Earticle

본 연구는 교육 분야에서 AI를 활용한 연구의 동향을 분석하여 향후 AI활용교육의 방향성과 시사점을 제시하 는 데 목적이 있다. 이를 위해 2019년부터 2021년 7월까지 최근 3년간 국내 학술지에 게재된 논문 중 검토를 통 해 최종 78편을 분석 대상으로 선정하였다. 분석 결과는 다음과 같다. 먼저, 3개년 중에서는 2020년에 게재된 논 문이, 연구방법으로는 질적연구가 가장 많이 나타났다. 또한, 연구대상별 분석에 따르면 초등학생을 대상으로 한 연구가 가장 많았고 대학·대학원생을 대상으로 한 연구가 뒤를 이었다. 교과목별 분석에서는 외국어 교육과 관 련된 연구가 가장 많았고, AI 테크놀로지 유형은 챗봇이 가장 많이 사용되었다. 마지막으로 교수학습 및 평가 영역에서는 실행 단계가, AI활용교육 시스템 유형으로는 학생 지원이 과반수를 차지하였다. 이러한 결과를 바탕 으로 향후 AI활용교육의 방향성과 시사점을 제시하였다. 본 연구는 전체적인 관점에서 국내 AI활용교육 연구의 동향을 파악하였고, 교수자-학습자와 교수학습설계과정을 중심으로 AI활용교육을 살펴보았다는 점에서 의의가 있다.

The purpose of this study is to suggest the direction and implications of learning with AI in the future by analyzing the trends of research learning with AI in the field of education. For doing this, the final 78 papers published in domestic journals over the past three years from 2019 to July 2021 were selected for analysis through review. The analysis results are as follows. First of all, papers in 2020 among the three years were most published, and the most utilized research method was the qualitative research. In addition, according to the analysis by study subject, studies on elementary school students were the most common, followed by studies on college and graduate students. In the analysis by subject, research related to foreign language education was most utilized and chatbot was most used in the AI technology type. Finally, the research learning with AI accounted for the majority, and student support accounted for the majority as the type of education system learning with AI at the implementation stage among the areas of teaching and learning and evaluation. Based on these results, the direction and implications of learning with AI in the future were presented. This study is meaningful in that it grasped research trends of learning with AI in domestic from an overall perspective, and examined learning with AI focusing on the instructor-learner and the teaching and learning design process.

The need to detect lung cancer accurately and early has been brought out through the fact that lung cancer has remained one of the top causes of cancer-related deaths in the world. A CT scan is usually interpreted manually which makes this time consuming and subjective. This paper hypothesizes that an explainable implementation of automated deep learning can be used to classify lung cancer based on the transfer learning models (VGG16, VGG19, ResNet50, and EfficientNetB3) on the IQ-OTH/NCCD dataset. The dataset was stratified by means of the SMOTE and was split into 80 and 20 percent training and validation subsets, respectively. All the pretrained CNNs were fine-tuned with the Adam optimizer and categorical cross-entropy loss. VGG16 performed the best and had a validation accuracy of 98.64, precision and recall of 98, and ROCAUC of 99. Visualization of tumor regions was done using explainable AI techniques (Grad-CAM, LIME), which are interpretable and have diagnostic transparency. The suggested framework proves that transfer learning combined with XAI is more effective in terms of accuracy and reliability in diagnosing medical images and is one of the steps to clinically reliable smart healthcare systems.

4차 산업혁명 이후 인공지능 기술의 발달로 인해 변화하는 사회를 대비하고 실생활에서 인공지능을 다양하게 활용하기 위해, 학교 현장에서는 초등학생 때부터 기초 소양으로서 인공지능을 교육하여 학생들이 인공지능에 관한 기본적인 지식과 역량을 갖출 수 있도록 애쓰고 있다. 하지만, 현재 초등학교 교육과정 체제에서는 컴퓨팅 사고력을 비롯하여 초등학교 5-6 학년군에서 정보교 육이 시작되고 있다. 미래사회와 환경변화에 대응하는 교육이 강화되기 위해서는 인공지능 교육의 시작이 기존 교육과정에서보다 는 좀 더 빨리 진행될 필요가 있다. 본 연구에서는 초등학교 3-4학년군 학생에게 적절한 온작품읽기 기반 인공지능 교육 프로그램 을 개발하고 적용하여 개발한 프로그램의 효과성을 검증하였다. 온작품읽기로 맥락과 상황을 인지하여 학생들이 인공지능에 대한 기본 개념을 배울 수 있게 되었다. 본 연구에서 개발한 프로그램을 적용한 초등학생들의 지식・이해・과정 및 기능 영역에서 유의 미한 인식변화의 결과를 얻어 향후 초등학교에서의 인공지능 융합교육 프로그램으로 활용할 수 있음을 알게 되었다.

After the 4th industrial revolution, in order to prepare for a society that is changing due to the development of artificial intelligence(AI) technology and to utilize AI in various ways in real life, in the school field, AI is taught as a basic knowledge from elementary school students, so that students can acquire basic knowledge and capabilities related to AI. However, in the current elementary school curriculum system, Informatics education, including computational thinking, is starting in the 5th and 6th grade groups of elementary school. In order to strengthen education that responds to future society and environmental changes, the start of AI education needs to proceed more quickly than in the existing curriculum. In this research, an AI education program based on reading whole works suitable for students in the 3rd and 4th grade of elementary school was developed and applied to verify the effectiveness of the developed program. By recognizing the context and situation by reading the whole work, students can learn the basic concept of AI. It was found that it can be used as an AI convergence education program in elementary school in the future by obtaining results of significant cognitive change in the areas of knowledge, understanding, process, and function of elementary school students who applied the program developed in this research.

본 연구는 기존 인공지능 교육의 한계를 보완하고 교육 대전화의 패러다임을 반영한 AI교육 2.0 모델을 개발하는 데 목적이 있다 이 를 위해 국내외 AI 교육 등의 문헌 연구와 사례 분석, 집단지성 워크숍, 델파이 조사, 내용 타당도 검증을 포함한 설계기반 연구 방법 을 적용하여 AI교육 2.0 모델을 체계적으로 설계·개발하였다. AI교육 2.0 모델은 기초 역량으로 디지터시를 설정하고, 이를 토대로 AI 이해교육, AI협력교육, AI가치교육의 세 가지 핵심 교육 영역을 구조화하였다. AI교육 2.0을 통해 학습자가 신장하는 핵심 역량은 AI 사고력(AI Thinking)으로 추상화, 자동화, 지능화, 자율화의 4가지 요소를 제안하였다. 연구의 타당도 검증 결과, AI교육 2.0 모델은 전 반적으로 높은 타당도를 확보하였으며, 교육 현장에서의 적용 가능성과 이론적 정합성을 동시에 갖춘 모델로 평가되었다. 본 연구는 AI 기술의 이해를 넘어, AI와 함께 살아가는 시민을 기르는 f교육 모델을 제시하였다는 점에서 의의가 있다.

This study aims to develop the AI Education 2.0 model, which addresses the limitations of existing artificial intelligence education and reflects the paradigm shift toward educational transformation. To this end, a design-based research methodology was employed, incorporating literature review and case analysis of domestic and international AI education, collective intelligence workshops, Delphi surveys, and content validity verification, to systematically design and develop the AI Education 2.0 model. The model establishes Digitacy as its foundational competency and structures three core educational domains: Understanding AI Education, AI Collaboration Education, and AI Value Education. The central competency fostered through AI Education 2.0 is AI Thinking, which consists of four components: abstraction, automation, intelligence, and autonomy. The results of the content validity evaluation indicate that the AI Education 2.0 model demonstrates a high level of overall validity and is assessed as a model that simultaneously ensures theoretical coherence and practical applicability in educational settings. This study is significant in that it proposes an educational model that goes beyond understanding AI technologies to cultivating citizens capable of living and collaborating with AI in an AI-driven society.

본 연구는 간호학과 병태생리학 수업에서 발표 형식 선택과 인공지능(AI) 활용 가이드가 학습자의 학습 경험과 인식에 미치는 영향을 질적 사례연구를 통해 탐색하였다. 연구 대상자는 일개 대학교 간호학과 1학년 2학기 병태생리 학 교과목 수강생 중 발표 활동에 참여한 학생 6명으로, FameLab 형식 발표 경험자 3명과 일반 PPT 발표 경험자 3명 을 의도적 표집하였다. 성적 평가 종료 후 교수법 평가를 목적으로 포커스 그룹 인터뷰, 서면 응답, 개별보완 인터뷰 를 통해 자료를 수집하였으며, 주제분석 방법을 적용하여 분석하였다. 연구 결과, 발표 과제는 학습자가 자신의 이해 수준을 점검하고 학습 내용을 재구성하도록 유도하는 학습 촉진 요인으로 작용하였다. FameLab 발표는 개념의 내면화 와 설명 중심 학습 경험을 강조한 반면, 일반 PPT 발표는 학습 내용의 구조화와 복습 가능성을 강화하는 경험으로 인 식되었다. 또한 AI 활용은 학습을 대체하는 도구가 아니라 이해를 보조하고 학습 과정을 점검하는 수단으로 인식되었 으며, 교수자의 활용 가이드는 학습자의 비판적 사고를 강화하는 데 기여하였다. 본 연구는 병태생리학과 같이 개념적 난이도가 높은 교과목에서 발표 형식 선택과 AI 활용 가이드를 결합한 수업 설계의 교육적 가능성을 제시한다.

This qualitative case study explored nursing students’ learning experiences and perceptions of presentation format choice and artificial intelligence(AI) usage guidelines in a pathophysiology course. Participants were six first-year nursing students enrolled in a second-semester pathophysiology course at a single university. Using purposive sampling, three students who delivered FameLab-style presentations and three students who conducted conventional PowerPoint presentations were selected. Data were collected after course grading through focus group interviews, written responses, and follow-up individual interviews, and analyzed using thematic analysis. The findings indicated that presentation tasks functioned as learning facilitators by encouraging students to recognize gaps in their understanding and to reorganize learning content. FameLab presentations emphasized conceptual internalization and explanation-centered learning, whereas conventional PowerPoint presentations were associated with content structuring and opportunities for review. AI was perceived not as a substitute for learning but as a supportive tool for comprehension and self-checking. Instructor-provided AI usage guidelines, which emphasized source verification and conceptual accuracy, promoted students’ critical use of AI. These findings suggest that integrating presentation format choice with clear AI usage guidelines can be an effective instructional strategy for supporting learning in conceptually challenging courses such as pathophysiology. This study highlights the role of instructors as learning designers and learning consultants in AI-integrated nursing education.

본 연구에서는 초등학생의 영어 어휘 학습을 향상시키기 위해 AI 챗봇, 특히 ChatGPT를 활용하여 학습하고 그 효과를 분석하였다. 학습 도구는 Excel에 통합된 고유한 학습 프로그램이 포함되어 대화식 학습 방식을 제공 하며 상황에 맞는 실시간 응답을 생성하여 효과적인 학습 환경을 지원한다. 이 프로그램은 학생들의 단어 이해 와 사용을 점진적으로 향상시키기 위해 고안된 세션으로 구성된다. AI 챗봇은 이 프로세스를 안내하여 낱말의 정의, 동의어를 제공하고 의미 있는 대화를 촉진한다. 또한, 개인 맞춤형 및 상호작용형 학습 경험을 제공하여, 본 연구에서는 학생들이 감기, 열, 두통, 콧물, 복통과 같은 증상에 관련된 표현들을 습득할 수 있도록 돕는다. 실용적인 언어 능력을 키우고 학생들의 참여도를 향상시킴으로써, AI 챗봇이 언어 교육에서의 효과가 있음을 입 증하며 디지털 시대에 새로운 교육 방법의 방향성을 제시한다.

In this study, AI chatbots, particularly ChatGPT, were utilized to enhance the English vocabulary learning of elementary school students, and their effects were analyzed. The learning program, integrated into Excel, offered a unique approach, providing an interactive learning environment with contextual, real-time responses. The program comprised sessions aimed at progressively improving students' comprehension and utilization of words. AI chatbots guided students throughout the process, offering word definitions, synonyms, and facilitating meaningful conversations. Moreover, this study enabled students to acquire expressions pertaining to symptoms like cold, fever, headache, runny nose, and stomachache through a personalized and interactive learning experience. By nurturing practical language skills and boosting student engagement, AI chatbots demonstrate efficacy in language education, paving the way for innovative teaching methods in the digital era.

Injection-molded products frequently exhibit localized surface defects such as weld lines, flow marks, scratches, bubbles, and burn marks due to variations in material flow, mold temperature, and cooling conditions. Conventional visual inspection is highly dependent on operator experience, while rule-based machine vision methods are limited under variations in lighting and surface texture. This study proposes a deep learning–based defect detection model using YOLOv8 combined with a novel Defect-Aware Augmentation technique designed to enhance robustness for small, local defect regions. The proposed augmentation pipeline includes geometric transformations, optical perturbations, local defect patch synthesis, and diffusion-based synthetic defect generation. Experiments were conducted on a custom dataset of 5,000 images (3,000 normal and 2,000 defective). Results show that the proposed model achieves significant improvements over baseline models, obtaining 95% precision, 90% recall, and 0.96 mAP@0.5, outperforming the default YOLOv8 model by 7%p in mAP. Ablation studies verify that defect-aware augmentation is the dominant factor contributing to the performance gain. The proposed system demonstrates high applicability for automated quality inspection in injection- molding production lines.

2024년 기준 이주배경학생 수는 193,814명으로 전년 대비 7% 증가하였으나, 이들을 위한 교육에 관한 지원은 아직 미흡한 실정이다. 모국어가 아닌 언어를 이용하여 여러 교과를 학습할 때, 단어의 뜻을 파악하고 문장의 의 미를 이해하는 과정은 중요하다. 최근 초·중등학교에서 인공지능 디지털교과서(AIDT)의 도입이 현실로 다가왔 다. 만일, AIDT가 이주배경학생의 학습을 도울 수 있는 단어학습 도우미를 제공한다면, 교육적 소외를 줄일 수 있다. 본 연구에서는 이주배경학생의 한국어 이해를 돕기 위해 AI 디지털교과서 내 단어를 쉽게 이해할 수 있도 록 돕는 단어학습 도우미를 개발하였다. 본 연구에서 개발한 시스템은 학생들이 교과서의 단어를 쉬운 언어로 재구성한 정의와 학생의 모국어로 번역된 정의를 제공한다. 또한, 교과서에서 단어를 선택하면 해당 단어를 활용 한 예시 문장과 예시 문장에 대한 설명을 제시한다. 또한, 이주배경학생의 비율이 높은 베트남, 중국, 필리핀, 일 본어를 포함한 다양한 언어 지원을 통해 학생들의 한국어 이해 및 소외감 없이 자신의 문화와 언어가 존중받는 환경을 조성할 수 있어 교육 효과를 향상시킬 수 있다.

As of 2024, the number of students with migrant backgrounds in South Korea has reached 193,814, marking a 7% increase compared to the previous year. However, support for their education remains insufficient. Understanding the meanings of words and the context of sentences is essential when learning various subjects in a non-native language. Recently, the introduction of AI Digital Textbooks (AIDT) in primary and secondary schools has become a reality. If AIDT provides a vocabulary learning assistant to support the learning of students with migrant backgrounds, it could help reduce educational marginalization. This research developed a vocabulary learning assistant within AI Digital Textbooks to help Korean language understanding for students with migrant backgrounds. The system provides students with simplified definitions of textbook vocabulary and translations into their native languages. Additionally, when students select a word from the textbook, the system presents example sentences using the word and contextual explanations. Moreover, by supporting multiple languages, including Vietnamese, Chinese, Filipino, and Japanese—languages common among students with migrant backgrounds—the system can improve students’ understanding of Korean and foster an inclusive environment where their cultures and languages are respected. This approach enhances educational effectiveness and reduces feelings of alienation.

언어발달 장애를 가진 아동들은 일상생활 및 사회생활에서 많은 어려움을 겪으며 이는 생애 전반을 걸쳐 지속된 다. 언어발달 장애 아동들은 의사소통 수단인 언어를 이해하거나 사용하는 데에 어려움을 겪기 때문에 종종 사회적 활동 에 참여할 기회를 박탈당하곤 한다. 이와 관련해서 Augmentative and Alternative Communication(AAC, 보완대체 의사소통)는 언어장애를 앓는 이들에게 실직적인 의사소통 수단으로 사용될 수 있다. 본 논문은 픽토그램을 AAC의 수 단으로써 최대한 활용하여 언어발달 장애 아동이 타인과 의사소통하고 언어 이해 능력을 향상시킬 수 있도록 돕는 딥러 닝 기반 인공지능 서비스를 제안한다. 본 서비스를 통해 언어 문제를 겪고 있는 이들이 자신의 의도 혹은 욕구를 보다 수월하게 표현하여 삶의 질이 향상 될 수 있을 것으로 기대한다.

Children with language developmental disabilities often struggle through their lives from a lot of challenges in everyday life and social activities. They’re often easily deprived of the opportunity to engage in social activities, because they find difficulty in understanding or using language, a core means of communication. With regard to this issue, AAC(Augmentative and Alternative Communication) can be an effective communication tool for children who are suffering from language disabilities. In this paper, we propose a deep learning-based AI service to make full use of the pictogram as an AAC tool for children with language developmental disabilities to improve not only the ability to interact with others but the capacity to understand language. Using this service, we strive to help these children to more effectively communicate their intention or desire and enhance the quality of life.

항공 분야에도 머신러닝 (ML; machine learning) 기반의 인공지능(AI; artificial intelligence)를 활용하는 시스템 개발이 본격적으로 시작되었다. 항공용 소프트웨어는 항공무선기술위원회(RTCA; Radio Technical Commission for Aeronautics) DO-178C 또는 DO-278A 등의 표준을 통해 안전성 보증을 하고 있으며, 이 표준들은 결정론적 특성과 설명가능성을 내재한 소프트웨어를 대상으로 개발되었고 잘 적용된다. 반면 ML 기반 AI는 그 특성을 고려할 때, 이러한 기존 소프트웨어 인증 표준 적용만으로는 그 신뢰성을 제대로 보증하기 어렵다. 본 논문에서는 유럽항공안전청(EASA; european union aviation safety agency)이 이에 대응하기 위해 제시하는 새로운 인증 방법론에 대해 알아보고, AI가 적용된 항공 소프트웨어 인증을 위해 국내 규제당국과 산업계가 어떤 준비를 해야 하는지 논의한다.

Recently, the Machine Learning based Artificial Intelligence has introduced in aviation field. In most cases, safety assurance of aviation software is achieved by applying RTCA DO-178C or DO-278A or similar standards. These standards were developed for and are well-suited to software that has inherent deterministic properties and explainability. Considering the characteristics of AI software based on ML, it is not feasible to assure the integrity of those new aviation systems using traditional software assurance standards mentioned above. In this paper, we research the certification framework that is newly suggested by EASA to deal with the aviation system including ML AI functions, and discuss what should the Korean authority and related industries prepare to cope with this issue.

본 연구는 학습에서 생성형 AI Chat 서비스를 이용한 학습 경험을 SERVPERF 기반 서비스품질 관점에서 분석하고, 서비스품질 요인과 자기효능감, 지각된 가치 간의 구조적 관계를 실증적으로 검증하였다. 이를 위해 ChatGPT와 Gemini를 학습에 활용한 대학생을 대상으로 설문조사를 실시하고 구조방정식모형을 적용하였다. 분석 결과, SERVPERF의 서비스품질 요인은 모두 자기효능감에 유의한 긍정적 영향을 미치는 것으로 나타났다. 또한, 신뢰성, 반응성, 확신성은 지각된 가치에 직접적인 영향을 미쳤다. 그러나 유형성과 공감성은 자기효능감을 매개로 간접적인 영향을 미치는 것으로 나타났다. 본 연구는 생성형 AI Chat 서비스의 학습 가치를 이해하는 데 있어 서비스품질과 학습자의 심리적 요인을 통합적으로 고려해야 함을 증명한 것이다.

This study examines university students' learning experiences with generative AI-based chat services through a SERVPERF-oriented service quality framework and empirically tests the structural relationships among service quality dimensions, self-efficacy, and perceived value. Survey data were collected from students who used ChatGPT and Gemini for learning purposes, and structural equation modeling was employed for hypothesis testing. The results indicate that all SERVPERF service quality dimensions exert significant positive effects on learners' self-efficacy. Reliability, responsiveness, and assurance were found to have direct effects on perceived value, whereas usefulness and empathy influenced perceived value indirectly through self-efficacy. These findings suggest that understanding the educational value of generative AI chat services requires an integrated consideration of both service quality attributes and learners' psychological factors.

HTHT 교육은 최신 AI 기술(High-Tech)과 교수-학습상호작용(High-Touch)을 결합하여 학생 중심의 맞춤형 학 습환경을 조성하는 것을 목표로 한다. 이에 본 연구에서는 Google Arts and Culture, ChatGPT for Edu 등 사례조사 를 통해 생성형 AI (Perplexity, DALL·E, Suno 등) 활용 문화예술 교육에서 HTHT 방식의 효과성과 적용 가능성을 탐색하였다. 그 결과 High-Tech 관점에서 메타버스를 활용한 루브르미술관 가상투어, 학습경험플랫폼(LXP)로서 로블 록스 스튜디오 실습이 가능했다. 나아가 High-Touch 관점에서 AI는 문화예술 교육에서 창작과정의 효율성을 높이고, 창의성을 보조하는 중요한 도구로 활용될 수 있어 레오나르도 AI 아트스쿨, 베토벤 10번 교향곡 AI 연주감상 및 비평 등이 가용사례로 확인되었다. 요컨대 HTHT 기반 교육이 학습자의 참여도와 몰입도를 높이는 데 기여하고 문화예술 교육의 혁신적인 모델로 활용될 가능성이 있음을 검증하고, 실무적으로 이를 준비할 수 있는 단계별 로드맵을 제시한다

The High-Touch High-Tech (HTHT) approach integrates advanced AI technologies (High-Tech) with interactive and student-centered learning to create a personalized educational environment. This study explores the application of Generative AI (Perplexity, DALL·E, Suno, etc.) in cultural and artistic education and examines its effectiveness in Google Arts & Culture and ChatGPT for Edu. The findings suggest that HTHT-based education enhances students’ creativity, and engagement in learning. As a result, from a High-Tech perspective, it was possible to conduct a virtual tour of the Louvre Museum, and Roblox Lua as the LXP implementation practice using metaverse technology. Also, as a High-Touch perspective, AI can be used as an important tool to increase the efficiency of the creative process in arts education and assist human creativity. Leonardo. AI Art School and Beethoven's 10th Symphony AI composition were confirmed as good practices. In conclusion, the results verified that HTHT-based education can contribute to increasing learner participation and immersion and can be utilized as an innovative model for cultural arts education, and a roadmap could be presented.

The current study examines whether AI-assisted TOEIC learning program positively affects TOEIC scores of students majoring in airline services. The study focuses on how the effectiveness of AI technology differs by gender and proficiency. The experiment was conducted during the 2020 academic year. The participants were 119 EFL students from three intact classes at a university in Korea. For the study, the participants studied English through AI-assisted TOEIC learning technology called Soljam for the experimental period of 10 weeks. Before the treatment, all participants took a pre-test based on TOEIC and assigned to three different proficiency groups: Basic, Elementary, and Elementary Plus. After the treatment, all the participants took a post-TOEIC test. Major findings are as follows: First, AI technology positively affected the students’ TOEIC scores at a significance level of .05. Second, students’ improvement in TOEIC scores significantly differed by gender. However, statistically significant difference was not found regarding proficiency levels. There was no significant interaction effect of gender and proficiency. The findings of this study will shed light on the educational value of AI technology to promote language skills of Korean airline service students.

This study investigates the experiences of Korean university students in online English learning environments, focusing on self-directed strategies, instructor and AI feedback, and emotional and technical challenges. As online education continues to expand, understanding learners’ lived experiences has become increasingly important. Employing a qualitative research design, the study analyzed open-ended questionnaire responses and semi-structured interview data from 65 students enrolled in an English language course. Thematic analysis revealed six core themes: the importance of instructor interaction, the role of learning strategies, perceived benefits of online learning, emerging challenges, limitations of AI feedback systems, and the factors influencing learner confidence and persistence. Findings indicate that while students appreciate the flexibility and self-paced learning opportunities offered by online platforms, they face difficulties related to isolation, system usability, and inconsistent feedback. Instructor presence and teacher-centered scaffolding were identified as key elements in maintaining learner motivation, emotional security, and conceptual clarity. Unlike AI-driven feedback, which often lacks contextual nuance and affective support, teacher interaction provided adaptive and empathetic guidance aligned with learners’ needs. The study underscores the necessity of hybrid learning models that leverage both technological tools and human-mediated support, highlighting the central role of instructors in ensuring responsive and effective AI-based online English education.

This comparative case study examined the experiences of urban and rural elementary school students in Korea using PengTalk, an AI-based English learning application. Using a mixed-methods approach, the study analyzed student surveys, usage records, and interviews. Quantitative data were examined through descriptive statistics and Mann-Whitney U tests, while qualitative responses were analyzed thematically. Both groups expressed positive perceptions, especially regarding PengTalk’s gamified and interactive features. Urban students outperformed their rural peers in average scores (82.5 vs. 75.8) and cumulative totals (28,224 vs. 12,369). However, gamification alone was insufficient to sustain engagement, particularly in the rural school. Major challenges included limited digital literacy and inadequate instructional support in the rural school, as well as technical issues such as voice recognition errors in the urban school. The findings indicate that student engagement with AI tools depends not only on application design but also on contextual factors such as access, support, and digital infrastructure. This study concludes with recommendations for more inclusive and equitable AI-assisted language learning environments.

Practicum is essential for pre-service teachers to bridge the gap between theories and practice and prepare for the real classroom. However, it is often limited due to various reasons, such as time, cost, and a lack of partnership schools, in many countries. As an alternative to an actual practicum, the current paper explored a virtual practicum simulation, simSchool, to mitigate a shortage in reality-based preparation and investigated its effectiveness for pre-service teacher education. The study applied a mixed-method to perform more robust research. The study used quantitative content analysis of two versions of 37 pre-service teachers’ lesson plans (original and revised), reflection papers, and interviews as a primary method. As a secondary method, pre- and post-surveys on teacher efficacy were used. The results showed that pre-service teachers valued their experience with the virtual practicum and increased their abilities in diverse areas of teaching, including instruction, activities, facilitation, and material use. Their simSchool experience also helped the pre-service teachers realize their weaknesses in teaching and enabled them to transfer their newly learned knowledge to practice. Based on the results, the study argues that a virtual practicum offers pre-service teachers a valuable extended learning opportunity to develop their teaching skills, although it cannot replace real classroom teaching experience.

Objectives: The demand for Korean language learning has increased with global interest in Korean culture. However, non-native speakers struggle with pronunciation due to Korean's unique phonological features. This study aims to develop an AI-based pronunciation evaluation application to provide structured and personalized feedback. Methods: The system uses a client?server architecture, integrating an Android app, a FastAPI server, and an SQLite database. The AI model evaluates pronunciation based on accuracy, fluency, and completeness, and checks keyword usage in learners' spoken responses. Results: A pilot study with non-native Korean learners showed high user satisfaction with the system’s learning content and feedback quality. Conclusions: The AI-based application effectively supports Korean pronunciation learning by offering real-time, structured feedback. Pilot results suggest strong potential for broader application in Korean language education.

Artificial intelligence techniques have been developed in the prediction of corporate bankruptcy over time. The first step in the analysis of real-world data on bankrupt companies is to include multicollinearity, which has an adverse impact on biased estimation and possibly causes large errors. Studies in finance have presented potential problems affected by a strong association between the features and the outcome in the dataset, but examining the role of the factor analysis techniques which handle bias in the corporate data is at a nascent stage. The integrative big data analytics can be utilized to combine unsupervised learning to understand the structures of high-dimensional data with supervised learning to classify the target outcome efficiently. In this study, the results of the big data analytics show that random forest classification with factor analysis outperforms other big data analysis techniques in terms of predictive accuracy. The goal of this study is to minimize the gap between the theoretical concepts of artificial intelligence techniques and the analysis of real-world financial data, in addition to developing big data analytics methods on high-dimensional data with strongly associated corporate features. The method proposed in this study can be applied to similarly structured data which may contribute to understanding the interplay between corporate bankruptcy and financial features.

이 연구는 AI를 활용한 교육의 시의성에 입각해 AI활용교육 이슈별 교원의 정책 요구를분석하고 시사점을 도출하는 데 목적이 있다. 이를 위해 ‘AI 디지털교과서’, ‘데이터 관리 및표준화’, ‘학생용 AI 튜터링 시스템 및 교사용 AI 보조교사 시스템’의 AI활용교육 이슈를 중심으로 국내 교원 대상 설문조사를 하였고, 수집된 자료를 대상으로 Borich 요구도 분석, TheLocus for Focus 모델 활용 분석을 하였다. 이 연구의 분석결과는 다음과 같다. 첫째, 국내교원은 AI활용교육 관련 정책의 중요성을 높게 평가했지만 준비도는 낮은 것으로 인식하고있었다. 특히, ‘데이터 관리 및 표준화’ 관련 요구도가 높은 것으로 나타났다. 둘째, ‘AI 디지털교과서’의 경우 ‘AI 디지털교과서 관련 법⋅제도 개선'이 제일 높은 요구도를 보였고, ‘데이터 관리 및 표준화’는 ‘데이터 관리 환경 조성’이 제일 시급한 항목인 것으로 밝혀졌다. 또한, ‘학생용 AI 튜터링 시스템 및 교사용 AI 보조교사 시스템’에서는 ‘안전성 확보’가 최우선 정책과제로 도출되었다. 이와 같은 결과를 토대로 향후 AI활용교육의 현장 안착을 위한 정책적제언을 제시하였다.

The purpose of this study is to analyze teachers' policy demands in Learningwith AI, focusing on AI Digital Textbook, Data Management andStandardization, and AI Tutoring Systems for Students and AI Assistant Systemsfor Teachers. A survey of South Korean teachers was analyzed using the BorichNeeds Assessment and the Locus for Focus Model. The findings show that while teachers recognize the importance of Learningwith AI policies, they perceive a lack of readiness for implementation. Amongthe three issues, Data Management and Standardization had the highest policydemand. For AI Digital Textbook, legal and institutional improvements were themost urgent need. In Data Management and Standardization, establishing arobust data management environment was the top priority. For AI TutoringSystems for Students and AI Assistant Systems for Teachers, ensuring safety wasthe greatest concern. Based on these findings, this study proposes policy recommendations aimed atfacilitating the effective integration of Learning with AI into real-worldeducational settings.

본 연구는 UTAUT모형을 기반으로 초등교사의 AI활용교육에 대한 성과기대, 노력기대, 사회적 영향, 촉진조건, 수용의도, 실제수용 간의 구조적 관계를 분석하고 성별, 교직경력, 담당학년, 사전실행경험의 조절효과 분석을 목적으로 수행되었다. 이를 위해 서울 소재 초등학교 교사 273명이 참여하였으며, 제안된 모형을 검증하고 변인 간의 구조적 관계를 구조방정식으로 분석하였다. 분석 결과 첫째, 성과기대, 사회적 영향, 촉진조건이 AI활용교육의 수용의도에 유의한 영향을 미쳤다. 둘째, 수용의도는 실제 행동에 유의한 영향을 미쳤다. 셋째, 수용의도의 간접효과는 수행기대, 사회적영향, 촉진조건, 실제수용 사이에서 유의한 것으로 나타났다. 넷째, 사전실행경험의 조절효과는 모든 경로에서 유의하지 않았지만, 촉진조건과 수용의도 간에 성별, 교직경력, 담당학년의 조절효과는 유의하였다. 또한 노력기대와 수용의도 사이에서 담당학년의 조절효과가 입증되었다. 이러한 결과를 바탕으로 초등학교 교사의 AI활용교육의 수용의도와 실제수용을 촉진하기 위한 방안을 제안하였다. 첫째, 초등교사가 AI활용교육을 지도할 때 수업의 성과를 향상시키며 효율성을 높이는 데에 유용할 것이라 생각할 수 있게 하는 방안을 간구해야 할 것이다. 둘째, 초등교사가 AI활용교육을 지도할 때 개개 교사가 소속된 학교 구성원의 AI활용교육에 대한 전반적인 인식을 함께 키울 수 있는 지원 정책이 필요하다. 셋째, AI활용을 위한 조직적, 기술적 및 인적 지원의 인프라를 구축해야 한다. 이러한 제언을 바탕으로 초등교사의 AI활용을 높이기 위한 체계적인 전략을 고민해 볼 필요가 있다.

The purpose of this study is to verify the structural relationships among performance expectancy, effort expectancy, social influences, and facilitating conditions, which are expected to affect elementary school teachers’ usage intention and actual usage of ‘Learning with AI’. This research also investigated moderating effects of gender, teaching experiences, grades, and prior experience. To examine this purpose, 273 elementary school teachers in Seoul participated, and structural equation modeling was employed to validate the proposed model and analyze the causal relationships among variables. As a result, the proposed model was acceptable with a good fit, and there were significant direct effects between factors. Research results were as follows: First, performance expectancy, social influences, facilitating conditions significantly affected usage intention of ‘learning with AI’; second, usage intention significantly affected actual usage; third, the indirect effects of usage intention were significant between performance expectancy, social influences, facilitation conditions, and actual usage; fourth, the moderating effect of prior experience was not significant on all paths, however, the moderating effects of gender, teaching experiences, and grade between facilitating conditions and usage intention were significant. In addition, between effort expectancy and acceptance intention was moderated by grade. Based on the results, this research suggests practical guidelines to facilitate ‘learning with AI’ of elementary school teachers.