Review Arcade: ACL ARR Paper Reviews with Accepted/Rejected Splits

ACL ARR Reviews - Review Arcade Project contains paper reviews from the ACL ARR (Association for Computational Linguistics - Annual Review of Research) program. The dataset is organized into two splits for accepted and rejected papers, with each record stored as a JSONL line. Author G4KMU uploaded the dataset to Hugging Face, with a last recorded update on 2026-04-13.

Use Cases

• Analyze review content and sentiment differences between accepted and rejected papers based on the split structure.
• Train models to predict paper acceptance based on review text features mentioned in the description.
• Study linguistic patterns and argumentation in academic peer review within computational linguistics.
• Benchmark fairness or bias detection algorithms using the labeled acceptance outcomes.

Strengths

• Provides a structured split between accepted and rejected paper reviews, enabling comparative analysis.
• Records follow a defined JSONL schema, suggesting a consistent format for parsing.

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

Association for Computational Linguistics - Annual Review of Research (ACL ARR) program.

Collection Method

Likely collected through the official ACL ARR peer-review submission system.

Time Range

null

Freshness

Last updated 2026-04-13 15:56:48; freshness should be verified.

Geography

null