https://github.com/ahmedbesbes/keywords-extractor-with-bert

A Streamlit app to extract keywords using KeyBert
https://github.com/ahmedbesbes/keywords-extractor-with-bert

bert keybert keyphrase-extraction keyword-extraction maxsum mmr nlp nlproc streamlit streamlit-application transformers

Last synced: 2 months ago
JSON representation

A Streamlit app to extract keywords using KeyBert

• Host: GitHub
• URL: https://github.com/ahmedbesbes/keywords-extractor-with-bert
• Owner: ahmedbesbes
• Created: 2021-05-08T10:11:55.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2021-05-08T10:31:49.000Z (over 3 years ago)
• Last Synced: 2024-05-08T00:26:06.221Z (9 months ago)
• Topics: bert, keybert, keyphrase-extraction, keyword-extraction, maxsum, mmr, nlp, nlproc, streamlit, streamlit-application, transformers
• Language: Jupyter Notebook
• Homepage:
• Size: 168 KB
• Stars: 33
• Watchers: 3
• Forks: 12
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# How to extract meaningful and semantic keywords using BERT and Streamlit

This small Streamlit app uses **KeyBert** to extract meaningful keywords from text documents.

KeyBert can be an alternative to bag of words techniques (e.g. Count or Tfidf vectorizers) that might suffer from noisy results.

You can try it out by:

- pasting a text or picking a sample

- selecting the top n keywords to extract

- specifying the `keyphrase_ngram_range`

- adding diversity in the results by using Max Sum Similarity or Maximal Marginal Relevance

---

![](./images/screenshot.png)

### Run the app locally

```sh

pip install pipenv

cd keywords-extractor-with-bert

pipenv install

pipenv shell

streamlit run app.py

```

### Credits

- [KeyBert](https://github.com/MaartenGr/KeyBERT)