Teachers, educational book authors, and instructors often invest considerable time in crafting assignments, quizzes, and questionnaires. The challenge arises when their focus shifts from teaching the material to designing meaningful, non-trivial questions - many of which may already be addressed in the textbook.
This task becomes even more demanding for larger courses or comprehensive textbooks, as generating a diverse range of questions, along with their variations, can be a time-consuming process. To simplify this process, there was a clear need for an intuitive and easy test maker that could automatically generate various question types and assemble assignments directly from textbooks or other educational resources.
Our client is a textbook publisher based in Austria.
Austria
Ongoing
6 specialists
The proposed solution is an AI question generator from text in various fields and topics. The system is based on user materials containing information in both German and English. This information can include books, articles, textbooks from the curriculum, or any other materials used in education or business.
The system generates questions using Azure OpenAI API (ChatGPT) and supports various types of questions (open-ended, multiple-choice, yes-no), generating correct answer options. An important addition is the ability to paraphrase questions and answers.
Step 1:
The user interface is pretty simple to navigate. First, the user can either upload a PDF file (whether it’s a book, article, or publication) or insert a link to the content. To generate questions, the user just chooses a specific section or range of pages to focus on. It also possible to set the level of creativity of the questions on a scale of 1 to 10 using the "model options" function.
Step 2:
We have downloaded the book titled "Introduction to Scrum – Student Guide"1 to provide you with an illustrative example. This is what the settings panel for question generation looks like. During this phase, the user inputs a topic (using a few keywords or a sentence) for which questions should be generated. The user also chooses the question type (open-ended, multiple-choice, yes-no) and specifies the desired number of questions and corresponding answers. For example, for the topic "PROJECT TEAM" we need to get 2 multiple-choice questions and 3 answer choices.
Step 3:
For this query, we got the following result. You can use the "Paraphrase" function to view other variants of questions. The results can be exported to a file feed
The implementation process of such a solution involves several stages, starting from requirements identification and concluding with system deployment and support. Here is a general overview of the stages:
The implementation of the proposed solution will bring several significant benefits to the educational organization or publisher house as a whole and to the teachers and authors in particular. Here are some potential outcomes.
To put it simply, we’ve designed an AI test question generator that creates high-quality questions and answers from any book or text. This tool saves authors and instructors hours of work every day, while also making student assessments more interactive and enjoyable - transforming the learning experience for everyone involved.
"This tool works on its own but can seamlessly integrate with other products. The AI quiz generator isn’t just in education – it’s also great for e-commerce, helping shoppers understand products and make better decisions. It can even improve customer support by giving more detailed answers and be useful in marketing for creating content and planning campaigns."