ClinScope Corpus - Clinical Notes Annotated for Hedge and Negation

Anonymous

Abstract: There has been great interest in knowledge extraction from biomedical texts. Part of this research involves hedge and negation assertion detection as doctors often use these assertions during the diagnostic process to specify likelihood or ruling out other possible diseases and conditions. Although natural language processing has been growing rapidly in the biomedical field, available corpora for clinical free-texts are still limited with research relying on limited available corpora where many are not annotated. In addressing this issue, we propose this ClinScope Corpus, a new clinical text corpus focused negation and hedge annotations. Our sampling allows for higher concentrations of assertion cues along with their scope and medical foci to aid in detecting when cues directly negate or mark medical entities uncertain.

Paper Type: short

Research Area: Resources and Evaluation

Contribution Types: Model analysis & interpretability, Data resources

Languages Studied: English

Consent To Share Submission Details: On behalf of all authors, we agree to the terms above to share our submission details.