• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 1999.03a
• /
• pp.435-444
• /
• 1999

현재까지 연구되고 있던 정보보호관련분야의 계량화방법을 좀 더 다른 방법으로 접근하여, 정보시스템 환경 하에서 보안 및 관리 운영 평가 지수에 계량화하여 1차 집단과 2차 집단간의 차이를 연구하였다. 정보화 관련항목에 대하여 빈도 분석을 적용함으로서 군별, 항목별 분류를 통한 항목 비례 가중치법을 산출하였다. 또한, 선지정 가중치법을 이용하여, 보호지수와 관리운용지수에 따른 상관관계를 조사하여 안전관리 지수를 계량화 하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07b
• /
• pp.760-762
• /
• 2005

Support vector machine(SVM)은 졸은 일반화 능력을 가진 학습시스템으로, 최근 다양한 패턴 인식 분야에서 적용되고 있다. SVM은 기본적으로 이진 분류기이므로 두 개 이상의 클래스를 분류하기 위해서는 다중 클래스 분류가 가능한 형태로의 설계 방법이 필요하다. 이를 위해 각 클래스별로 독립적인 SVM들을 만들어 결과를 병합하는 방식이 주로 사용되어 왔다. 그러나 이러한 방법은 클래스의 수는 않고 한 클래스 내의 데이터의 수가 많지 않은 경우에는 SVM의 일반화 성능을 저하시키고 노이즈에 민감해지는 문제점을 가지고 있다. 이를 해결하기 위해 본 논문에서는 각 클래스내의 데이터간의 유사도 측정을 위한 통계적 정보를 안정적으로 추출하기 위해 두 데이터의 쌍을 입력으로 받는 새로운 SVM 설계 방법을 제시한다. 제안한 방법을 실제 생체인식 데이터에 적용한 실험에서 기존의 방법보다 우수한 분류 성능을 보임을 확인할 수 있었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.10a
• /
• pp.609-612
• /
• 2018

기계학습 분야에서 분류 문제를 해결하기 위해 다양한 알고리즘들이 연구되고 있다. 하지만 기존에 연구된 분류 알고리즘 대부분은 각 클래스에 속한 데이터 수가 거의 같다는 가정하에 학습을 진행하기 때문에 각 클래스의 데이터 수가 불균형한 경우 분류 정확도가 다소 떨어지는 현상을 보인다. 이러한 문제를 해결하기 위해 본 논문에서는 Conditional Generative Adversarial Networks(CGAN)을 활용하여 데이터 수의 균형을 맞추는 오버샘플링 기법을 제안한다. CGAN은 데이터 수가 적은 클래스에 속한 데이터 특징을 학습하고 실제 데이터와 유사한 데이터를 생성한다. 이를 통해 클래스별 데이터의 수를 맞춰 분류 알고리즘의 분류 정확도를 높인다. 실제 수집된 데이터를 이용하여 CGAN을 활용한 오버샘플링 기법이 효과가 있음을 보이고 기존 오버샘플링 기법들과 비교하여 기존 기법들보다 우수함을 입증하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.678-680
• /
• 2023

본 논문은 정확하면서도 효율적인 한국어 문장 분류 기법에 대해서 논의한다. 최근 자연어처리 분야에서 사전 학습된 언어 모델(Pre-trained Language Models, PLM)은 미세조정(fine-tuning)을 통해 문장 분류 하위 작업(downstream task)에서 성공적인 결과를 보여주고 있다. 하지만, 이러한 미세조정은 하위 작업이 바뀔 때마다 사전 학습된 언어 모델의 전체 매개변수(model parameters)를 학습해야 한다는 단점을 갖고 있다. 본 논문에서는 이러한 문제를 해결할 수 있도록 도메인 적응기(domain adapter)를 활용한 한국어 문장 분류 프레임워크인 DAKS(Domain Adaptation-based Korean Sentence classification framework)를 제안한다. 해당 프레임워크는 학습되는 매개변수의 규모를 크게 줄임으로써 효율적인 성능을 보였다. 또한 문장 분류를 위한 특징(feature)으로써 한국어 사전학습 모델(KLUE-RoBERTa)의 다양한 은닉 계층 별 은닉 상태(hidden states)를 활용하였을 때 결과를 비교 분석하고 가장 적합한 은닉 계층을 제시한다.

• The Journal of the Korea Contents Association
• /
• v.18 no.8
• /
• pp.178-186
• /
• 2018

Subject classification of thesis units is essential to serve scholarly information deliverables. However, to date, there is a journal-based topic classification, and there are not many article-level subject classification services. In the case of academic papers among domestic works, subject classification can be a more important information because it can cover a larger area of service and can provide service by setting a range. However, the problem of classifying themes by field requires the hands of experts in various fields, and various methods of verification are needed to increase accuracy. In this paper, we try to classify topics using the unsupervised learning algorithm to find the correct answer in the unknown state and compare the results of the subject classification algorithms using the coherence and perplexity. The unsupervised learning algorithms are a well-known Hierarchical Dirichlet Process (HDP), Latent Dirichlet Allocation (LDA) and Latent Semantic Indexing (LSI) algorithm.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.30 no.1
• /
• pp.71-76
• /
• 2017

In BIM design, the construction and use of a library are very important. Since the existing contents can be re-used, the design can be executed more effectively and efficiently. Unlike the construction, the civil engineering, in particular, the railroad sees an inappropriate development and standardization of libraries. Thus, this study aims to develop and standardize the 3D object library in the railroad facility. We first gather and analyze the railroad facility breakdown structure and relevant drawings. We then match the items of analyzed standard drawings and the breakdown structure items. It was reviewed whether the library was required according to all items, and if required, it was reviewed what software was proper. Available software were found to be Civil 3D, Revit, etc. Based on this analysis, we will design the attribute items and specifications of the 3D railroad infrastructure library, as well as construct the library thereof.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.4
• /
• pp.53-63
• /
• 2023

The utilization of department information in bibliometric analysis using scientific and technological literature is highly advantageous. In this paper, the department information dataset was built through the screening, data refinement, and classification processing of authors' department type belonging to university institutions appearing in academic journals in the field of science and technology published in Korea, and the automatic classification model based on deep learning was developed using the department information dataset as learning data and verification data. In addition, we analyzed the co-authorship structure and network in the field of computer science using the department information dataset and affiliation information of authors from domestic academic journals. The research resulted in a 98.6% accuracy rate for the automatic classification model using Korean department information. Moreover, the co-authorship patterns of Korean researchers in the computer science and engineering field, along with the characteristics and centralities of the co-author network based on institution type, region, institution, and department type, were identified in detail and visually presented on a map.

• The Korean Journal of Archival Studies
• /
• no.45
• /
• pp.189-217
• /
• 2015

The National Archives's presidential records in 1960s are neither complete nor accurate. However these records are worth enough to researchers as a source of the main policy stream and informations of the time. In the past, the catalog of National Archives only contained simple record information without the analysis of content. Therefore this article designed the subject classification scheme based on analysis of presidential records in economic sector. The distribution aspect of the subject record type is the clue to understand the main project of presidential secretary in 1960s during the industrialization process. Also the information of the content of record such as the character, association, location invigorate the collection of record and its use to related organizations.

• Journal of the Korean Society for information Management
• /
• v.20 no.1
• /
• pp.165-198
• /
• 2003

This study analyzed authors' use of theory in 654 articles that appeared in two core library and information science journals during last three decades. In order to analyze degree of theory use of LIS, such as, publication productivity, growth and distribution of theory in subfields. name and origin of theory, usability of each theory, subfields and journals, and so on, content analysis of LIS theories was performed through conceptual and empirical study. For the purpose of this study, we suggested a couple of new analytical methods, so called, ‘Subfield Classification Scheme’ within LIS, and ‘5 Degrees of Theory Use’ model for the first time.

• The Korean Journal of Applied Statistics
• /
• v.28 no.1
• /
• pp.9-21
• /
• 2015

The support vector machine has been successfully applied to various classification areas due to its flexibility and a high level of classification accuracy. However, when analyzing imbalanced data with uneven class sizes, the classification accuracy of SVM may drop significantly in predicting minority class because the SVM classifiers are undesirably biased toward the majority class. The weighted $L_2$-norm SVM was developed for the analysis of imbalanced data; however, it cannot identify irrelevant input variables due to the characteristics of the ridge penalty. Therefore, we propose the weighted $L_1$-norm SVM, which uses lasso penalty to select important input variables and weights to differentiate the misclassification of data points between classes. We demonstrate the satisfactory performance of the proposed method through simulation studies and a real data analysis.