This study investigated the relationships among learners’ perceptions of AI chatbot use, self-directed learning, and self-efficacy in an interview English course utilizing ChatGPT, and explored the educational implications of generative AI-based instructional design. The course was redesigned around AI tools and implemented in a cyclical structure consisting of model answer generation, instructor feedback, and speaking practice. Answer sharing through Padlet and TTS-based pronunciation practice were also incorporated. Survey data from 49 out of 63 enrolled students majoring in Airline Cabin Service were analyzed. The results indicated that learners perceived the AI chatbot as practically helpful for organizing interview responses and preparing for speaking tasks, with generally high levels of positive responses across related items. Experiences of adjusting question difficulty through prompting and generating responses were associated with increased learning efficacy and speaking confidence. Subcomponents of self-directed learning and self-efficacy showed significant correlations with perceptions of AI chatbot use, and regression analysis revealed that the model including these two variables significantly explained learners’ perceptions of AI use. This study shows that systematically integrating generative AI with instructor feedback can enhance learners’ self-directedness, self-efficacy, and speaking confidence, suggesting the value of positioning AI as a structured learning partner in speaking-focused courses.

목적: 본 연구는 감각통합 평가 도구인 Sensory Profile 2 (SP-2)의 해석 복잡성, 주관성, 시각화 한계를 개선하기 위해 인공지능(artificial intelligence) 기반 챗봇(ChatBot) 시스템인 Sensory Profile 2 Summary and Interpretation (SPSI)를 개발하고 그 활용 가능성을 탐색하였다. 방법: SP-2의 평가 항목과 점수 산출 방식을 기반으로 자동화된 ChatBot 인터페이스를 구현하였으며, 자연어 기반의 문답 시스템을 통해 사용자가 직관적으로 평가에 참여할 수 있도록 설계하였다. ChatBot은 자동 점수 산출, quadrant grid/score별 해석, 중재 전략 제시, 시각적 요약 등의 기능을 포함하고 있으며, OpenAPI 연동을 통해 다양한 출력 형태를 지원한다. 결과: SPSI는 평가자의 주관 개입을 최소화하면서도 일관된 해석과 전략 제시가 가능하며, 시각화 자료(파이차트, 행동 예시 등)를 통해 보호자 및 비전공자에게도 직관적인 결과 전달이 가능하였다. 특히 설문 입력에서 결과 도출까지의 전 과정을 자동화함으로써 평가 소요 시간과 해석 부담을 줄이는 데 기여할 수 있다. 결론: SPSI는 작업치료 분야에서 기존 감각통합 평가 도구의 한계를 보완할 수 있는 사용자 중심의 디지털 평가 도구로서의 가능성을 보여주었다. 향후에는 임상적 유용성을 확보하기 위해 신뢰도 검증, 설명 가능한 인공지능(explainable artificial intelligence) 도입, 보안 체계 강화 등의 보완이 필요하다.

The purpose of this study is to develop a chatbot for Korean language education using Google Dialogflow, aiming to determine its effectiveness as a learning tool and to investigate learners’ perceptions of prototype chatbots. Six Korean language education chatbots were developed for beginner Korean language learners, and 30 beginner learners—15 from Mongolia studying in Korea and 15 from Japan studying abroad—were recruited. These learners practiced Korean conversation with the chatbots for about three weeks, and pre- and post-surveys were conducted. Analysis showed that using the chatbots for conversation practice positively impacted learners’ confidence. However, the chatbots’ overall effectiveness fell short of expectations, as significant results were observed only in confidence among the three investigated areas: interest, confidence, and motivation. Learners were generally satisfied with the chatbots, although Japanese learners rated its effectiveness in improving expressions and comprehension skills low, suggesting a need for further analysis. Both groups found voice chatbots more beneficial, underscoring a significant need for individual speaking practice, especially among Japanese learners. Despite limitations such as the inability to engage in flexible conversations, the study demonstrates the potential of chatbots developed as a learning tool and identifies learners’ perceptions from various perspectives.

The purpose of this study was to investigate how the English speaking ability of Korean EFL college students was affected by their interactions with Talk-to-ChatGPT while taking an ‘English Interview’ class. Thirty pieces of English conversation scripts with thirty chatbot conversations created by five students were collected for analysis. Two online text analysis programs, Quillbot including word counter and grammar checker and T.E.R.A.(Text Ease and Readability Assessor), were used for data analysis. The findings of data analysis revealed that 1) The average length of the sentences and words spoken by the participants has increased through English speaking practice using Talk-to-ChatGPT, and 2) There was no significant change in text ease and readability, and coherence of students’ utterances through English speaking practice using a chatbot while there were differences depending on their English proficiency levels. 3) Students A, B, and D, who had relatively low levels of English proficiency, showed a slight increase in syntactic accuracy and semantic clarity in their English interview practice. Based on the study findings, pedagogical implications for the effective use of AI-based apps or programs in English speaking classes were presented.

Due to the aging of a building, 38.8% (about 2.82 million buildings) of the total buildings are old for more than 30 years after completion and are located in a blind spot for an inspection, except for buildings subject to regular legal inspection (about 3%). Such existing buildings require users to self-inspect themselves and make efforts to take preemptive risks. The scope of this study was defined as the general public's visual self-inspection of buildings and was limited to structural members that affect the structural stability of old buildings. This study categorized possible damage to reinforced concrete to check the structural safety of buildings and proposed a checklist to prevent the damage. A damage assessment methodology was presented during the inspection, and a self-inspection scenario was tested through a chatbot connection. It is believed that it can increase the accessibility and convenience of non-experts and induce equalized results when performing inspections, according to the chatbot guide.

Chatbot-based services in online travel agency (OTA) are rapidly spreading in order to respond more agilely to consumers' needs based on the digitalization of the travel industry. Although AI chatbots use anthropomorphism to provide social experiences on behalf of humans, research results on its effects are mixed. Therefore, based on construal level theory, this study suggests the degree of anthropomorphism (low vs. high) of chatbots prime mental representations of different construal levels (low vs. high) and the fit between anthropomorphism and communication context (communication types and conversation types) has a positive effect on use behavior. This research method consisted of sentimental analysis for exploring use behavior of AI chatbots and two experimental studies (study 1 and study 2) to examine the hypotheses. The results of this study expand construal level theory and avatar research to provide an understanding of the anthropomorphism of AI chatbots.

Emojis have become an important element in human-chatbot interaction to communicate emotion. In addition to facilitating emotional communication, emojis are able to engage consumers, enhance relationship strength and influence consumer behavioural intention. Therefore, it is crucial to understand how the use of emojis affects customer-chatbot rapport and purchase intention in the consumption context.

인터넷은 전통적인 미디어를 대체하고 주요 뉴스 미디어 플랫폼 중 하나가 되었습니다. 인터넷 소스의 뉴스는 접근성이 좋고 편리하기 때문에 기존 뉴스 소스에 비해 빠르고 간단하게 이동할 수 있습니다. 그러나 가짜 뉴 스가 대량으로 발생하고 정치적, 상업적 이유로 온라인 커뮤니티에 퍼지면서 확인되지 않은 소식통으로부터 입수한 모든 언론 보도가 진짜인 것은 아니다. 가짜 뉴스는 이론적으로나 의도적으로 독자들을 속이거나 잘못 알릴 수 있다. 왜냐하면 사람들은 오프라인 커뮤니티에 영향을 미칠 수 있는 어떤 정보에도 쉽게 얽히게 되기 때문이다. 일부 수동 웹사이트는 정보가 사실인지 확인하도록 설계되어 있지만 온라인, 특히 웹에서 빠르게 확 산되는 정보의 양은 확장되지 않습니다. 이 문제를 해결하기 위해 자동 팩트체크 어플리케이션은 확장성과 자 동화의 요건에 대응하도록 설계되었습니다. 그러나 현재 애플리케이션 방법에는 기계 학습 분류 모델 성능을 개선하기 위해 가짜 뉴스 특징을 식별하는 포괄적인 다차원 데이터 세트가 없다. 이 문제를 해결하기 위해 본 연구논문에서는 사용자가 기사의 제목을 입력하면 데이터를 분류하는 Formb 챗봇을 제안했다. 이 연구 작업 에서 데이터 집합의 분류는 반복 신경망(RNN)과 장기 단기 기억(LSTM) 모델을 사용하여 수행되었다. 가짜 및 실제 뉴스 데이터 세트는 사전 처리되어 모델을 교육하는 데 사용됩니다. 저장된 모델은 지정된 입력 텍스트 의 신뢰성을 확인하기 위해 불일치 서버에 배포됩니다. Disconsid API는 python 파일을 chatbot으로 실행할 수 있는 액세스를 제공합니다. 분석 측면에서, 제안된 모델은 96.77%의 정확도로 CNN와 같은 기존 뉴럴 네트 워크 모델을 능가한다.