• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.04a
• /
• pp.209-212
• /
• 2001

기존의 클라이언트 컴포넌트 기술로부터 EJB(Enterprise JavaBeans)와 마이크로소프트의 DCOM 모델을 기반으로 하는 서버 환경의 컴포넌트 기술이 발표 되고있다. 이에 따라 비즈니스 분야에서 활용할 수 있는 비교적 규모가 큰 컴포넌트가 개발되고 있어 이들 컴포넌트를 새로운 소프트웨어 개발에 활용할 수 있도록 컴포넌트를 재정의(Customization)하고 합성하는 과정을 지원하는 도구의 개발이 필요하다. 본 연구에서는 컴포넌트 기반 소프트웨어 개발시 컴포넌트 저장소에 구축되어 있는 컴포넌트를 검색하여 사용자의 요구사항에 맞게 재정의 하고 기존 컴포넌트 패키지에서 컴포넌트를 조립 및 삭제할 때 컴포넌트간의 종속성을 유지할 수 있도록 하는 방법 및 도구를 개발하였다. 본 도구는 재정의 도구를 관리하는 재정의기(Customizer), 컴포넌트 저장소에 구축되어 있는 컴포넌트 패키지를 보여주고 수정, 삭제를 지원하는 컴포넌트 브라우저, 컴포넌트 브라우저로부터 선택한 컴포넌트의 속성을 나타내고 수정, 삭제 등을 지원하는 속성 편집기와 컴포넌트 브라우저로부터 가져온 컴포넌트를 시각적으로 편집할 수 있게 하는 디자이너(Designer)로 구성되며, 컴포넌트의 조립 및 삭제를 할 매 컴포넌트 인터페이스의 종속성을 확인할 수 있게 하는 종속성 브라우저(Dependency Browser), 종속성 유지를 위하여 대체 컴포넌트 및 인터페이스를 선택할 수 있게 하는 인터페이스 편집기(Interface Editor)를 제공한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.05a
• /
• pp.500-503
• /
• 2020

특허를 통해 기술의 권리를 정의하고 보호하는 일이 매우 중요해짐에 따라 특허 문서를 분석하는 연구 또한 중요해지고 있다. 특히 특허의 청구항을 종속항과 독립항을 구분하고, 관련된 인용을 찾아내는 일은 관련 특허들을 분석하는데 매우 중요하다. 본 연구는 최근 텍스트 분석 분야에 획기적 성능 개선을 이끈 BERT(Bidirectional Encoder Representations From Transformers) 언어 모델을 사용하고 Neural Network 의 파인 튜닝 과정을 통해 청구항의 독립과 종속을 구분하였고, 인용하는 항의 번호와 인용 문구로 이루어진 인용 패턴을 통해 종속항의 인용 항을 찾아내었다. 이 방법을 2003 년 이후의 xml 형식의 미국 특허 데이터에 사용한 결과, 정확도 99% 의 성능을 확보하였다.

• Proceedings of the Korea Society for Industrial Systems Conference
• /
• 1997.11a
• /
• pp.353-363
• /
• 1997

중첩 루프의 전체적인 병렬화를 하기 위해서 자료 종속을 효과적으로 제거하는 알고리즘이다. 즉 순차 루프를 중첩된 DOALL루프로의 자동 변환에 대한 절차이다.

• Journal of radiological science and technology
• /
• v.30 no.4
• /
• pp.419-426
• /
• 2007

In this paper, the input pattern vectors are extracted and the learning algorithms is designed to recognize taste(bitter, sweet, sour and salty) pattern vectors. The signal intensity of taste are used to compose the input pattern vectors. The SOM(Self Organizing Maps) algorithm for taste pattern recognition is used to learn initial reference vectors and the ot-star learning algorithm is used to determine the class of the output neurons of the sunclass layer. The weights of the proposed algorithm which is between the input layer and the subclass layer can be learned to determine initial reference vectors by using SOM algorithm and to learn reference vectors by using LVQ(Learning Vector Quantization) algorithm. The pattern vectors are classified into subclasses by neurons in the subclass layer, and the weights between subclass layer and output layer are learned to classify the classified subclass, which is enclosed a class. To classify the pattern vectors, the proposed algorithm is simulated with ones of the conventional LVQ, and it is confirmed that the proposed learning method is more successful classification than the conventional LVQ.

• Journal of Korea Multimedia Society
• /
• v.5 no.3
• /
• pp.331-341
• /
• 2002

Traditional slicing techniques make slices through dependence graph and improve the accuracy of slices. However, traditional slicing techniques require many vertices and edges in order to express a data communication link because they are based on static slicing techniques. Therefore the graph becomes very complicated. We propose the representation of a dynamic system dependence graph so as to process the slicing of a software system that is composed of related programs in order to process certain jobs. We also propose programs on efficient slicing algorithm using relations of relative tables in order to compute dynamic slices of a software system. Using a marking table from results of the proposed algorithm can make dynamic system dependence graph for dynamic slice generation. Tracing this graph can generate final slices. We have illustrated our example with C program environment. Consequently, the efficiency of the proposed dynamic system dependence graph technique is also compared with the dependence graph techniques discussed previously. As the results, this is certifying that the dynamic system dependence graph is more efficient in comparison with system dependence graph.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.4
• /
• pp.593-602
• /
• 2018

This study examined the impact of trivial baselines obtained from GNSS single-baseline processing into a network adjustment in terms of accuracy and estimated coordinates sets. To this end, depending on the inclusion of the dependent baselines, three different types of experimental GNSS networks were composed and adjusted. The results showed that the networks including the trivial baselines are generally overestimated, but differences in the derived coordinates are limited at the millimeter level. A comparison of the adjusted coordinates with those published by the national geodetic agency showed that results of the network consisting of only the independent baselines are more constant than those of the networks with trivial baselines. Finally, a trivial baseline should be excluded from the GNSS network adjustment with a consideration of the realistic accuracy presentation and data processing burden.

• The Journal of the Korea Contents Association
• /
• v.10 no.11
• /
• pp.30-37
• /
• 2010

It is very important to analyze online customer reviews, which are small documents of writing opinions or experiences about products or services, for both customers and companies because the customers can get good informations and the companies can establish good marketing strategies. In this paper, we did not propose only dependency network model which is tool for analyzing online customer reviews, but also designed and implemented the system based on the dependency network model. The dependency network model analyzes both subjective and objective sentences, so that it can represent relative importance and relationship between the nouns in the sentences. In the result of implementing, we recognized that relative importance and relationship between the features of products or services, which can not be mined by opinion mining, can be represented by the dependency network model.

• Journal of Korea Multimedia Society
• /
• v.13 no.3
• /
• pp.369-377
• /
• 2010

In vocabulary recognition systems based on HMM(Hidden Markov Model)s, training process unseen model bring on show a low recognition rate. If recognition vocabulary modify and make an addition then recreated modeling of executed database collected and training sequence on account of bring on additional expenses and take more time. This study suggest efficient context dependent process modeling method using decision tree-based state tying. On study suggest method is reduce recreated of model and it's offered that robustness and accuracy of context dependent acoustic modeling. Also reduce amount of model and offered training process unseen model as concerns context dependent a likely phoneme model has been used unseen model solve the matter. System performance as a result of represent vocabulary dependence recognition rate of 98.01%, vocabulary independence recognition rate of 97.38%.

• Journal of Internet Computing and Services
• /
• v.1 no.1
• /
• pp.105-113
• /
• 2000

The representation of a procedural program cannot be applied directly to object oriented program representation consisting of class, object, inheritance, and dynamic binding. Furthermore, preexisting program dependence represented the dependence among statements, but not among variables. That is, it could not solve the problem of which variables make an effect on given variables. Consequently, this study presents the method dependence model representing implementation level information including the dependence among variables in an object oriented program. I also propose implementation-based class testing technique based on the test adequacy criterion of an object-oriented program. Considering inter-data member dependences and a set of axioms for test data adequacy, it generates sequences of methods as test cases which satisfy a flow graph-based testing criterion. For a derived class testing, it considers inheritance relationship and the resuability of the testing information for its parent classes which verified the reduction of test cost through the experiment.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.5 s.43
• /
• pp.95-103
• /
• 2006

In this paper, the input pattern vectors are extracted and the learning algorithms is designed to recognize EEG pattern vectors. The frequency and amplitude of alpha rhythms and beta rhythms are used to compose the input pattern vectors. And the algorithm for EEG pattern recognition is used SOM to learn initial reference vectors and out-star learning algorithm to determine the class of the output neurons of the subclass layer. The weights of the proposed algorithm which is between the input layer and the subclass layer can be learned to determine initial reference vectors by using SOM algorithm and to learn reference vectors by using LVQ algorithm, and pattern vectors is classified into subclasses by neurons which is being in the subclass layer, and the weights between subclass layer and output layer is learned to classify the classified subclass, which is enclosed a class. To classify the pattern vectors of EEG, the proposed algorithm is simulated with ones of the conventional LVQ, and it was a confirmation that the proposed learning method is more successful classification than the conventional LVQ.