Welcome to SciAssist’s documentation!

SciAssist is a toolkit to assist scientists’ research. SciAssist currently supports reference string parsing, more functions are under active development by WING@NUS, Singapore. The project was built upon an open-sourced template by ashleve.

Check out the Usage section for further information, including how to use the tool.

Note

This project is under active development.

Contents

• Installation
  • Prepare the environment
  • Set up PDF parsing engine s2orc-doc2json
  • Install Grobid
• Usage
  • How to Parse Reference Strings
    • Parse strings from PDFs
    • Parse strings from TEXT or strings
    • Extract strings from a PDF
  • How to Summarize a document
    • Summarize a PDF
    • Summarize a TEXT or a string
    • Extract bodytext from a PDF
• Development
  • How to train on a new task
    • To prepare your dataset
    • To add a model
    • To create a Trainer and train
  • How to build a pipeline for a new task
  • Other points
    • Default directories
    • (Other standards and regulations are to be added here)