• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.485-490
• /
• 2023

언어 모델 발전에 따라 사람과 유사하게 글을 생성하고 태스크를 수행하는 LLM들이 등장하고 있다. 하지만 아직까지도 기계와 사람의 수행 과정에 초점을 맞추어 차이점을 드러내는 연구는 활성화되지 않았다. 본 연구는 자연어 이해 및 생성 태스크 수행 시 기계와 인간의 수행 과정 차이를 밝히고자 하였다. 이에 이해 태스크로는 문법성 판단, 생성 태스크로는 요약 태스크를 대상 태스크로 선정하였고, 기존 주류 사전학습 모델이었던 transformer 계열 모델과 LLM인 ChatGPT 3.5를 사용하여 실험을 진행하였다. 실험 결과 문법성 판단 시 기계들이 인간의 언어적 직관을 반영하지 못하는 양상을 발견하였고, 요약 태스크에서는 인간과 기계의 성능 판단 기준이 다름을 확인하였다.

• Journal of the Microelectronics and Packaging Society
• /
• v.30 no.3
• /
• pp.11-19
• /
• 2023

Recently with the rapid advancement of artificial intelligence (AI) technologies such as Chat GPT, AI semiconductors have become important. AI technologies require the ability to process large volumes of data quickly, as they perform tasks such as big data processing, deep learning, and algorithms. However, AI semiconductors encounter challenges with excessive power consumption and data bottlenecks during the processing of large-scale data. Thus, the latest packaging technologies are required for AI semiconductor computations. In this study, the authors have described packaging technologies applicable to AI semiconductors, including interposers, Through-Silicon-Via (TSV), bumping, Chiplet, and hybrid bonding. These technologies are expected to contribute to enhance the power efficiency and processing speed of AI semiconductors.

• Journal of Practical Engineering Education
• /
• v.16 no.2
• /
• pp.143-150
• /
• 2024

With the development of AI, multimedia tools in education offer personalized learning environments, which foster individual competencies. This study aims to examine the effectiveness of smart education as perceived by learners through a case study of university writing classes utilizing multimedia learning tools, and to explore potential applications. To achieve this, a writing course incorporating various multimedia tools to promote interaction was designed and implemented over the course of one semester, targeting 42 university students. Through the semester, student reactions and survey results were analyzed to investigate the effects and satisfaction levels regarding the use of multimedia learning tools in writing instruction as perceived by students. The analysis revealed that multimedia-assisted writing classes effectively fostered learners' autonomy by focusing on individual needs, while also promoting interaction and encouraging spontaneous participation. Students reported recognizing the presence of diverse perspectives by comparing and communicating about each other's writing, leading to an expansion of their own thinking. In using ChatGPT, it was found that students attempted to refine their questions until they obtained the desired answers. They reported that this process deepened their understanding of the essence of the questions. These benefits led to results of high levels of students' active class engagement and satisfaction. This study contributes foundational and empirical data regarding the effectiveness and potential applications of learner-centered smart education as part of fourth industrial revolution integration research.

• Journal of the korean academy of Pediatric Dentistry
• /
• v.51 no.3
• /
• pp.299-309
• /
• 2024

This study aimed to assess the competency of artificial intelligence (AI) in pediatric dentistry and compare it with that of dentists. We used open-source data obtained from the Korea Health Personnel Licensing Examination Institute. A total of 32 item multiple-choice pediatric dentistry exam questions were included. Two AI-based chatbots (ChatGPT 3.5 and Gemini) were evaluated. Each chatbot received the same questions seven times in separate chat sessions initiated on April 25, 2024. The accuracy was assessed by measuring the percentage of correct answers, and consistency was evaluated using Cronbach's alpha coefficient. Both ChatGPT 3.5 and Gemini demonstrated similar accuracy, with no significant differences observed between them. However, neither chatbot achieved the minimum passing score set by the Pediatric Dentistry National Examination. However, both chatbots exhibited acceptable consistency in their responses. Within the limits of this study, both AI-based chatbots did not sufficiently answer the pediatric dentistry exam questions. This finding suggests that pediatric dentists should be aware of the advantages and limitations of this new tool and effectively utilize it to promote patient health.

• International Journal of Advanced Culture Technology
• /
• v.12 no.1
• /
• pp.305-310
• /
• 2024

The purpose of this study is to implement game content within the generative artificial intelligence module Chat-GPTs, grounded in the humanistic discourse of self-reflection. This content aims to empower the dignity of the silver generation, which has been marginalized by digital technology. Simultaneously, we intend to prototype a digital psychotherapeutic tool. The development of a flexible device that adapts to the silver generation's living environment and temporal constraints is also part of our plan. However, there are still few commercially available products, and digital therapeutics developed in the form of content are virtually nonexistent. The goal is to implement game content that allows the elderly, who have been marginalized by digital technology, to find their true dignity. Simultaneously, we plan to commercialize a prototype of digital psychotherapy that can flexibly adapt to the range of living environments and time constraints of the elderly. This study has been extended based on the game content 'Daily Run' created by Hyein Kwon, an undergraduate student at Kyungil University.

• Journal of Christian Education in Korea
• /
• v.77
• /
• pp.171-201
• /
• 2024

Purpose of Research : Since ChatGPT's 2022 release, the educational sector faces mixed reactions to generative AI, sparking innovation but raising concerns about student cognition and communication. While Christian colleges employ AI reflecting their values, secular institutions stress ethical usage. This study explores ethical AI use in these settings, aiming to integrate findings into educational practices. Research content and method : Analyzing AI use and ethics guidelines from 50 domestic and international universities, differences between Christian and secular institutions were explored. Data was categorized, conceptualized via open coding, and components were identified through axial coding. The importance of components for Christian colleges' AI guides was assessed based on the initial data and previous research, leading to the development of tailored AI utilization components for Christian universities. Conclusion : Studies revealed secular institutions have six AI guide components, while Christian colleges found seven in both utilization and ethics guides, focusing on truthfulness, responsibility, and diversity. Emphasizing the need for ethical AI use in Christian colleges, the findings advocate developing AI ethics guidelines to aid marginalized groups and establish a new educational paradigm through further research.

• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.457-460
• /
• 2021

대화 시스템은 인공지능과 사람이 자연어로 의사 소통을 하는 시스템으로 크게 목적 지향 대화와 일상대화 시스템으로 연구되고 있다. 목적 지향 대화 시스템의 경우 날씨 확인, 호텔 및 항공권 예약, 일정 관리 등의 사용자가 생활에 필요한 도메인들로 이루어져 있으며 각 도메인 별로 목적에 따른 시나리오들이 존재한다. 이러한 대화는 사용자에게 명확한 발화을 제공할 수 있으나 자연스러움은 떨어진다. 일상 대화의 경우 다양한 도메인이 존재하며, 시나리오가 존재하지 않기 때문에 사용자에게 자연스러운 발화를 제공할 수 있다. 또한 일상 대화의 경우 검색 기반이나 생성 기반으로 시스템이 개발되고 있다. 검색 기반의 경우 발화 쌍에 대한 데이터베이스가 필요하지만, 생성 기반의 경우 이러한 데이터베이스가 없이 모델의 Language Modeling (LM)으로 부터 생성된 발화에 의존한다. 따라서 모델의 성능에 따라 발화의 품질이 달라진다. 최근에는 사전학습 모델이 자연어처리 작업에서 높은 성능을 보이고 있으며, 일상 대화 도메인에서도 역시 높은 성능을 보이고 있다. 일상 대화에서 가장 높은 성능을 보이고 있는 사전학습 모델은 Auto Regressive 기반 생성모델이고, 한국어에서는 대표적으로 KoGPT2가 존재한다. 그러나, KoGPT2의 경우 문어체 데이터만 학습되어 있기 때문에 대화체에서는 낮은 성능을 보이고 있다. 본 논문에서는 대화체에서 높은 성능을 보이는 한국어 기반 KoDialoGPT2를 개발하였고, 기존의 KoGPT2보다 높은 성능을 보였다.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.258-262
• /
• 2023

과제 지향 대화 시스템은 발화 의도 및 요구사항을 파악하여 사용자가 원하는 과제를 달성한다는 점에서 유용하다. 대화 상태 추적은 과제 지향 대화 시스템의 핵심 모듈이며, 최근에는 텍스트뿐만 아니라 시각 정보까지 활용하여 대화 상태를 추적하는 멀티모달 대화 상태 추적 연구가 활발히 진행되는 중이다. 본 논문에서는 멀티모달 공황장애 상담 대화 속 내담자의 상태를 추적하는 과제를 제안하였다. ChatGPT를 통한 멀티모달 공황장애 상담 과제 지향 대화 데이터셋 구축 프레임워크와, 구축한 데이터셋의 품질을 증명하기 위한 분석도 함께 제시하였다. 사전학습 언어 모델인 GPT-2를 벤치마크 데이터셋에 대해 학습한 성능을 측정함으로써 향후 멀티모달 대화 추적 성능이 능가해야 할 베이스라인 성능을 제시하였다.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.707-710
• /
• 2023

관계추출은 문장 내 두 개체 간의 의미적 관계를 추론하는 자연어분석 태스크이다. 딥러닝의 발전과 더불어 관계추출은 BERT 계열의 이해형 언어모델을 이용하였다. 그러나, ChatGPT의 혁신적인 등장과 함께, GPT계열의 생성형 언어모델에 대한 연구가 활발해졌다. 본 논문에서는 소규모의 생성형 언어모델(Kebyt5)을 이용하여 관계추출 성능개선을 위한 프롬프트 구성 및 생각의 사슬(CoT) 학습 방법을 제안한다. 실험결과 Kebyt5-large 모델에서 CoT 학습을 수행하였을 경우, Klue-RoBERTa-base 모델보다 3.05%의 성능개선이 있었다.

• Electronics and Telecommunications Trends
• /
• v.39 no.2
• /
• pp.54-65
• /
• 2024

This paper examines the trends of AI Foundation Model development and the competition to lead the related ecosystem, which have been rapidly unfolding since the emergence of ChatGPT, focusing on big tech companies in the United States. Based on this understanding of background knowledge, I analyzed and presented the main contents of the initial report reviewed by the UK competition authority, CMA, on potential competition issues that may arise in the process of innovations resulting from FM development. In addition, the trend and background of the CMA's investigation into the OpenAI-Microsoft partnership, whose importance has recently been highlighted, were also explained. It is expected that a reasonable domestic policy plan will be established by referring to these UK policy trends and monitoring & analyzing domestic industries.