Institutional Repository of Vilnius University
Edukacinių žaidimų konstravimo įrankiai /
Abstract
dc:descriptionKnowledge transmission is one of the crucial factors in the development and improvement of society. Therefore, the education sector has a significant role in the lives of both young people and adults. Felicitous tasks develop people's critical thinking, professional skills and personal development. As can be seen, today's youth are described as the 'always connected' generation. It is due to the online presence and use of digital technologies. Therefore, young people are more interested in learning through the digital or technological side, as they grow up using smartphones, computers and video calls from a young age. Digital technologies are increasingly taking their lives into the virtual space. For this reason, educational games are increasingly becoming a focus in education. Consequently, pupils are more interested and focused when tasks are in digital space. Design tools are becoming an essential part of educational games, and AI-based models are increasingly being used to build them. The use of chatbots and the understanding of their structure is increasingly mentioned. Chatbots are based on large language models for natural language processing and are capable of generating text, answering questions, creating pictures and many other functions. „OpenAI“ has developed „ChatGPT“, a chatbot competent in understanding human speech, processing it and providing the necessary output. „HuggingFace“ is an open-source platform, provides a library of transformers, natural language models and open collaboration with developers. Hence, the Llama2 model developed by „Meta AI“ for natural language understanding was tested. Large language models based on the transformer architecture, such as GPT-3.5-turbo-0125, GPT-4-turbo or GPT-4o, are convenient for generating large amounts of data. This can help educators to generate large amounts of information. Hence, in this work, these three models were used for creating a large number of mathematics problems for 4th-grade students including topics such as addition, subtraction, multiplication, division and simple algebra. In this context, the large language models can support the development of personalised learning content, tailoring it to the individual needs of the student. This could be a useful aid for teachers when faced with teaching challenges, lack of ingenuity in the design of tasks and the desire to make tasks unique and non-repetitive. Certainly, it is crucial to check the information generated, because there may be errors or contextual inconsistencies.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Švetkauskaitė, Dainora,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD210646945&prefLang=en_US