• 한국정보통신학회논문지
• /
• 제15권7호
• /
• pp.1531-1536
• /
• 2011

온라인 게임의 활성화로 온라인 게임의 폐해도 증가하고 있는데 온라인 게임의 대표적인 폐해 중 하나인 언어 폭력 문제가 심각한 사회문제를 야기하고 있다. 본 논문은 온라인 게임의 채팅에 나타나는 비속어를 자동으로 차단하는 시스템을 제안한다. 우리는 온라인 게임의 채팅창에 나타나는 문장을 수집하였고 비속어 포함 문장과 정상 문장으로 수동으로 분류하였다. 음절 n-gram과 어휘-품사 쌍을 자질로 사용하며 카이제곱 통계량을 이용하여 자질을 선택한다. 선택된 자질들을 이진가중치로 표현하여 지지벡터기계(SVM)를 학습한 후, SVM 분류기로 각 문장의 차단 여부를 결정하였다. 실험 결과, 수집된 데이터에 대해 약 90.4%의 F1 정확률을 얻었다.

• 한국정보통신학회논문지
• /
• 제15권6호
• /
• pp.1213-1220
• /
• 2011

본 논문에서는 분류자들 속에 중요도 정보를 삽입하여 동적 중요도 결정이 가능한 앙상블 시스템을 제안하였다. 그동안 앙상블 시스템에서 중요도는 훈련이 끝나고 결정된 중요도를 사용하였다. 한 번 결정된 중요도는 테스트 데이터에 상관없이 정적으로 사용되었다. 이 문제를 푸는 방법으로 관문 네트워크에서 구조적으로 계층을 두는 프로세스를 추가하여 동적 중요도 결정이 가능하게 하는 방법이 있지만 프로세스가 추가된다는 단점이 있다. 본 논문에서는 이런 추가적인 프로세스 없이 간단하게 동적 중요도 결정이 가능한 방법을 보여주고 구조적 변경 없이 기존의 시스템에 쉽게 적용할 수 있으며 AdaBoost보다 나은 성능을 보여주는 알고리즘을 제안한다.

• 한국정보통신학회논문지
• /
• 제7권4호
• /
• pp.767-778
• /
• 2003

이 논문은 유전자 알고리즘을 이용한 최적의 퍼지 규칙을 만드는 방법을 제시한다. 퍼지 규칙은 첫 번째 단계에서 학습 데이터를 이용해 생성된다. 이 단계에서 퍼지 c-Means 군집화 알고리즘과 군집 유효성을 사용해 구조를 결정하고 퍼지 규칙 수가 되는 군집 수를 결정한다. 첫 번째 단계에서 구조가 결정되면 퍼지규칙의 매개변수들은 유전자 알고리즘을 이용해서 조율된다. 또한, 비대칭 가우시안 소속 함수를 위해 분산 매개변수는 좌ㆍ우값을 따로 관리하여 조율한다. 이 방법은 가중치와 분산 공간에서 유전자 알고리즘을 사용함으로서 전역 최소 쪽으로 수렴하도록 한다.

• 한국의류산업학회지
• /
• 제15권4호
• /
• pp.554-564
• /
• 2013

We provide an empirical assessment that examines the differences in appearance management behavior, life satisfaction and demographic variables between groups classified by individual lifestyle. Questionnaires were administered to 513 female and male adults over 17 years of age in the Daegu and Kyungbok metropolitan regions. Descriptive statistics, cluster analysis, Cronbach's ${\alpha}$, ANOVA, Duncan test and ${\chi}^2$ test were applied to analyze data from 513 respondents. The results are as follows. First, we did a cluster analysis on the appearance management behavior of weight training, skin care, hair care, make-up and clothing selection. Four groups (passive, rational, fashion oriented, and active typed) where classified according to individual lifestyle. Second, the rational and active groups were more interested in the social life, environmental stability, health, fashion and economic seeking life. They were also more involved in appearance management behavior and in a higher level of life satisfaction. However, the differences of life satisfaction among the lifestyle group (male) were not statistically significant. Third, females with higher level of income and education (among the demographic variables) belonged to the active group. We found significant differences in appearance management behavior, life satisfaction and demographic variables among male and female groups classified by lifestyle.

• Journal of Mechanical Science and Technology
• /
• 제18권6호
• /
• pp.904-914
• /
• 2004

The fiber reinforced composite material is widely used in the multi-industrial field because of their high specific modulus and specific strength. It has two main merits which are to cut down energy by reducing weight and to prevent explosive damage proceeding to the sudden bursting which is generated by the pressure leakage condition. Therefore, Pressure vessels using this composite material can be applied in the field such as defence industry and aerospace industry. In this paper, for nonlinear finite element analysis of E-glass/epoxy filament winding of composite vessel subjected to internal pressure, the standard interpretation model is developed by using the ANSYS with AutoLISP and ANSYS APDL languages, general commercial software, which is verified as useful characteristic of the solution. Among the modules of the system, both the process planning module for carrying out the process planning of filament wound composite pressure vessel and the autofrettage process module for obtaining higher residual stress will minimize trial and error and reduce the period for developing new products. The system can serve as a valuable system for experts and as a dependable training aid for beginners.

• 한국의류학회지
• /
• 제38권4호
• /
• pp.528-539
• /
• 2014

This study analyzed the influence of self-construal on sociocultural attitude toward physical appearance, body satisfaction, and appearance management behavior through a structural equation model. The empirical study was based on the response of 369 adult females between the ages of 20 and 49 in Seoul. Self-construal was presented as an independent self-construal and interdependent self-construal, respectively. The sociocultural attitude toward physical appearance as an intermediate variable in the research model was composed of two sub-factors that included internalization and awareness. The other (body satisfaction) was measured by two factors (body and face). Appearance management behavior (as a final outcome variable) were composed of various factors that included clothing concern, skin care, hair care, and weight training. The findings of this study were: 1) the effect of independent self-construal on the sociocultural attitude toward physical appearance was not significantly meaningful; however, interdependent self-construal influenced it positively. 2) Sociocultural attitude toward physical appearance appeared to have a negative effect on body satisfaction. 3) The body satisfaction also had a negative effect on appearance management behavior in this study.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권1호
• /
• pp.364-380
• /
• 2016

The most popular approach in object classification is based on the bag of visual-words model, which has several fundamental problems that restricting the performance of this method, such as low time efficiency, the synonym and polysemy of visual words, and the lack of spatial information between visual words. In view of this, an object classification based on weakly supervised E2LSH and saliency map weighting is proposed. Firstly, E2LSH (Exact Euclidean Locality Sensitive Hashing) is employed to generate a group of weakly randomized visual dictionary by clustering SIFT features of the training dataset, and the selecting process of hash functions is effectively supervised inspired by the random forest ideas to reduce the randomcity of E2LSH. Secondly, graph-based visual saliency (GBVS) algorithm is applied to detect the saliency map of different images and weight the visual words according to the saliency prior. Finally, saliency map weighted visual language model is carried out to accomplish object classification. Experimental results datasets of Pascal 2007 and Caltech-256 indicate that the distinguishability of objects is effectively improved and our method is superior to the state-of-the-art object classification methods.

• 한국항공우주학회지
• /
• 제38권4호
• /
• pp.335-341
• /
• 2010

PD 제어기 기반으로 설계된 무인헬리콥터의 내부루프 제어기의 성능을 향상시키기 위 하여 한 개의 신경회로망이 적용되었다. 오차방정식의 응답특성 기반으로 설계된 PD 제어기는 운동모델의 비선형성에 의해 성능이 저하된다. 이러한 비선형성은 운동모델로부터 변형된 운동 역변환 모델(Modified Dynamic Inversion Model, MDIM)로 분리되었고 신경회로망의 출력에 의해 보상되었다. 신경회로망의 학습에는 제어기 안정성 보장을 위하여 리야프노프의 직접방법(Lyapunov's direct method)으로부터 유도된 온라인 가중치 적응법칙이 이용되었다. 신경회로망에 의한 PD제어기의 성능향상은 비선형성을 갖고 있는 무인헬리콥터의 수치시뮬레이션 결과로 보였다.

• East Asian mathematical journal
• /
• 제28권2호
• /
• pp.215-232
• /
• 2012

The methods of selection through observations and recommendations were introduced in the process of recruiting new students for the science education institutes for the gifted attached to 25 universities recently. This paper itemized the methods of screening through observations and recommendations. This paper also analyzed the problems with the methods and attempted to create plans for their improvement. The methods of selection through observations and recommendations led to the positive results that students' usual activities and attitudes in the classroom were reflected on the evaluation and that the cost of their private lessons was also reduced. However, the methods showed a few problems that need to be corrected. We point out problems occurring with examining their documents for submission and interviews. It was not easy to grade candidates' gifts, creativity, potential and development within the contents of the documents and the limited time of conducting interviews. On the plans for the developments of the implemented methods of selection through observations and recommendations, we have several suggestions. The chances for teachers' in-service training of learning the methods of selection through observations and recommendations need to be expanded. The interview needs to be enhanced and to have the same weight as the document screening. To secure the continuity of the education for the gifted, the clear guidelines from the Ministry of Education, Science, Technology along with the cooperation of the education institutes for the gifted are essential.

• 한국인터넷방송통신학회논문지
• /
• 제13권6호
• /
• pp.237-242
• /
• 2013

헬스장을 이용하고는 싶지만 과한 업무량과 지리적 위치, 시간 등 여러 가지 이유로 이용에 제한이 있다. 그래서 집에서 간단히 할 수 있는 유/무산소 운동을 즐기되, 혼자 하는 지루함을 잊게 하고 옆에서 누군가 코치(케어)해 줄 수 있는가에 대한 발상에 힘입어 헬스케어 어플리케이션 개발에 착수하게 되었다. 본 연구는 인터넷/매장에서 쉽게 접할 수 있는 MicroSoft Kinect 하드웨어를 사용하였고 키넥트를 이용 사용자를 인식하고, 3D 모델을 모델링하여 인식된 사용자와 맵핑시켜 동작을 컨트롤하고 몸무게, 칼로리를 케어하는 프로그램을 구현하였다.