• Annual Conference of KIPS
• /
• 2017.04a
• /
• pp.910-913
• /
• 2017

데이터의 양이 증가하면서 인공신경망을 통한 데이터 분석 기술이 주목받고 있으며, 텍스트, 그림, 동영상 등에 이르기까지 다양한 종류의 데이터를 자동으로 분석하여, 번역기, 채팅봇, 그림 캡션 자동 생성 등에 대한 연구 및 서비스 개발에 활용되고 있다. 인공신경망 기반으로 수행된 많은 연구들이 공통적으로 가진 한계가 있는데, 그것은 은닉층에 대한 해석이 어렵다는 것이다. 가령, 입력층, 은닉층, 그리고 결과층으로 이루어진 인공신경망을 임의의 데이터로 학습시키면, 입력층과 은닝층 사이에 존재하는 행렬은 해당 데이터에 존재하는 패턴 정보를 내포하게 된다. 따라서, 행렬에 존재하는 패턴 정보를 직접 분석할 수 있다면, 인공신경망 결과물에 대한 해석이 가능할 뿐만 아니라 성능을 높이기 위해 어떤 조정이 필요한지에 대한 직관도 얻을 수 있을 것이다. 하지만, 이 행렬의 실체는 숫자로 이루어진 벡터이므로 사람이 직접 해석하는 것은 불가능하며, 지금까지 수행되어온 대부분의 인공신경망 연구들은 공통적으로 이러한 한계점을 가지고 있다. 본 연구는 데이터에 존재하는 패턴을 잡아내면서도 해석이 가능한 토픽 모델과 인공신경망의 결과물을 비교함으로써, 인공신경망 은닉층 해석에 대한 실마리를 찾기 위한 연구이다. 실험을 통해 토픽과 은닉층 패턴의 유사성을 검증하고, 향후 인공신경망 연구에서 은닉층에 대한 가능성을 논한다.

• Archives of design research
• /
• v.11 no.3
• /
• pp.47-55
• /
• 1998

Homepage design on the Web is incredibly growing fast as the integrative information method than any other media. At the beginning the homepage was designed by text mostly, however, it has been changed to use the multimedia. Design in the Web is changed by the development of computer technology. Rnal destination of Web is an user platform that consists of pixel of monitor screen. Monitor requires different approach from printing material because of the limited of range of presentation and the property of light. This paper has studied to find the possibilities of expression of type which can be the basic structure for HTML. By the case study of homepage, it has analyzed the kind of types, presentation method, the number of frequency, and variation of design.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.91-94
• /
• 2023

자연스러운 상호작용이 가능한 인공지능 에이전트를 개발하기 위해서는 언어적 표현뿐 아니라, 비언어적 표현 또한 고려되어야 한다. 본 논문에서는 한국어 발화문으로부터 비언어적 표현인 모션을 생성하는 연구를 소개한다. 유튜브 영상으로부터 데이터셋을 구축하고, Text to Motion의 기존 모델인 T2M-GPT와 이종 모달리티 데이터를 연계 학습한 VL-KE-T5의 언어 인코더를 활용하여 구현한 모델로 실험을 진행하였다. 실험 결과, 한국어 발화 텍스트에 대해 생성된 모션 표현은 FID 스코어 0.11의 성능으로 나타났으며, 한국어 발화 정보 기반 비언어 표현 정보 생성의 가능성을 보여주었다.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.2
• /
• pp.29-45
• /
• 2015

Systematic theory, concepts, and methodology for the biological evolution have been developed while patterns and principles of the evolution have been actively studied in the past 200 years. Furthermore, they are applied to various fields such as evolutionary economics, evolutionary psychology, evolutionary linguistics, making significant progress in research. In addition, existing studies have applied main biological evolutionary models to artifacts although such methods do not fit to them. These models are also limited to generalize evolutionary patterns of artifacts because they are designed in terms of a subjective point of view of experts who know well about the artifacts. Unlike biological organisms, because artifacts are likely to reflect the imagination of the human will, it is known that the theory of biological evolution cannot be directly applied to artifacts. In this paper, beyond the individual's subjective, the aim of our research is to present evolutionary patterns of a given artifact based on peeping the idea of the public. For this, we propose a text mining approach that presents a systematic framework that can find out the evolutionary patterns of a given artifact and then visualize effectively. In particular, based on our proposal, we focus mainly on a case study of mobile phone that has emerged as an icon of innovation in recent years. We collect and analyze review posts on mobile phone available in the domestic market over the past decade, and discuss the detailed results about evolutionary patterns of the mobile phone. Moreover, this kind of task is a tedious work over a long period of time because a small number of experts carry out an extensive literature survey and summarize a huge number of materials to finally draw a diagram of evolutionary patterns of the mobile phone. However, in this work, to minimize the human efforts, we present a semi-automatic mining algorithm, and through this research we can understand how human creativity and imagination are implemented. In addition, it is a big help to predict the future trend of mobile phone in business and industries.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.23 no.4
• /
• pp.374-390
• /
• 2021

Growth and yield of field vegetable crops would be affected by climate conditions, which cause a relatively large fluctuation in crop production and consumer price over years. The yield prediction system for these crops would support decision-making on policies to manage supply and demands. The objectives of this study were to compile literatures related to onion and garlic and to perform data-mining analysis, which would shed lights on the development of crop models for these major field vegetable crops in Korea. The literatures on crop growth and yield were collected from the databases operated by Research Information Sharing Service, National Science & Technology Information Service and SCOPUS. The keywords were chosen to retrieve research outcomes related to crop growth and yield of onion and garlic. These literatures were analyzed using text mining approaches including word cloud and semantic networks. It was found that the number of publications was considerably less for the field vegetable crops compared with rice. Still, specific patterns between previous research outcomes were identified using the text mining methods. For example, climate change and remote sensing were major topics of interest for growth and yield of onion and garlic. The impact of temperature and irrigation on crop growth was also assessed in the previous studies. It was also found that yield of onion and garlic would be affected by both environment and crop management conditions including sowing time, variety, seed treatment method, irrigation interval, fertilization amount and fertilizer composition. For meteorological conditions, temperature, precipitation, solar radiation and humidity were found to be the major factors in the literatures. These indicate that crop models need to take into account both environmental and crop management practices for reliable prediction of crop yield.

• Information Systems Review
• /
• v.26 no.1
• /
• pp.269-288
• /
• 2024

The main objective of this study is to empirically explore how the organizational culture influences financial performance of companies. To achieve this, 58 companies included in the KOSPI 200 were selected from an online job platform in South Korea, JobPlanet. In order to understand the organizational culture of these companies, data was collected and analyzed from 81,067 reviews written by current and former members of these companies on JobPlanet over a period of 9 years from 2014 to 2022. To define the organizational culture of each company based on the review data, this study utilized well-known text analysis techniques, namely Word2Vec and FastText analysis methods. By modifying, supplementing, and extending the keywords associated with the five organizational culture values (Innovation, Integrity, Quality, Respect, and Teamwork) defined by Guiso et al. (2015), this study created a new Culture Dictionary. By using this dictionary, this study explored which cultural values-related keywords appear most often in the review data of each company, revealing the relative strength of specific cultural values within companies. Going a step further, the study also investigated which cultural values statistically impact financial performance. The results indicated that the organizational culture focusing on innovation and creativity (Innovation) and on customers and the market (Quality) positively influenced Tobin's Q, an indicator of a company's future value and growth. For the indicator of profitability, ROA, only the organizational culture emphasizing customers and the market (Quality) showed statistically significant impact. This study distinguishes itself from traditional surveys and case analysis-based research on organizational culture by analyzing large-scale text data to explore organizational culture.

• Journal of The Korean Association of Information Education
• /
• v.19 no.3
• /
• pp.311-322
• /
• 2015

This study analyzed the trend and problems in the design of contents developed by teachers, and suggested strategies for improvement. It analyzed the contents ranked as the first level in the Research Competition on Educational Informatization for the last 3 years. Concerning the 8 types of instructional activities and the 6 types of knowledge acquisition, most contents took limited types(i.e., the individual tutoring type, the concept learning type and the principle learning type). In addition, when the contents were evaluated according to the quality certification criteria for educational software, it was found that the quality level of the design was low in many criteria. When the content analysis was applied for the in-depth analysis of design characteristics, various problems were found in the areas such as evaluation, feedback and learning objectives. Also other common problems were found in the design areas such as level-based differentiated learning, interaction between students and contents, presentation of text and narration, utilization of information on a student, screen design, the content level appropriate for students. In relation to the problems found from the analysis, some strategies for improvement were suggested concerning the following topics: question selection and guidance for evaluation, content and types of feedback, statement of learning objectives, selection of content, interaction, and screen design.

• The Journal of Society for e-Business Studies
• /
• v.25 no.1
• /
• pp.83-107
• /
• 2020

Selecting business charts based on the nature of the data and the purpose of the visualization is useful in business analysis. However, current visualization tools lack the ability to help choose the right business chart for the context. Also, soliciting expert help about visualization methods for every analysis is inefficient. Therefore, the purpose of this study is to propose an accessible method to improve business chart productivity by creating rules for selecting business charts from online published documents. To this end, Korean, English, and Chinese unstructured data describing business charts were collected from the Internet, and the relationships between the contexts and the business charts were calculated using TF-IDF. We also used a Galois lattice to create rules for business chart selection. In order to evaluate the adequacy of the rules generated by the proposed method, experiments were conducted on experimental and control groups. The results confirmed that meaningful rules were extracted by the proposed method. To the best of our knowledge, this is the first study to recommend customizing business charts through open unstructured data analysis and to propose a method that enables efficient selection of business charts for office workers without expert assistance. This method should be useful for staff training by recommending business charts based on the document that he/she is working on.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2000.11a
• /
• pp.66-69
• /
• 2000

웹이 등장하면서 지금까지 인터넷 상에서 텍스트와 이미지를 이용하여 정보를 표현하고 전달하는 방법이 가장 많이 사용되어왔다. 그러나 웹 관련 기술의 비약적인 발달과 네트워크 속도의 증가 및 인터넷의 급속한 보급으로 단순한 텍스트와 이미지 중심의 HTML 문서를 이용한 정보의 전달이 아닌 멀티미디어 데이터를 이용한 정보의 표현과 전달이 점차 증대되고 있다. 이에 따라 멀티미디어 데이터를 전송하기 위한 스트리밍 프로토콜도 등장하였다. 최근에는 컴퓨터의 성능 증가 및 네트워크 속도의 증가(초고속 통신 서비스의 보급)에 의해 멀티미디어 데이터의 전송이 가능하게 됨으로써 기존의 공중파나 CATV 방송국의 형태 지니고 인터넷 상에서 실시간 생방송 서비스와 VOD(Video On Demand) 서비스를 제공하는 인터넷 방송국이 급속하게 생겨나고 있다. (11) 인터넷 방송은 동영상과 오디오의 실시간 전달을 가능하게 하는 멀티미디어 스트리밍 기술과 멀티미디어를 실시 간으로 전송할 수 있는 실시간 전송 프로토콜을 기반으로 발전하고 있다. 인터넷 상에서 멀티미디어 스트리밍 서비스를 하는 대부분의 인터넷 방송은 스트리밍 서버로서 RealNetworks사의 RealSystem과 Microsoft사의 WMT(Windows Media Technologies)를 사용하고 있다. 본 논문은 Real Server와 WMT의 비교 분석을 통해 실시간 전송 프로토콜을 지원하고, 멀티미디어 스트리밍 기술을 지원하는 자바를 기반으로 한 분산 서버 구조의 스트리밍 서버, 서버간의 부하를 제어하는 미들웨어, 멀티미디어 스트림을 재생할 수 있는 클라이언트를 설계하고 구현한다.있다.구현한다. 이렇게 구현된 시스템은 전자 상거래, 가상 쇼핑몰, 가상 전시화, 또는 3차원 게임이나 가상교육 시스템과 같은 웹기반 응용프로그램에 사용될 수 있다.물을 보존·관리하는 것이 필요하다. 이는 도서관의 기능만으로는 감당하기 어렵기 때문에 대학정보화의 센터로서의 도서관과 공공기록물 전문 담당자로서의 대학아카이브즈가 함께 하여 대학의 공식적인 직무 관련 업무를 원활하게 지원하고, 그럼으로써 양 기관의 위상을 높이는 상승효과를 낼 수 있다.하여는, 인쇄된 일차적 정보자료의 검색방법등을 개선하고, 나아가서는 법령과 판례정보를 위한 효율적인 시스템을 구축하며, 뿐만 아니라 이용자의 요구에 충분히 대처할 수 잇는 도서관으로 변화되는 것이다. 이와 함께 가장 중요한 것은 법과대학과 사법연수원에서 법학 연구방법에 관한 강좌를 개설하여 각종 법률정보원의 활용 내지 도서관 이용방법에 관하여 교육하는 것이다.글을 연구하고, 그 결과에 의존하여서 우리의 실제의 생활에 사용하는 $\boxDr$한국어사전$\boxUl$등을 만드는 과정에서, 어떤 의미에서 실험되었다고 말할 수가 있는 언어과학의 연구의 결과에 의존하여서 수행되는 철학적인 작업이다. 여기에서는 하나의 철학적인 연구의 시작으로 받아들여지는 이 의미분석의 문제를 반성하여 본다. 것이 필요하다고 사료된다.크기에 의존하며, 또한 이러한 영향은 $(Ti_{1-x}AI_{x})N$ 피막에 존재하는 AI의 함량이 높고, 초기에 증착된 막의 업자 크기가 작을 수록 클 것으로 여겨진다. 그리고 환경의 의미의 차이에 따라 경관의 미학적 평가가 달라진 것으로

• Journal of Korean Library and Information Science Society
• /
• v.51 no.1
• /
• pp.179-200
• /
• 2020

This study aims to detect the knowledge structure of Korean studies using document co-citation analysis and text mining techniques. This study divided Korean corpus into two perspectives: Self-perceived and others' perceived Korean studies. To this end, we collected 10,929 humanities and social literature containing the word Korea or Korean as a keyword in the SCOPUS database. As a result of analysis, a total of 20 subdomains were found in the knowledge structure of self-perception, and a total of 14 subdomains were found in the knowledge structure of otherts' perception. Differences in Korean Studies between two are: First, the sub-area of self-perceived Korean studies is subdivided into more diverse areas than the sub-area of other-perceived Korean studies. Second the major areas in self-perceived Korean studies are customers and services, industrialization, multiculturalism, mental health, tourism, Korean language, environment, and cities. Others' perceptions of Korean Studies are grouped into domestic and foreign situations of Korea, Korean pop culture, Koreans as US immigrants, and Korean language. Finally, the common areas of self-perception and others' perception were mental health, tourism, Korean language, North-Korean defectors, and juvenile delinquency.