Brazilian Court Decision Summaries with Categorical Labels

4,043 court decision summaries (Ementa) from the Tribunal de Justiça de Alagoas (TJAL) in Brazil. The decisions are labeled according to 7 categories and include metadata on whether the decisions were unanimous. It supports the task of Legal Judgment Prediction.

Use Cases

• Train a text classification model on the Ementa summaries to predict one of the 7 legal categories.
• Analyze the relationship between the unanimous decision label and the assigned legal category.
• Use the labeled Ementa summaries for legal judgment prediction research in the Brazilian context.

Strengths

• Contains 4,043 court decision summaries.
• Each decision is labeled with one of 7 legal categories.
• Includes a label indicating whether the decision was unanimous.

Limitations

• The dataset is limited to decisions from a single Brazilian state court (TJAL), which may not generalize nationally.
• The sample size of 4,043 rows is relatively small for training complex machine learning models.
• The specific metadata columns beyond the summary, category, and unanimity label are unknown.

Provenance

Source

Tribunal de Justiça de Alagoas (TJAL), the State Supreme Court of Alagoas, Brazil.

Collection Method

Collection of court decision summaries and their metadata.

Freshness

The dataset was last updated on 2022-09-22.

Geography

State of Alagoas, Brazil.