DuoWikiBias: A Spanish Wikipedia Corpus for Bias Detection

DuoWikiBias is a novel parallel corpus derived from Wikipedia for Spanish bias classification. The dataset, 165.0 KB in size and released under CC-BY-4.0 by Karla Salas-Jimenez, supports research on framing, epistemological, and demographic biases. It was last updated on May 13, 2026.

Use Cases

• Evaluating Large Language Models for bias detection based on the described advanced prompting techniques.
• Training traditional classifiers like logistic regression or SVM using S-BERT embeddings derived from the corpus.
• Benchmarking hybrid approaches that combine LLM-based representation learning with classical machine learning methods.
• Studying framing, epistemological, and demographic biases in Spanish-language text from Wikipedia.

Strengths

• Provides a Spanish-language resource for bias detection, addressing a noted scarcity in NLP.
• Corpus is derived from Wikipedia, a widely used and structured source of text.
• Dataset is released under the permissive CC-BY-4.0 license for open use.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is small at 165.0 KB, indicating a limited scope.

Provenance

Source

Derived from Wikipedia.

Collection Method

Created as a parallel corpus for bias classification.

Freshness

Last updated 2026-05-13 05:34:03; freshness should be verified.