slides2questions

a code repo to generate questions from a given slide or study material

Warning

This project is still in development and may not work as expected. Please report any issues you encounter.

Getting Started

Prerequisites

• Python 3.10 or higher
• pip

Installation

1. Clone the repo

   git clone

2. Install the required packages

   pip install -r requirements.txt

3. Set up environment variables

   cp .env.example .env

   replace the placeholder GOOGLE_API_KEY with your own key in the .env file. You can get the key from here.

Usage

1. Put the PDF files you want to generate questions from in a directory. Let's say the directory is pdfs/.

2. Run the following command

   python src/cli.py pdfs/

   or

   python src/cli.py pdfs/ --extract-text-from-images

3. The questions will be generated in the questions_and_answers.json file in current directory by default. You can change the output file by using the --output option.

   python src/cli.py pdfs/ --output my_questions.json

Options

> python src/cli.py -h 

 usage: pdf2questions [-h] [--verbose] [--extract-text-from-images] [--number-of-topics NUMBER_OF_TOPICS] [--passes-over-corpus PASSES_OVER_CORPUS]
                     [--max-answers MAX_ANSWERS] [--min-answers MIN_ANSWERS] [--correct-answers CORRECT_ANSWERS]
                     pdf_directory

 Generate questions from PDF

 positional arguments:
 pdf_directory         Directory containing PDF files

 options:
 -h, --help            show this help message and exit
 --verbose, -v         Print more information (default: False)

 PDF options:
 --extract-text-from-images, -e
                         Extract text from images in the PDF (slower, requires `pip install rapidocr-onnxruntime`) (default: False)

 LDA options:
 --number-of-topics NUMBER_OF_TOPICS, -n NUMBER_OF_TOPICS
                         Number of topics to extract from the text (default: 10)
 --passes-over-corpus PASSES_OVER_CORPUS, -p PASSES_OVER_CORPUS
                         Number of passes over the corpus when training the LDA model (higher values may improve the quality of the topics but also
                         increase the training time) (default: 5)

 Multiple choice question options:
 --max-answers MAX_ANSWERS, -m MAX_ANSWERS
                         Maximum number of answers to generate for each question (default: 5)
 --min-answers MIN_ANSWERS, -i MIN_ANSWERS
                         Minimum number of answers to generate for each question (default: 4)
 --correct-answers CORRECT_ANSWERS, -c CORRECT_ANSWERS
                         Number of correct answers to generate for each question (default: 1)