# Event Extraction from Biomedical Literature

## TL;DR

This paper reviews recent advances in using natural language processing to extract events from biomedical literature, focusing on tools tested in BioNLP competitions.

## Contribution

The paper provides a comprehensive review of leading BioNLP methods and their innovative contributions to event extraction in the biomedical domain.

## Key findings

- Biomedical literature is vast and growing rapidly, making manual comprehension impractical.
- NLP methods can extract entities and relationships to help summarize and query biomedical articles.
- BioNLP Shared Task competitions have driven significant progress in automated event extraction.

## Abstract

The breadth and scope of the biomedical literature hinders a timely and thorough comprehension of its content. PubMed, the leading repository for biomedical literature, currently holds over 26 million records, and is growing at a rate of over 1.2 million records per year, with about 300 records added daily that mention ‘cancer’ in the title or abstract. Natural language processing (NLP) can assist in accessing and interpreting this massive volume of literature, including its quality NLP approaches to the automatic extraction of biomedical entities and relationships may assist the development of explanatory models that can comprehensively scan and summarize biomedical articles for end users. Users can also formulate structured queries against these entities, and their interactions, to mine the latest developments in related areas of interest. In this article, we explore the latest advances in automated event extraction methods in the biomedical domain, focusing primarily on tools participated in the Biomedical NLP (BioNLP) Shared Task (ST) competitions. We review the leading BioNLP methods, summarize their results, and their innovative contributions in this field.