생태환경 이론을 기반으로한 공간디자인 분류 모델 연구 - 중국 취락 공간을 중심으로 -

정조룡,이상해,최경란 한국공간디자인학회 2023 한국공간디자인학회논문집 Vol.18 No.8

(연구배경 및 목적) 현재 유엔과 전 세계는 생태환경 취락의 지속 가능한 발전을 중시하고 있으며, 2050년을 목표로‘인간과 자연의 조화로운 공생’을 전면적으로 실현하기로 공동으로 합의하였다. 그러나 오늘날 생태환경 취락 발전에는 여전히 많은 문제가 존재하며 취락 발전이 직면한 문제를 해결하기 위한 시스템적이고 종합적인 연구와 대책이 부족하다. 본 논문은 생태환경 사상 및 취락 공간의 디자인에 대한 공통 관점을 기반으로 하여 취락의 ‘생태-문화-생활-생산’ 공간의 디자인 시스템 분류 모델의 구축 방안을 연구한 것이다. 이를 통해 현재의 생태환경 취락이 당면한 문제를 전면적이고 시스템적으로 해결하기 위한 이론과 실천적인 근거를 제공할 것으로 기대한다. (연구방법) 첫째, 세계 및 중국의 대표적인 선행연구 문헌에서 출발하여 CAJViewerv 8.0 소프트웨어를 사용하여 키워드 데이터를 추출하고 순위를 매겨 생태 환경과 취락 공간의 디자인 지표가 되는 18가지 키워드 공간을 선별하였다. 둘째, 중국 학술 데이터베이스(CNKI)에 수록된 학술 문헌을 데이터 소스로, 그리고 CNKI 출판물의 통계적 분석과 시각화 분석 방법을 사용하여 취락 연구 맥락에 대한 객관성 검증 및 전체적 분석을 수행하여 지표가 되는 18가지 키워드 공간이 현재 이론 및 실천의 추세이자 방향이라고 결론을 내렸다. (결과) 18가지 지표 키워드 공간을 기반으로 취락 공간의 본체가 되는 이론에 근거하여 취락의 ‘생태-문화-생활-생산’ 공간의 디자인 시스템 분류 모델의 구축 방안을 제안하였다. 취락의 ‘생태-문화-생활-생산’ 공간의 디자인 시스템 분류 모델을 바탕으로 생태환경 취락 공간의 디자인 주체 기능의 구축을 제안하였다. (결론) 문헌 분석법과 CNKI의 출판물의 통계자료에 대한 분석을 통해 현재 생태환경 취락 연구에서 추출한 18가지 키워드의 지표 공간을 시스템적으로 분석하고, 생산-생활-생태 3중 공간 재구성 모델을 기반으로 문화 공간 요소를 추가하여 생태환경 취락 ‘생태-문화-생활-생산’ 공간의 디자인 시스템 분류 모델을 구축할 것을 제안하였다. 이 모델의 제안은 취락 공간의 디자인 생태계의 지속 가능성, 문화 계승 및 보호, 거주 환경의 개선 및 경제 및 사회 발전에 도움을 줄 것이다. 이를 통해 현재 취락 공간을 구분하기 어려운 상황에서 벗어나 생태환경 취락 공간을 위한 시스템적이고 전면적이며 표준화된 이론적 기반 및 실천적 설계 방향을 제시하고 중국 생태환경 취락의 발전을 촉진하며 취락 공간의 디자인에 관한 이론 및 실천 지침을 제공할 것으로 기대된다. (Background and Purpose) Currently, the United Nations and the world are focusing on the sustainable development of ecological environment villages and have jointly agreed to fully realize 'harmonious symbiosis between humans and nature' with the goal of 2050. However, there are still many problems in the development of ecological environments today, and systematic and comprehensive research and measures to solve the problems faced by village development are lacking. Based on a common perspective on the ecological environment idea and the design of the village space, this paper studied the construction of a design system classification model for the 'eco-culture-life-production' space of the village. Through this, it is expected to provide a theoretical and practical basis for solving the problems facing the current ecological environment fully and systematically. (Method) First, starting from the world and China's leading prior research literature, keywords data was extracted and ranked using CAJViewerv 8.0 software to select 18 keyword spaces that are design indicators of ecological environments and village spaces. Second, objectivity verification was conducted and an analysis of the context of village research was carried out using statistical analysis and visualization analysis methods of CNKI publications as a data source which academic literature contained in the Chinese Academic Database (CNKI). Finally, the conclusion that the 18 keyword spaces that serve as indicators are the current trend and direction of theory and practice was presented. (Results) Based on the theory that becomes the main body of the village space based on the 18 indicator keyword spaces, a plan to establish a design system classification model for the 'eco-culture-life-production' space of the village was proposed. Based on the design system classification model of the 'eco-culture-life-production' space of the village, it was proposed to establish the design subject function of the ecological environment village space. (Conclusions) Through an analysis of literature and statistical data from CNKI's publications, it was proposed to systematically analyze the surface space of 18 keywords extracted from the current ecological environment study and to establish a design system classification model for the ecological environment villages 'eco-culture-life-production' space by adding cultural space elements based on the production-life-ecological triple spatial reconstruction model. The proposal of this model will help the sustainability of the design ecosystem of village space, inherit and protect culture, the improvement of living environment, and economic and social development. This is expected to present a systematic, comprehensive, and standardized theoretical basis and practical design direction for ecological environment village spaces, promote the development of Chinese ecological environment villages, and provide theoretical and practical guidelines for the design of village spaces.

BRM 정비를 통한 기록관리기준표 개선사례: 서울시 BRM 및 기록관리기준표 정비사례를 중심으로

이세진,김화경 한국기록학회 2016 기록학연구 Vol.0 No.50

Unlike other government agencies, the city of Seoul experienced a three-year gap between the establishment of a function classification system and the introduction of a business management system. As a result, the city has been unable to manage the current status of the function classification system, and this impeded the establishment of standards for records management. In September 2012, the Seoul Metropolitan Government integrated the department in charge of the standard sheet for record management with the department of function classification system into a new department: “Information Disclosure Policy Division.” This new department is mainly responsible for record management and information disclosure, and taking this as an opportunity, the city government has pushed ahead with the maintenance project on BRM and Standards for Record Management (hereby “BRM maintenance project”) over the past two years, from 2013 to 2014. The study was thus conducted to introduce the case for the improvement of standards for record management through the BRM maintenance project by mainly exploring the case of Seoul. During the BRM maintenance project, Seoul established a unique methodology to minimize the gap between the operation of a business management system and the burden of the person in charge of the BRM maintenance project. Furthermore, after the introduction of the business management system, the city government developed its own processes and applied the maintenance result to the system in close cooperation with the related departments, despite the lack of precedence on the maintenance of the classification system. In addition, training for the BRM managers of the department has taken place twice —before and after the maintenance—for the successful performance of the BRM maintenance project and the stable operation of the project in the future. During the period of maintenance, newsletters were distributed to all employees in an effort to induce their active participation and increase the importance of records management. To keep the performance of the maintenance project and to systematically manage BRM in the future, the city government has mapped out several plans for improvement: to apply the “BRM classification system of each purpose” to the service of the“Seoul Open Data Plaza”; to reinforce the function for task management in the business management system; and to develop the function of a records management system for the unit tasks. As such, the researchers hope that this study would serve as a helpful reference so that the organizations-which had planned to introduce BRM or to perform the maintenance project on classification system-experience fewer trials and errors. 서울시는 다른 기관과 달리 기능분류시스템 구축과 업무관리시스템도입 사이에 3년의 시간차가 발생하여 기능분류체계의 현행관리가 되지 못했고, 이것은 기록관리기준표를 수립하지 못하는 문제로 이어졌다. 2012년 9월 서울시는 기록관리와 정보공개를 전담하는 정보공개정책과의 신설을 통해 기능분류시스템과 기록관리기준표의 운영부서를통합하였다. 이를 계기로 2013~2014년 2년에 걸쳐 ‘BRM 및 기록관리기준표 정비사업(이하 BRM 정비사업)’을 추진하게 되었다. 본 연구는 서울시 사례를 중심으로 BRM 정비를 통한 기록관리기준표 개선사례를소개하고자 한다. 서울시 BRM 정비사업은 업무담당자의 부담과 업무관리시스템의 운영공백을 최소화하기 위한 정비방법론을 수립하였다. 업무관리시스템 도입 후 분류체계를 정비하는 선례가 없는 상황에서관련부서와의 긴밀한 협의를 통해 독자적인 절차를 개발하고 정비결과를 시스템에 탑재하였다. 또한 BRM 정비사업을 성공적으로 수행하고 향후 안정적인 운영을 위해 BRM 부서관리자를 대상으로 정비 전· 후 2차례에 걸쳐 교육을 진행하였고, 정비사업 기간 동안 뉴스레터를배포하여 전 직원들의 적극적인 참여를 유도하고 기록관리 인식을 높이기 위해 노력하였다. 뿐만 아니라 정비성과를 유지하고 향후 BRM의체계적인 관리를 위해 기록관리시스템에 단위과제 현황기능 개발, 업무관리시스템에 과제관리 기능 보완, BRM 목적별 분류체계를 서울시정보소통광장의 서비스에 적용하는 등 여러 개선사항을 마련하였다. 본 연구가 새롭게 BRM을 도입하려는 기관 또는 분류체계 정비사업을준비하는 기관에서 시행착오를 줄이는데 참고사례가 될 수 있기를 바란다.

Automatic Linkage Model of Classification Systems Based on a Pretraining Language Model for Interconnecting Science and Technology with Job Information

정현지,김태현,장광선,신동구 한국과학기술정보연구원 2022 Journal of Information Science Theory and Practice Vol.10 No.-

For national industrial development in the Fourth Industrial Revolution, it is necessary to provide researchers with appropriatejob information. This can be achieved by interconnecting the National Science and Technology Standard Classification Systemused for management of research activity with the Korean Employment Classification of Occupations used for job informationmanagement. In the present study, an automatic linkage model of classification systems is introduced based on a pre-trainedlanguage model for interconnecting science and technology information with job information. We propose for the first time anautomatic model for linkage of classification systems. Our model effectively maps similar classes between the National Science &Technology Standard Classification System and Korean Employment Classification of Occupations. Moreover, the model increasesinterconnection performance by considering hierarchical features of classification systems. Experimental results show thatprecision and recall of the proposed model are about 0.82 and 0.84, respectively.

웹기반 전문가시스템의 구조 분류

임규건(Gyoo Gun Lim) 한국지능정보시스템학회 2007 지능정보연구 Vol.13 No.4

According to the expansion of the Internet use and the utilization of e-business, there are an increasing number of studies of intelligent?based systems for the preparation of ubiquitous environment. In addition, expert systems have been developed from Stand Alone types to web?based Client-Server types, which are now used in various Internet environments. In this paper, we investigated the environment of development for web?based expert systems, we classified and analyzed them according to type, and suggested general typical models of web?based expert systems and their architectures. We classified the web?based expert systems with two perspectives. First, we classified them into the Server Oriented model and Client Oriented model based on the Load Balancing aspect between client and server. Second, based on the degree of knowledge and inference?sharing, we classified them into the No Sharing model, Server Sharing model, Client Sharing model and Client-Server Sharing model. By combining them we derived eight types of web-based expert systems. We also analyzed the location problems of Knowledge Bases, Fact Bases, and Inference Engines on the Internet, and analyzed the pros & cons, the technologies, the considerations, and the service types for each model. With the framework proposed from this study, we can develop more efficient expert systems in future environments.

네트워크 중심성 척도가 추천 성능에 미치는 영향에 대한 연구

이동원(Dongwon Lee) 한국지능정보시스템학회 2021 지능정보연구 Vol.27 No.1

Collaborative filtering, which is often used in personalization recommendations, is recognized as a very useful technique to find similar customers and recommend products to them based on their purchase history. However, the traditional collaborative filtering technique has raised the question of having difficulty calculating the similarity for new customers or products due to the method of calculating similaritiesbased on direct connections and common features among customers. For this reason, a hybrid technique was designed to use content-based filtering techniques together. On the one hand, efforts have been made to solve these problems by applying the structural characteristics of social networks. This applies a method of indirectly calculating similarities through their similar customers placed between them. This means creating a customer’s network based on purchasing data and calculating the similarity between the two based on the features of the network that indirectly connects the two customers within this network. Such similarity can be used as a measure to predict whether the target customer accepts recommendations. The centrality metrics of networks can be utilized for the calculation of these similarities. Different centrality metrics have important implications in that they may have different effects on recommended performance. In this study, furthermore, the effect of these centrality metrics on the performance of recommendation may vary depending on recommender algorithms. In addition, recommendation techniques using network analysis can be expected to contribute to increasing recommendation performance even if they apply not only to new customers or products but also to entire customers or products. By considering a customer’s purchase of an item as a link generated between the customer and the item on the network, the prediction of user acceptance of recommendation is solved as a prediction of whether a new link will be created between them. As the classification models fit the purpose of solving the binary problem of whether the link is engaged or not, decision tree, k-nearest neighbors (KNN), logistic regression, artificial neural network, and support vector machine (SVM) are selected in the research. The data for performance evaluation used order data collected from an online shopping mall over four years and two months. Among them, the previous three years and eight months constitute social networks composed of and the experiment was conducted by organizing the data collected into the social network. The next four months’ records were used to train and evaluate recommender models. Experiments with the centrality metrics applied to each model show that the recommendation acceptance rates of the centrality metrics are different for each algorithm at a meaningful level. In this work, we analyzed only four commonly used centrality metrics: degree centrality, betweenness centrality, closeness centrality, and eigenvector centrality. Eigenvector centrality records the lowest performance in all models except support vector machines. Closeness centrality and betweenness centrality show similar performance across all models. Degree centrality ranking moderate across overall models while betweenness centrality always ranking higher than degree centrality. Finally, closeness centrality is characterized by distinct differences in performance according to the model. It ranks first in logistic regression, artificial neural network, and decision tree withnumerically high performance. However, it only records very low rankings in support vector machine and K-neighborhood with low-performance levels. As the experiment results reveal, in a classification model, network centrality metrics over a subnetwork that connects the two nodes can effectively predict the connectivity between two nodes in a social network. Furthermore, each metric has a different performance depending on the classification model type. This result implie

BERT 기반의 모델을 이용한 무기체계 소프트웨어 정적시험 거짓경보 분류 모델 개발 방법 연구

남효주,이인섭,정남훈,정성윤,조규태,노성규 한국정보과학회 2024 정보과학회논문지 Vol.51 No.7

최근 무기체계에서 소프트웨어의 규모와 복잡도가 커짐에 따라 소프트웨어의 신뢰성 및 안정성 확보가 요구되고 있다. 이를 위해 개발자는 정적 및 동적 신뢰성 시험을 수행해야한다. 하지만 정적시험 과정에서 많은 거짓경보들이 발생하여 이를 분석하고 처리하는데 많은 시간과 자원을 할애하고 있다. 기존 연구에서는 이러한 문제를 해결하기 위해 SVM, LSTM 등의 모델을 활용하여 거짓 경보를 분류한다. 하지만 연구들에서 사용된 모델의 입력값은 코드 관련 정보이거나, Word2Vec기반 코드 임베딩이므로 결함 발생 부분과 연관된 코드 간의 관계를 표현하지 못한다는 한계점이 존재한다. BERT기반의 모델은 양방향 트랜스포머의 적용을 통해 문장 간 앞뒤 관계를 학습하므로 코드 간 관계를 분석하는데 용이하다. 따라서 이를 거짓 경보 분류 문제에 활용하면 위 한계점을 극복할 수 있다. 본 논문에서는 정적시험 결과를 효율적으로 분석하기 위해 BERT기반의 모델을 활용한 거짓경보 분류 모델 개발 방법을 제안한다. 개발 환경에서 데이터셋을 구축하는 방법을 설명하고, 실험을 통해 분류 모델의 성능이 우수함을 보인다. Recently, as the size and complexity of software in weapon systems have increased, securing the reliability and stability is required. To achieve this, developers perform static and dynamic reliability testing during development. However, a lot of false alarms occur in static testing progress that cause wasting resources such as time and cost for reconsider them. Recent studies have tried to solve this problem by using models such as SVM and LSTM. However, they have a critical limitation in that these models do not reflect correlation between defect code line and other lines since they use Word2Vec-based code embedding or only code information. The BERT-based model learns the front-to-back relationship between sentences through the application of a bidirectional transformer. Therefore, it can be used to classify false alarms by analyzing the relationship between code. In this paper, we proposed a method for developing a false alarm classification model using a BERT-based model to efficiently analyze static test results. We demonstrated the ability of the proposed method to generate a dataset in a development environment and showed the superiority of our model.

딥러닝 기반 과일 선별 시스템

정수호,이명훈,여현 한국지식정보기술학회 2018 한국지식정보기술학회 논문지 Vol.13 No.5

Deep learning technology among artificial intelligence technologies has shown good results in image recognition field. In this paper, we use a learning model that is based on a Tensorflow based model that utilizes this deep learning technique and that has been repaired by Inception-v3 model. Based on the characteristics of the fruit, we construct a fruit classification system that classifies into four categories : Healthy apple, Damaged apple, Diseased apple and Discolored apple. To do this, we designed a learning model in which the number of learning iterations was 500 times based on 1,280 apple image data of four kinds and conducted a model evaluation experiment based on the fruit image data taken by the user. Experiments were based on images taken in three directions for accurate model evaluation. Experimental results show that the accuracy of the learning model is more than 90%. However, since fruit showed different classification results according to direction, it suggested the necessity of classification algorithm according to image direction in the future. If such a deep learning based fruit classification system is applied to farmers, fruit quality classifiers due to farm labor shortage are essential, and it will be possible to construct a fruit quality screening system with high accuracy and low cost.

한국 전통건축 분류체계에 따른 건축문화재 BIM 연구 동향

최현상(Choi, Hyun Sang),김성우(Kim, Sung Woo) 한국디지털디자인협의회 2014 (사)한국디지털디자인협의회 conference Vol.2014 No.5

Today"s construction based on building information modeling(BIM) can be approach Architectural heritage management to be background with Classification of Korean wooden building system. In this study, we define the 3D Classification system of Architectural heritage. first, we reviewed researches related to classification system of architectural heritage (CS-AH) and BIM based architectural heritage (BIM-AH). As a result, we found that CS-AH is focused on building elevation and type, and BIM-AH is biased on the Library and Parametric Modeling. Second, we figured out a relationship between the CS-AH and BIM-AH. From this analysis, we found that BIM-AH is biased on Library and Parametric since the building elevation and type was focused on CS-AH. This review suggests a potential of the 3D CS-AH to expand the range of research for BIM-AH.

지식증류 방안을 활용한 무인 군사 이미지 분류 AI 모델의 데이터 부족 및 경량화 모델 한계 극복

정자훈,송윤호,강인욱,류준열 한국산학기술학회 2024 한국산학기술학회논문지 Vol.25 No.3

Developing AI models for military unmanned systems requires consideration of the unique operational environment. Constraints like limited battery power and the high risk of destruction at the frontline necessitate restrictions on using costly, high-performance chips. In this study, we explored methods to enhance image classification performance of AI models under two key challenges. Firstly, constraints such as power and cost limit the utilization of high-capacity, high-performance models in unmanned systems. Secondly, there's a shortage of sufficient training data to ensure the performance of military AI models. To address these issues, we propose knowledge distillation. We selected EfficientNetB4 as the Teacher model, known for its superior performance despite high computational complexity, and SqueezeNet, ShuffleNetV2, and MobileNetV3 small as Student models. Through knowledge distillation, the high-accuracy knowledge of the Teacher model effectively enhanced the Student models, improving classification performance even under constraints. Such results are expected to enhance military utility by addressing the performance limitations of lightweight models applied to on device AI model in scenarios with limited training data.

앙상블 기법을 활용한 논문 주제 분류 모델

이수민,박민수,유재수,최도진 한국콘텐츠학회 2024 한국콘텐츠학회논문지 Vol.24 No.6

국내에서는 과학, 의학, 공학 등 다양한 학문 분야에서 많은 연구자들이 논문을 작성하고 있으며, 이러한 논문들은 DBPIA, KISS, RISS와 같은 학술지 논문 검색 플랫폼을 통해 널리 참조 및 인용되고 있다. 하지만, 현재의 논문 분류 체계는 저자가 선택한 주관적인 키워드에 의존하는 방식으로 운영되어 일관성과 표준화가 부족한 문제를 가지고 있다. 이러한 문제를 해결하기 위해, 본 논문에서는 KISTI(한국과학기술정보연구원)에서 제공하는 논문 데이터셋을 활용하고, 2018년 개정된 국가과학기술표준분류체계를 기반으로 하여 논문을 대분류 코드로 자동 분류하는 새로운 모델을 제안한다. 이 모델은 NTIS의 Open API를 이용한 데이터 증강과 XLM-RoBERTa 언어 모델을 활용하여 국내외 다양한 데이터에 대한 언어 이해력을 강화하였다. 또한, GRU(Gated Recurrent Unit)와 앙상블 기법을 사용한 계층적 접근 방식을 통해 논문의 초기 분야를 예측한 후, 이를 바탕으로 해당 분야의 대분류를 추가로 예측하는 방식으로 설계되었다. 결론적으로, 제안하는 모델은 논문의 주제를 더 정확하게 분류하고, 연구자들이 필요한 정보를 더 효과적으로 찾고 활용할 수 있도록 지원한다. 제안 모델의 우수성을 입증하기 위해 다양한 성능 평가를 수행하였으며, 그 결과는 기존 방식에 비해 정확도에서 상당한 개선을 보여준다. In South Korea, researchers across fields such as science, medicine, and engineering produce numerous papers that are widely referenced through platforms like DBPIA, KISS, and RISS. However, the current classification system relies on subjective keywords chosen by authors, leading to consistency and standardization issues. To address this, a new model is proposed using a dataset from the Korea Institute of Science and Technology Information (KISTI) and the revised 2018 National Science and Technology Classification System to automatically categorize papers into major categories. This model enhances linguistic comprehension for domestic and international data using data augmentation from NTIS's Open API and the XLM-RoBERTa language model. It employs a hierarchical approach with GRU (Gated Recurrent Unit) and ensemble techniques to first predict the initial field of the paper, and then further predict its major category. The proposed model more accurately classifies paper subjects, aiding researchers in finding and utilizing information effectively. Performance evaluations demonstrate significant improvements over existing methods.