Turkish Commercial Court Rulings for Natural Language Inference

Approximately 33,000 case ruling summaries from Turkish Commercial Courts, compiled for natural language inference tasks. The dataset was created by Turkish-NLI using web scraping methods from open sources. It was last updated in November 2024.

Use Cases

• Train NLI models to classify relationships between legal premises and hypotheses using case summary text.
• Fine-tune language models for Turkish legal domain understanding on summary sections of court rulings.
• Analyze linguistic patterns and reasoning in Turkish commercial law via the textual content of case summaries.

Strengths

• Approximately 33,000 case ruling records.
• Focuses on the summary sections where lawsuit reasons are described.

Limitations

• Dataset size and specific column structure are unknown.
• Limited to summary sections, excluding full case texts and other metadata.
• Potential geographic and legal domain bias, covering only Turkish Commercial Courts.

Provenance

Source

Case files from Turkish Commercial Courts, collected from open sources.

Collection Method

Web scraping.

Freshness

Last updated November 2024.

Geography

Turkey