Design of an Interactive Educational Application with Elements of Artificial Intelligence

In today's globalized world, proficiency in the English language is becoming increasingly vital, and mastering it unlocks a wealth of information and opportunities. While applications such as Duolingo allow for flexible learning, they have traditionally been limited to static, pre-defined study materials with fixed inputs and outputs. Artificial intelligence, particularly GPT models, offers the potential to transcend these limitations by enabling personalized and adaptive English language education. This thesis aims to design and implement a prototype mobile application for English language learning that leverages artificial intelligence and GPT models. The theoretical section presents a literature review of pedagogical methods and approaches to English language teaching, alongside a synthesis of research on language learning chatbots. Furthermore, a competitor analysis is conducted, primarily examining the integration of artificial intelligence in mobile language learning applications. The practical part details the design of the mobile application using the User-Centered Design (UCD) methodology. The design underwent iterative testing and was subsequently implemented using the React Native framework. The results indicate that users value the AI's ability to generate personalized lessons on arbitrary topics. With a SUS score of 69.5, the final application is deemed usable. Although the AI's performance is not flawless, this thesis confirms that employing GPT models in this context has the potential to enhance the accessibility of personalized English language learning. It is anticipated that this potential will expand as GPT models continue to evolve.