Curator extracts keywords from your documents and generates search queries to help you research better.
It works by parsing your document, extracting the text and running it through a keyword Bidirectional Encoder Representations from Transformers (keyBERT for short) to identify the keywords or key phrases in your document which is used to generate a search query. It's that straighforward.
The machine that I'm developing this application on does not have an nvidia gpu so the application has been configured to run on cpu which is a major overhead, please bear with me.
Licensed under MIT License