What question did this study set out to answer?

The aim is to improve legal research accessibility by creating multilingual judicial summaries from Swiss court decisions.

February 28, 2026Open Access

Unlocking Legal Knowledge: A Multilingual Dataset for Judicial Summarization in Switzerland

View Full Paper

Ask AI

Bookmark

Authors

This dataset enhances legal research through multilingual summaries of Swiss court rulings, suggesting improved access to legal information.

The aim is to improve legal research accessibility by creating multilingual judicial summaries from Swiss court decisions.
Introduced the Swiss Landmark Decisions Summarization (SLDS) dataset with 20K rulings in German, French, and Italian.
Fine-tuned various language models including Qwen2.5, Llama 3.2, Phi-3.5.
Compared model performance in legal summarization using an LLM-as-a-Judge framework.
Fine-tuned models showed good lexical similarity in summaries.
Larger models like GPT-4o provided more accurate and coherent legal summaries.
Reasoning-focused models did not consistently outperform others, indicating the importance of factual precision.

Rolshoven et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

This dataset enhances legal research through multilingual summaries of Swiss court rulings, suggesting improved access to legal information.