The Review-5K dataset is a collection of peer reviews and associated metadata from the ICLR 2024 conference. It was created by WestlakeNLP and published on Hugging Face on February 21, 2025. The dataset is designed to facilitate research on the analysis of the peer review process.
Use Cases
- Analyzing review quality and sentiment based on the textual content of peer reviews.
- Studying reviewer behavior and decision patterns based on the associated metadata.
- Developing NLP models for review summarization or classification based on the review text.
- Investigating the relationship between paper information and review outcomes based on the collected metadata.
Strengths
- Focuses on a specific, high-profile academic conference (ICLR 2024).
- Designed explicitly for research on the peer review process itself.
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
- WestlakeNLP
- Collection Method
- Constructed by gathering paper information and peer reviews from ICLR 2024.
- Time Range
- 2024
- Freshness
- Last updated 2025-02-21 09:21:29; freshness should be verified.