Dmytro Kalpakchi

My research is about helping computers to help teach humans

The overarching goal of my research is to help improving the way teaching is done by introducing technological solutions whenever possible to aid (NOT replace) teachers.

The goal of my PhD studies is, of course, more narrow than that! Firstly, the focus is only on one subject – Swedish as a second language. Secondly, the focus is only on one area within the subject – reading comprehension. Lastly, I focus only one aspect of teaching with that area – written examination.

Why do we need to improve examination? Because examination helps teachers understand the
needs of their students better. In turn, the clearer these needs are to the teachers, the faster their
students will learn.

How is it possible to improve examination? Well, one of the time-consuming parts of any written examination is actually creating the exam questions, which are very often in a form of multiple-choice items. This format involves creating a question and a number of alternatives, of which only one is typically correct. The main goal of my PhD studies is to let computers create multiple choice questions for the teacher, so that they can get a very good first draft of a test in a couple of minutes.

Taking steps towards achieving this goal is possible due to the generous funding coming from the following projects:

• AI-teacher for more effective learning of Swedish - part of the project, titled QuizMaster (funded by Vinnova)
• SWE-QUEST (funded by Digital Futures)

The pursued solution involves using what is commonly referred to as Artificial Intelligence (and more specifically Natural Language Processing) techniques. If you have heard of ChatGPT, then it's fair to say that my research attempts to use partially similar techniques, but not exclusively. Curious to know what exactly I experimented with? Feel free to visit my personal website for more information.