BioGPT: Relation Extraction is a language model fine-tuned to find a relation between entities.

Teaching machines to understand human language

Biomedical Language Models

AIMNET

Microsoft

BioGPT: Relation Extraction is a language model fine-tuned to find a relation between entities.
Included in this app are three models finetuned on different datasets each with a different use case.
<ul style="margin-left: 36pt">
<li style="list-style-type: disc">Finding the interaction between two drugs, finetuned on <a href="https://www.sciencedirect.com/science/article/pii/S1532046413001123" target="_blank" rel="noreferrer">DDI</a></li>
<li style="list-style-type: disc">Finding the interaction between a drug and its target, fine-tuned on <a href="https://academic.oup.com/bioinformatics/article/38/22/5100/6751771" target="_blank" rel="noreferrer">DTI</a></li>
<li style="list-style-type: disc">Finding the interaction between a chemical and a disease, finetuned on <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4860626/" target="_blank" rel="noreferrer">BC5CDR</a></li></ul>


<div>Example use case: <span style="display: inline !important; font-style: normal; color: rgb(51, 51, 51); text-decoration-color: initial; font-weight: 400; orphans: 2; font-size: 16px; font-variant-ligatures: normal; font-family: Poppins, sans-serif; text-decoration-style: initial; background-color: rgb(255, 255, 255); font-variant-caps: normal; text-decoration-thickness: initial; text-align: start; text-indent: 0px; word-spacing: 0px; widows: 2; float: none; letter-spacing: normal; text-transform: none; webkit-text-stroke-width: 0px">&nbsp;See examples below. Technology: GPT-2 Backbone Metrics: (Precision/Recall/F1) - DDI: (41.70/44.75/40.76), DTI: (40.00/39.72/38.42), BC5CDR: (49.52/43.25/46.17) Note: Input queries should be in the form of an abstract regarding the entities you wish to find the relation between.</div>

BioGPT: Relation Extraction - DDI, DTI, BC5CDR