• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.558-562
• /
• 2023

최근 ChatGPT의 등장으로 텍스트 생성 모델에 대한 관심이 높아지면서, 텍스트 생성 태스크의 성능평가를 위한 지표에 대한 연구가 활발히 이뤄지고 있다. 전통적인 단어 빈도수 기반의 성능 지표는 의미적인 유사도를 고려하지 못하기 때문에, 사전학습 언어모델을 활용한 지표인 BERTScore를 주로 활용해왔다. 하지만 이러한 방법은 사전학습 언어모델이 학습한 데이터에 존재하는 편향으로 인해 공정성에 대한 문제가 우려된다. 이에 따라 한국어 사전학습 언어모델의 편향에 대한 분석 연구가 필요한데, 기존의 한국어 사전학습 언어모델의 편향 분석 연구들은 사회에서 생성되는 다양한 속성 별 편향을 고려하지 못했다는 한계가 있다. 또한 서로 다른 언어를 기반으로 하는 사전학습 언어모델들의 속성 별 편향을 비교 분석하는 연구 또한 미비하였다. 이에 따라 본 논문에서는 한국어 사전학습 언어모델의 속성 별 편향을 비교 분석하며, 영어 사전학습 언어모델이 갖고 있는 속성 별 편향과 비교 분석하였고, 비교 가능한 데이터셋을 구축하였다. 더불어 한국어 사전학습 언어모델의 종류 및 크기 별 편향 분석을 통해 적합한 모델을 선택할 수 있도록 가이드를 제시한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10a
• /
• pp.118-120
• /
• 2003

현재 웹의 패러다임은 정적인 웹 페이지들의 모임에서부터 기업이 비즈니스를 더 신속하고 효율적으로 수행할 수 있게 하는 웹 서비스라는 동적인 웹으로 변화하고 있다. 이를 위하여 XML을 기반으로 하는 SOAP WSDL, UDDI, WSFL, XLANG, BPML. WSCL, WSCI 및 BPEL4WS 등과 같은 다양한 표준들과 더 불어 시맨틱 웹과 온톨로지를 기반으로 하는 컴퓨터에 의해서 자동으로 이해될 수 있는 방법으로 웹 서비스를 언급하는 UPML. WSMF 및 DAML-S와 같은 연구들이 최근에 대두되고 있다. 우리는 이 논문에서 시맨틱 웹을 기반으로 하는 지능형 웹 서비스를 가능하게 하기 위하여 새로운 시맨틱 웹 서비스 마크업 언어를 설계하기 위한 전략들을 제시한다. 이를 위하여 먼저 지능형 웹 서비스를 위한 서비스 마크업 언어의 요구사항들을 분석하고, 기존 웹 서비스를 위한 표준들과 연구들을 비교하고 분석한다. 그리고 효과적인 지능형 웹 서비스를 위한 새로운 마크업 언어의 설계를 위한 방법론을 제시한다. 이 논문의 연구결과들은 시맨틱 웹을 기반으로 하는 지능형 웹 서비스를 위한 마크업 언어의 필요성 및 기존 노력들의 문제점들을 지적함으로써 보다 개선된 웹 서비스를 위한 근간으로 사용될 수 있다. 뿐만 아니라 기존 연구들이 추후 보완해나가야 할 사항들을 언급함으로써 보다 향상된 지능형 웹 서비스의 실현을 가능하게 한다.

• Trans-
• /
• v.16
• /
• pp.107-133
• /
• 2024

Narratives are fundamental to human cognition and social culture, serving as the primary means by which individuals and societies construct meaning, share experiences, and convey cultural and moral values. The field of artificial intelligence, which seeks to mimic human thought and behavior, has long studied story generation and story understanding, and today's Large Language Models are demonstrating remarkable narrative capabilities based on advances in natural language processing. This situation raises a variety of changes and new issues, but a comprehensive discussion of them is hard to find. This paper aims to provide a holistic view of the current state and future changes by exploring the intersections and interactions of human and AI narrative intelligence. This paper begins with a review of multidisciplinary research on the intrinsic relationship between humans and narrative, represented by the term Homo narrans, and then provide a historical overview of how narrative has been studied in the field of AI. This paper then explore the possibilities and limitations of narrative intelligence as revealed by today's Large Language Models, and present three philosophical challenges for understanding the implications of AI with narrative intelligence.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.04a
• /
• pp.206-209
• /
• 2005

본 논문에서는 국내에서 상용화되고 있는 삼성전자의 SPC-10의 IEC1131-3 표준 언어(LD, SFC, FBD, ST)가 상호 호환성을 갖도록 목적 언어인 IL로 변환하여 컴파일 될 수 있는 IL 컴파일러를 개발하고자 한다. 개발하려는 IL 컴파일러는 IL언어의 활용성을 매우 높여주며, 실제네트워크 디바이스에 다운로드 하여 사용할 수 있는 IL 언어를 생성하고 실행시키는 점이 특징이다. 이 시스템은 SPC-10에서 많이 사용되는 언어를 패턴인식에 의해 클러스터링하여 자동적으로 IL 언어로 변환이 되며, 컴파일에 의해 PLC 프로그램의 동작이 가능하다. 이 시스템에서는 또한 사용자가 발생할 수 있는 문법오류는 물론 논리오류를 지능적 에이전트에 의해 검색하여 수정함으로써 최적화된 환경에서 PC 기반 제어가 가능하도록 해준다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.12 no.5
• /
• pp.417-423
• /
• 2002

In this paper, we propose intelligent steering control system that apply LIBL(Linguistic Instruction Based Learning) method to steering system of ship and take the place of process that linguistic instruction such as officer s steering instruction is achieved via ableman. We embody ableman s suitable steering manufacturing model using fuzzy inference rule by specific method of study, and apply LIBL method to present suitable meaning element and evaluation rule to steering system of ship, embody intelligent steering gear control system that respond more efficiently on officer s linguistic instruction. We presented evaluation rule to constructed steering manufacturing model based on ableman s experience, and propose rudder angle for steering system, compass bearing arrival time, meaning element of stationary state, and correct ableman manufacturing model rule using fuzzy inference. Also, we apply LIBL method to ship control simulator and confirmed the effectiveness.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2002.12a
• /
• pp.93-97
• /
• 2002

In this paper, we propose intelligent steering control system that apply LIBL(Linguistic Instruction Based Learning) method to steering system of ship and take the place of process that linguistic instruction such as officer's steering instruction is achieved via ableman. We embody ableman's suitable steering manufacturing model using fuzzy inference rule by specific method of study, and apply LIBL method to present suitable meaning element and evaluation rule to steering system of ship, embody intelligent steering gear control system that respond more efficiently on officer's linguistic instruction. We presented evaluation rule to constructed steering manufacturing model based on ableman's experience, and propose rudder angle for steering system, compass bearing arrival time, meaning element of stationary state, and correct ableman manufacturing model rule using fuzzy inference. Also, we apply LIBL method to ship control simulator and confirmed the effectiveness.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.2
• /
• pp.207-215
• /
• 2004

The PLC programming by IEC1131-3 is hard to handle to ordinary users as well as professionals. Also it has not a generality, so that it couldn't be debugging some logic errors easily. In order to be adapted for such environment, In this paper, I have developed the ISPLC(Intelligent Agent System based Software Programmable Logic Controller). In ISPLC system, LD programmed by a user is converted to the C code which can be used in a commercial editor such as Visual C++. The detection of logical errors in C code is more effective than PLC programming itself. ISPLC provides the GUI-based interface in web environment and an easy programming platform to such beginners as well as professionals. The study of code conversion of LD to IL as well as IL to C is firstly tried in the world as well as KOREA. To show the effectiveness of the developed system, I applied it to a practical case, a real time traffic control system. ISPLC is minimized the error debugging and programming time owing to be supported by windows application programs.

• Journal of Genetic Medicine
• /
• v.7 no.1
• /
• pp.45-52
• /
• 2010

Purpose: The objectives were to examine following 2 questions related to cognitive profile for the children with Williams syndrome (WS); 1) Is there a significant advantage for verbal IQ over performance IQ in WS?; 2) Is there selective impairment in visuospatial ability in the children with WS? Materials and Methods: Five children with WS with the age of $90.86{\pm}20.73$ months were compared with 12 children with Prader-Willi syndrome (PWS) or Down syndrome (DS) with comparable age and IQ. Results: All 5 children with WS showed intellectual disability whose mean scaled scores were $15.71{\pm}9.27$ in verbal subtests and $14.29{\pm}7.50$ in performance subtests, which did not show significant difference. There was no significant difference in the total sum of scaled scores of verbal subtests among WS, PWS and DS. There was no selective impairment in subtests which represented visuospatial tasks for the children with WS. However, the scaled score of object assembly was significantly lower in WS ($2.29{\pm}0.95$) compared to that of PWS ($4.75{\pm}2.77$; P <0.05). Conclusion: The general notion that the children with WS would be relatively strong in verbal function when compared with their overall cognitive function was not observed in this study. The verbal function of the children with WS was not better when compared to the children with DS or PWS. There was no selective impairment of visuospatial function in the children with WS at this age. However, the visuospatial function was significantly low in the children with WS only when compared to the children with PWS.

• Journal of the Korea Convergence Society
• /
• v.10 no.6
• /
• pp.251-257
• /
• 2019

The purpose of this study was to inquire into performance of Wechsler Intelligence Scale for Children based on HOME environment stimuli and age. For the purpose of this, a trained tester carried out K-WISC-IV oriented to 128 children aged 6-16, and this study conducted HOME to mothers. Regarding participants' full Scale IQ and 4 index score(Verbal Comprehension, Perceptual Reasoning, Working Memory, Processing Speed) and HOME, they were analyzed by age group of two: a group of children aged 6-10 and another group of children aged 11-16. According to the analysis results, both age groups showed a significant positive correlation between Full Scale IQ, Verbal Comprehension Index, and HOME environment stimuli. The results of this study were discussed and proposed the necessity of studies in relation to performance of diverse subtests in accordance with amendment of future intelligence scales.

• Journal of KIISE:Software and Applications
• /
• v.37 no.5
• /
• pp.370-377
• /
• 2010

An Intelligent service robot performs various complex tasks in dynamic environment, providing useful intelligent services for human users. The robot needs to continuously monitor dynamically changing environment and reactively choose the best behavior for the changing context. The selected behaviors may include nondeterministic or parallel actions. In this paper, we present a structured reactive robot programming language, SPRIT that is based on Structured Circuit Semantics (SCS). SPRIT is fully implemented as a task executor and tested for reactive robot tasks in dynamic environment to show that it can be used to explicitly represent and effectively implement the complex reactive behaviors of intelligent robot systems.