Gold Standard Dataset for Clinical Literature Screening

Gold standard dataset for the screening process in literature reviews on Clinical Natural Language Processing. Created by Christopher Norman from an Academic Medical Center. The dataset is sourced from the paperswithcode platform and is designated as a gold standard test set.

Use Cases

• Train text classifiers to identify relevant clinical studies for systematic reviews using article titles and abstracts.
• Benchmark information retrieval systems on the task of prioritizing medical literature for review inclusion.
• Develop models for automating the 'screening' phase of systematic reviews by predicting relevance labels.
• Evaluate the performance of Clinical NLP tools on a curated gold standard test set for literature filtering.

Strengths

• Designated as a 'gold standard' dataset for benchmarking, indicating high-quality human annotations.
• Curated specifically for the Clinical NLP domain, ensuring task relevance.

Limitations

• Specific row count, column details, and sample size are unknown, limiting reproducibility assessments.
• The temporal coverage and update recency are unspecified, which may affect model applicability to newer literature.

Provenance

Source

Christopher Norman, Academic Medical Center, via paperswithcode.

Collection Method

Curated as a gold standard test set for the literature screening process.

Time Range

null

Freshness

null

Geography

null