Temporal Framing: Multilingual News Articles Annotated for Persuasive Language

MBZUAI's Temporal Framing dataset is a sentence-level corpus introduced in an ACL 2026 paper. It contains 458 English and German news articles, with 2,365 sentences annotated for persuasive temporal framing. The dataset is designed to support research on temporal framing detection and classification across languages.

Use Cases

• Detecting persuasive temporal framing in news sentences based on the annotation scheme.
• Training classifiers to identify temporal framing categories across English and German text.
• Analyzing linguistic patterns of persuasion in multilingual news media.
• Conducting comparative studies on temporal framing strategies between languages.

Strengths

• Contains 458 news articles, providing a foundation for analysis.
• Includes 2,365 annotated sentences at the sentence level.
• Supports bilingual research with English and German content.
• Associated with peer-reviewed research from ACL 2026.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

MBZUAI

Collection Method

Likely manually annotated news articles, as described in the accompanying ACL 2026 paper.

Freshness

Last updated 2026-04-17 21:18:16; freshness should be verified.