• Journal of the Korea Society of Computer and Information
• /
• v.28 no.1
• /
• pp.27-38
• /
• 2023

In this paper, we propose a conversational AI agent based on continual learning that can continuously learn and grow with new data over time. A continual learning-based conversational AI agent consists of three main components: Task manager, User attribute extraction, and Auto-growing knowledge graph. When a task manager finds new data during a conversation with a user, it creates a new task with previously learned knowledge. The user attribute extraction model extracts the user's characteristics from the new task, and the auto-growing knowledge graph continuously learns the new external knowledge. Unlike the existing conversational AI agents that learned based on a limited dataset, our proposed method enables conversations based on continuous user attribute learning and knowledge learning. A conversational AI agent with continual learning technology can respond personally as conversations with users accumulate. And it can respond to new knowledge continuously. This paper validate the possibility of our proposed method through experiments on performance changes in dialogue generation models over time.

• Journal of The Korean Association of Information Education
• /
• v.25 no.5
• /
• pp.665-672
• /
• 2021

Humanity in the 21st century is ushering in an era of practical use of AI. Until now, even though the industrial structure has been advanced, mankind has seen that the abstraction of knowledge production is only their own domain, but they have doubts about that belief. Therefore, this paper tried to examine the identity of modern humanity from the perspective of the result of knowledge accumulated from the past. These discussions were summarized and presented in a historical model called "Changes in the way of accumulating knowledge step by step" starting from the emergence of the earth and mankind. The first stage of this analytical model is the "accumulation of DNA knowledge" until the emergence of human intelligence on Earth. The second stage is the process of "accumulating civilized knowledge" by human biological intelligence, which has become capable of producing knowledge on its own. It is currently classified into three stages and it is considered that it is entering the stage of "accumulating mechanical knowledge" using AI technology. This paper proposes human history as such a step-by-step knowledge accumulation model and describes related discussions.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.8
• /
• pp.1103-1110
• /
• 2022

Due to the rapid development of AI-related industries, countless edge devices are working in the real world. Since data generated within the smart space consisted of these devices is beyond imagination, it is becoming increasingly difficult for edge devices to process. To solve this issue, EdgeCPS has appeared. EdgeCPS is a technology to support harmonious execution of various application services including AI applications through interworking between edge devices and edge servers, and augmenting resources/functions. Therefore, we propose a knowledge-sharing graph-based componentized AI application support system applicable to the EdgeCPS platform. The graph is designed to effectively store information which are essential elements for creating AI applications. In order to easily change resource/function augmentation under the support of the EdgeCPS platform, AI applications are operated as components. The application support system is linked with the knowledge graph so that users can easily create and test applications, and visualizes the execution aspect of the application to users as a pipeline.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.301-305
• /
• 2023

최근 등장한 Large Language Models (LLM)은 자연어 처리 분야에서 눈에 띄는 성과를 보여주었지만, 주로 영어 중심의 연구로 진행되어 그 한계를 가지고 있다. 본 연구는 사전 학습된 LLM의 언어별 지식 전이 가능성을 한국어를 중심으로 탐구하였다. 이를 위해 한국어와 영어로 구성된 코드 스위칭 코퍼스를 구축하였으며, 기본 모델인 LLAMA-2와 코드 스위칭 코퍼스를 추가 학습한 모델 간의 성능 비교를 수행하였다. 결과적으로, 제안하는 방법론으로 학습한 모델은 두 언어 간의 희미론적 정보가 효과적으로 전이됐으며, 두 언어 간의 지식 정보 연계가 가능했다. 이 연구는 다양한 언어와 문화를 반영하는 다국어 LLM 연구와, 소수 언어를 포함한 AI 기술의 확산 및 민주화에 기여할 수 있을 것으로 기대된다.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.31-35
• /
• 2023

최근에는 질의응답(Question Answering, QA) 분야에서 거대 언어 모델(Large Language Models, LLMs)의 파라미터에 내재된 지식을 활용하는 방식이 활발히 연구되고 있다. Open Domain QA(ODQA) 분야에서는 기존에 정보 검색기(retriever)-독해기(reader) 파이프라인이 주로 사용되었으나, 최근에는 거대 언어 모델이 독해 뿐만 아니라 정보 검색기의 역할까지 대신하고 있다. 본 논문에서는 거대 언어 모델의 내재된 지식을 사용해서 질의 응답에 활용하는 방법을 제안한다. 질문에 대해 답변을 하기 전에 질문과 관련된 구절을 생성하고, 이를 바탕으로 질문에 대한 답변을 생성하는 방식이다. 이 방법은 Closed-Book QA 분야에서 기존 프롬프팅 방법 대비 우수한 성능을 보여주며, 이를 통해 대형 언어 모델에 내재된 지식을 활용하여 질의 응답 능력을 향상시킬 수 있음을 입증한다.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.409-414
• /
• 2023

최근 지식 기반 대화 생성에 많은 연구자가 초점을 맞추고 있다. 특히, 특정 도메인에서의 작업 지향형 대화 시스템을 구축하는 것은 다양한 도전 과제가 있으며, 이 중 하나는 거대 언어 모델이 입력과 관련된 지식을 활용하여 응답을 생성하는 데 있다. 하지만 현재 거대 언어 모델은 작업 지향형 대화에서 단순히 정보를 열거하는 방식으로 응답을 생성하는 경향이 있다. 이 논문에서는 전문 지식과 대화 정책 예측 모델을 결합한 프롬프트를 제시하고 작업 지향형 대화에서 사용자의 최근 입력에 대한 정보 제공 및 일상 대화를 지원하는 가능성을 탐구한다. 이러한 새로운 접근법은 모델 파인튜닝에 비해 비용 측면에서 효율적이며, 향후 대화 생성 분야에서 발전 가능성을 제시한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.1039-1040
• /
• 2023

본 논문에서는 무형 유산 DX 플랫폼의 AR(Augmented Reality) 기술과 ChatGPT 를 결합하여 전문가들의 지식을 보존하고 효과적으로 전달하는 시스템을 제안한다. 특히, 고령화 사회에서 은퇴한 전문가들의 지식이 소실될 위험을 방지하며, 사용자들의 교육 경험을 향상시키는 방법을 모색한다.

• Journal of Digital Convergence
• /
• v.18 no.8
• /
• pp.217-224
• /
• 2020

The purpose of this study is to discuss implications for government policy aimed at increasing public's intention to accept AI. Knowledge regarding AI and feelings regarding AI were found to influence acceptance to intention towards AI. Hierarchical regression analysis was then conducted to explore the moderation effect of confidence in government on knowledge and feelings regarding AI. Results showed that as advanced knowledge regarding AI has a positive influence on acceptance intention towards AI and negative feelings regarding AI has a negative influence on acceptance intention towards AI. Feelings regarding AI had the highest impact on acceptance intention towards AI, followed by confidence in government and knowledge regarding AI. Results also revealed that a high level of confidence in government regulations was associated with greater acceptance intention towards AI and a low level of confidence in government regulations acceptance intention towards AI was more influenced by feelings regarding AI than by knowledge regarding AI. Furthermore, religion had a significant influence on acceptance intention towards AI, which provides one insightful direction for future research.

• The Journal of the Korea Contents Association
• /
• v.20 no.8
• /
• pp.350-358
• /
• 2020

The objective of this empirical study is to examine the mediating and moderating effects of perceived risks on the relationship between knowledge, feelings and acceptance intention towards AI. Subjects in their teens to forties were surveyed and the final sample comprised 1,969 subjects. Data were analyzed using Mediation using Multiple Regression and Moderated Multiple Regression. Results showed that people's knowledge and feelings towards AI affected their acceptance intention of AI. Results also showed that the perceived risks of AI partially mediated and moderated the relationship between feelings and acceptance intention towards AI and moderated but not mediated the relationship between knowledge and acceptance intention towards AI. Overall, these results suggest that people's perceived risks of AI are associated more strongly with their feelings towards AI than their knowledge towards AI. Implications and directions for future research were discussed in relation to increasing general population's acceptance intention towards AI.

• Journal of the Korean Society for Precision Engineering
• /
• v.4 no.4
• /
• pp.6-12
• /
• 1987

지능화(Intelligence) 전반적인 경향 : 생산분야에 있어서의 지능화는 인공지능(AI)기법이 활용되면서 부터 관심을 모으게 되었다. AI기법이라고 하여도 그에 관련된 기술은 Software기술, 지식표현과 같은 기초기술 에서 부터, 지식공학, 화상인식 이해, 자연언어 이해, 음성이해 등의 응용기술에 이르기까지 다양한 분야를 포함하고 있다. 여기서는, 최근 실용화 면에서 주목을 받고 있는 지식공학의 실용 또는 그 응용시스템인 expert system(전문가시스템)에 주목하면서, 설계와 공정설계 중에 최근의 AI활용과 동향을 요약해 보기로 한다.