Answer Extraction and Question Generation

Overview

We present a tool capable of (1) extracting question-worthy answers from paragraphs and (2) generating questions from both paragraphs and answers.

Rationale

Question Generation (QG) aims to automatically generate questions given an input. This input often requires two components: the source text, and the word(s) that will make up the question's answer. When developing these QG systems, datasets consisting of paragraph-answer pairs are typically used. However, in real environment systems, pre-identified answers are not widely available. Thus, we propose a system capable of extracting question-worthy answers from paragraphs and generating questions from both paragraphs and answers. Formally, we define four components (aka agents). The first three (answer extractors) identify question-worthy answers from the paragraphs. The last one (generator) has the role of generating questions, considering both answer and paragraph pairs.

Main Features

• Question-Worthy Answer Extraction using:
  • Named Entity Recognition
  • Keywords using BERT
  • Clauses
• Question Generation with T5 model

Prerequisites

Python 3 (tested with version 3.9.13 on Windows 10)

Installation

1. Install the Python packages from requirements.txt. If you are using a virtual environment for Python package management, you can install all python packages needed by using the following bash command:

   pip install -r requirements.txt

2. Install spacy-clausie (as indicated on the authors' page)

   git clone https://github.com/mmxgn/spacy-clausie.git
   cd spacy-clausie
   python setup.py build 
   python setup.py install [--user]
   
   # Optionally
   python setup.py test
   

3. Download this t5 model checkpoint (for question generation) and save it in the models_checkpoints folder (you can create it in the project root)

Usage

1. Web Application

   cd web_app/
   python app.py

2. Command Line (to be done)

Issues and Usage Q&A

To ask questions, report issues or request features, please use the GitHub Issue Tracker.

Contributing

Contributions are what make the open source community such an amazing place to learn, inspire, and create. Any contributions you make are greatly appreciated.

If you have a suggestion that would make this better, please fork the repo and create a pull request. You can also simply open an issue with the tag "enhancement". Don't forget to give the project a star! Thanks in advance!

1. Fork the Project
2. Create your Feature Branch (git checkout -b feature/AmazingFeature)
3. Commit your Changes (git commit -m 'Add some AmazingFeature')
4. Push to the Branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

License

Project

This project is released under the General Public License Version 3.0 (or later). For details, please see the file LICENSE in the root directory.

Additionaly, this project includes third party software components: stanza, KeyBERT, spacy-clausie, and T5 model. Each of these software components have their own license. Please see stanza/license, KeyBERT/license, spacy-clausie/license, and T5 correspondingly.

Commercial Purposes

A commercial license may also be available for use in industrial projects, collaborations or distributors of proprietary software that do not wish to use an open-source license. Please contact the author if you are interested.

References

If you use this software in your research/job/work, please kindly cite our project:

@software{ae_and_qg,
  author = {Bernardo Leite},
  title = {Answer Extraction and Question Generation},
  year  = {2022},
  version = {1.0.0},
  url   = {https://github.com/bernardoleite/answer-extraction-and-qg},
}

Also consider citing the third party software components (see on their respective pages): stanza, KeyBERT, spacy-clausie and T5.

Other important references:

• A Multi-Agent Communication Framework for Question-Worthy Phrase Extraction and Question Generation

Contacts

Bernardo Leite, bernardo.leite@fe.up.pt