https://github.com/tomwalczak/pubmed-abstract-analyzer

NLP - exploring models across various tasks using PubMed RTC 200k dataset
https://github.com/tomwalczak/pubmed-abstract-analyzer

machine-learning nlp streamlit

Last synced: about 2 months ago
JSON representation

NLP - exploring models across various tasks using PubMed RTC 200k dataset

• Host: GitHub
• URL: https://github.com/tomwalczak/pubmed-abstract-analyzer
• Owner: tomwalczak
• Created: 2021-06-05T12:33:34.000Z (over 3 years ago)
• Default Branch: master
• Last Pushed: 2021-07-08T09:32:53.000Z (over 3 years ago)
• Last Synced: 2023-07-31T13:39:20.430Z (over 1 year ago)
• Topics: machine-learning, nlp, streamlit
• Language: Python
• Homepage: https://share.streamlit.io/tomwalczak/pubmed-abstract-analyzer
• Size: 143 MB
• Stars: 2
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# PubMed Abstract Analyzer 🍺 💊 📄 🔎

Explore how well different NLP models deal with summarization, classification and claim extraction, all on one page.

1. Improve readability of paper abstracts by breaking them down into sections. Inspired by the original discourse tagging paper: https://arxiv.org/abs/1710.06071

2. Extract material claims from each abstract by fine tuning the first model. Dataset and inspiration from https://arxiv.org/abs/1907.00962

### View the live app here: [PubMed Abstract Analyzer](http://tomwalczak.com/pubmed-abstract-analyzer/)

![Abstract Analyzer Demo](https://mldatatwk.s3.us-east-2.amazonaws.com/gifs/Animated+GIF-downsized_large.gif)