Knowledge Verificator

Knowledge Verificator is a tool for self-learning. It employs Natural Language Processing (NLP) techniques to facilitate and increase effectiveness of self-study. The project has been created as the Bachelor's Thesis of Igor Sosnowicz.

Using Knowledge Verificator is simple. Step by step:

1. Insert a paragraph you want to teach yourself or choose something from a databaset of predefined ones.
2. Knowledge Verificator generates a question for you.
3. You answer the question.
4. Your question is evaluated and you get the feedback.
5. The process repeats as long as you like.

Usage

Before running the Knowledge Verificator, you need to install Docker.

When Docker is installed, use the run.sh script to start the application:

./run.sh

The script handles building the image (approximately 10 minutes) and running it (approximately 30 seconds). Building the image happens only once, before the first usage of the application.

Notice: By default, only CPU is used. To enable GPU support, you have to install Nvidia Container Toolkit and configure it to cooperate with Docker by yourself. Then, you may use:

./run.sh --gpu

Development

If you are interested in contributing to the project by submitting source code, you have to have finish all the steps described below.

Prerequisites

You have to have the following tools installed:

• the Python build and dependency management system: poetry
• the JavaScript package manager: npm

Installation Steps

Backend

1. Clone the repository.

   git clone [email protected]:Iamhexi/knowledge_verificator.git

2. Enter its directory.

   cd knowledge_verificator

3. Install all dependencies, also including the optional ones.

   poetry install --with test

Frontend

Install the modules required by the frontend.

npm install --prefix frontend

Running

And then run the application (make sure you are in the root directory of the repository).

poetry run python knowledge_verificator/main.py