The availability of annotated corpora is an important step towards the development of well-performing and generalizable NLP tools for temporal information extraction.
In this work, we present a corpus of clinical texts (for four different specialties) annotated with five types of time expressions, including their normalized values.
This corpus will be made available to the NLP community.
Learning Objective: 1) Understand the importance of creating and sharing clinical corpora annotated with temporal information;
2) Learn how temporal information can be documented and modelled in clinical text;
Natalia Viani (Presenter)
King's College London
Hegler Tissot, University College London
Ariane Bernardino, King's College London
Sumithra Velupillai, King's College London