UppsalaNLP/swedish-causality-ranking

Swedish Causality Ranking Dataset

Causality ranking dataset for Swedish text, comparing sentence pairs on how well they match a causal prompt.

Dataset Description

This dataset contains pairs of Swedish sentences annotated on a 6-point scale for their causal relevance to a given prompt containing a cause or effect.

Fields

• —prompt: Query containing a cause or effect (e.g., "Verkan: växthuseffekt")
• —sentence_1_left_context: Context before sentence 1
• —sentence_1_target: First target sentence
• —sentence_1_right_context: Context after sentence 1
• —sentence_2_left_context: Context before sentence 2
• —sentence_2_target: Second target sentence
• —sentence_2_right_context: Context after sentence 2
• —annotation: Ranking score (1-6 scale)

Annotation Scale

The 6-point scale compares which sentence better expresses a causal relation matching the prompt:

• —Lower scores: Sentence 1 is more relevant
• —Higher scores: Sentence 2 is more relevant
• —Middle scores: Both sentences are similarly relevant

Usage

from datasets import load_dataset

dataset = load_dataset("UppsalaNLP/swedish-causality-ranking")

# Access the data
data = dataset["train"]

# Example
print(data[0]["prompt"])
print(data[0]["sentence_1_target"])
print(data[0]["sentence_2_target"])
print(data[0]["annotation"])

Source

Text extracted from the SOU-corpus (Swedish Government Official Reports).

Citation

@inproceedings{durlich-etal-2022-cause,
    title = "Cause and Effect in Governmental Reports: Two Data Sets for Causality Detection in Swedish",
    author = "D{\"u}rlich, Luise and Reimann, Sebastian and Finnveden, Gustav and Nivre, Joakim and Stymne, Sara",
    booktitle = "Proceedings of the First Workshop on Natural Language Processing for Political Sciences",
    month = jun,
    year = "2022",
    address = "Marseilles, France"
}

License

This dataset is licensed under CC BY 4.0.