen Eger, MENLI: Robust Evaluation Metrics from Natural Language Inference, 2023, doi: https://doi.org/10.48550/arXiv.2208.07316
- Choi, Jonathan H.. Measuring Clarity in Legal Text, , The University of Chicago Law Review, 2022, https://lawreview.uchicago.edu/print-archive/measuring-clarity-legal-text
- Dušek, Ondřej and Zdeněk Kasner, Evaluating Semantic Accuracy of Data-to-Text Generation with Natural Language Inference, 2020, Proceedings of the 13th International Conference on Natural Language Generation (INLG), doi: https://doi.org/10.48550/arXiv.2011.10819
- Katz, Daniel Martin, Dirk Hartung, Lauritz Gerlach, Abhik Jana, Michael J. Bommarito II. Natural Language Processing in the Legal Domain. arXiv, Feb. 2023. doi: https://doi.org/10.48550/arXiv.2302.12039
- Keller, T.A., Mason, R.A., Legg, A.E. et al. The neural and cognitive basis of expository text comprehension. npj Sci. Learn. 9, 21 (2024). https://doi.org/10.1038/s41539-024-00232-y
- Sai, Ananya B. and Akash Kumar Mohankumar and Mitesh M. Khapra, A Survey of Evaluation Metrics Used for NLG Systems, 2022, doi: https://doi.org/10.48550/arXiv.2008.12009
- Scott A. Crossley, Text readability and intuitive simplification: A comparison of readability formulas, Reading in a Foreign Language, 2011, doi: 10125/66657
- Sichelman, Ted M., Quantifying Legal Entropy (June 17, 2021). 9 Frontiers in Physics 665054 (2021), Available at SSRN: https://ssrn.com/abstract=2293015