Entity Recognition in Medical Research

Abstract

The COVID-19 pandemic has generated an extensive amount of literature, posing challenges in efficiently extracting relevant information. To address this, we developed a web-based entity recognition model that automatically extracts entities from medical research literature. Our model, known as the Entity Recognition in Medical Research web framework, aims to evaluate the performance of medical research related to COYID-19 and other associated diseases. Its primary goal is to recognize entities in medical literature, with a specific focus on COYID-19 research. To train our model, we utilized a large corpus of COVID-19 literature that underwent meticulous annotation to identify and classify entities. The training process involved leveraging the Spacy library, using its blank model architecture as the foundation for our machine learning model. We employed a custom pipeline that incorporated named entity recognition (NER) techniques. Our model encompassed various entity labels, including COMPOUND, DISEASE, SYMPTOMS, GENE, ORG, PROTEIN, CELL, AGENT, MEDICINE, TREATMENT, TOOL, PERCENTAGE, ENZYME, NUCLEICACID, CARDINAL, DATE, LOC, MICRO-ORGANISM, CONDITION, and AMINOACID, to comprehensively cover relevant entities in medical research literature. Our proposed model holds significant implications for the medical research community as it enables researchers and medical professionals to efficiently extract pertinent information from the vast COVID-19 literature. By automatically identifying and categorizing entities, our model empowers users to make better-informed decisions and gain crucial insights to combat the ongoing pandemic. The Entity Recognition in Medical Research web framework has the potential to revolutionize information extraction in the medical domain, fostering breakthroughs and advancements in COVID-19 research and related fields.