• Journal of KIISE:Software and Applications
• /
• v.34 no.6
• /
• pp.549-561
• /
• 2007

Vision-based scene understanding is to infer and interpret the context of a scene based on the evidences by analyzing the images. A probabilistic approach using Bayesian networks is actively researched, which is favorable for modeling and inferencing cause-and-effects. However, it is difficult to gather meaningful evidences sufficiently and design the model by human because the real situations are dynamic and uncertain. In this paper, we propose a learning method of Bayesian network that reduces the computational complexity and enhances the accuracy by searching an efficient BN structure in spite of insufficient evidences and training data. This method represents the domain knowledge as ontology and builds an efficient hierarchical BN structure under constraint rules that come from the ontology. To evaluate the proposed method, we have collected 90 images in nine types of circumstances. The result of experiments indicates that the proposed method shows good performance in the uncertain environment in spite of few evidences and it takes less time to learn.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2016.05a
• /
• pp.191-193
• /
• 2016

인적오류 예방은 해양사고 예방에 가장 중요한 이슈로 현재 인식되고 있다. 현재 이러한 인적오류를 예방하기 위한 다양한 과학적인 기법들이 등장하고 있으나, 실제 인적오류를 예방할 수 있는 기법은 아직 개발되어 있지 못한 실정이다. 그 이유는 인적오류의 발생 원인과 특징이 사람을 대상으로 하기 때문에 실로 방대하고 원인식별이 어려우며, 원인과 결과 사이의 인과관계 구축에는 한계가 있기 때문이다. 기존 개발된 다양한 기법들은 이론적으로는 완벽할 수 있으나, 실제 방대한 원인과 결과 사이에 형성된 연계체인을 모두 흡수하기가 곤란하기 때문이다. 현재 IMO의 공식안전성평가(FSA) 기법이 해상분야에 널리 적용되고 있으나 구체적으로 어떠한 기법을 적용하여 인적오류를 적용할 수 있는지에 대해서는 아직도 애매모호한 실정이다. FTA, ETA, FEMA, SWIFT 등 다양한 분석기법의 등장과 AI, Fuzzy, MMC, Kalman 등 기초과학분야의 기본적인 이론과 기술을 적용할 수 있으나 인간의 인적오류 식별과 분석 및 평가와 예측에는 한계가 있는 것이 현재의 실정이다. 한편 최근에는 기존에 많은 문제점을 내포하고 있는 것으로 고려되었던 베이지안 네트워크(Bayesian Network, BN)가 다시 분석과 예측 분야에 등장하고 있는데, BN의 장점을 수용하고 단점을 해결할 수 있는 방법들이 연구되고 있기 때문이다. BN의 장점은 전방추론과 후방추론을 적용하여 사고의 원인과 결과를 분석한 후, 이에 대한 해결 방안을 식별할 수 있기 때문이다. BN의 단점은 이진(binary) 구조의 데이터만을 수용할 수 있기 때문에 상관 변수들이 방대한 경우 계산시간이 방대해지고 이를 모두 수용할 수 있는 방법이 없기 때문이다. 따라서 BN 구조를 어떻게 설계하는냐가 최근의 이수로 등장하고 있다. 본 연구에서는 이러한 제 문제점을 고찰하고 인적오류 모델 개발에 최적인 방법 또는 기술을 모색하는데 있다.

• Journal of Digital Convergence
• /
• v.19 no.3
• /
• pp.167-176
• /
• 2021

With development of big data analysis technology, many studies to find social issues using texts mining techniques have been conducted. In order to derive social issues, previous studies performed in a way that collects a large amount of text data from news or SNS, and then analyzes issues based on text mining techniques such as topic modeling and terms network analysis. Social issues are the results of various social phenomena and factors. However, since previous studies focused on deriving social issues that are results of various causes, there are limitations to revealing the cause of the issues. In order to effectively respond to social issues, it is necessary not only to derive social issues, but also to be able to identify the causes of social issues. In this study, in order to overcome these limitations, we proposed a method of deriving the factors that cause social issues from texts related to social issues based on the theory of part of Korean linguistics. To do this, we collected news data related to social issues for three years from 2017 to 2019 and proposed a methodology to find causes based causal sentences based on text mining techniques.

• Korean Journal of Social Welfare
• /
• v.62 no.4
• /
• pp.223-248
• /
• 2010

This study is to explore the contents and interaction of residents voluntary network and propose the strategies to promote residents voluntary network. The grounded theory was utilized to attain our object. Total of seven social worker and 17 residents participated in the study. Data were collected through in-depth interviews and documents. The data were analyzed by using Strauss and Corbin's method. Results are the followings. In open coding 13 categories, 32 subcategories and 133 concepts were constructed. In axial coding causal conditions were qualitative ascent of needs, emergence of the right welfare consumer. Phenomenon was agitation of praxis ground and grope of exist. Contextual conditions were crisis resources, skepticism of welfare. Intervention conditions were maturation of welfare cognition and proliferation of the sense of community responsibility. Strategy were resocialization of voluntary organization and construction of field related service delivery system. Consequence were grass routing welfare strategic fitting service system. In selective coding we constructed the core category: The praxis revolution from bottom for break social welfare environment.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.12
• /
• pp.502-512
• /
• 2019

To examine the problem structure of balanced regional development policy and identify the causes of the gap between the capital and non-capital areas, this study examined the causal structure in terms of the concentration and distribution of functions in the metropolitan area and non-capital area and explored the policy leverage. As a research method, the causal map was drawn using System Dynamics techniques and policy leverage was derived through an exploration of the feedback structure. In particular, the causes of the problems in balanced regional development policies were approached by system accidents rather than by single-circuit accidents, and causality analysis was performed among the variables constituting balanced regional development policies based on system accidents. In particular, it singled out 31 variables, developed 13 feedback loops, and confirmed four major policy leverages, including relocation of the capital function, local decentralization policies, consistent development, and investment practices centered on the Non­Capital Area. Subsequently, a follow-up study by computer simulation is needed by modeling the structure of the system to identify the ripple effect of the policies of the Capital Area on the Non­Capital Area.

• Journal of Korea Port Economic Association
• /
• v.38 no.3
• /
• pp.87-104
• /
• 2022

Considering the complex structure and high volatility in the shipping market, it is important to investigate the connectedness amongst influencing factors. This study explores the dynamic relationship between supply-demand factors and shipping freight indices. We choose Capesize and Panamax in the bulk carrier market and use quarterly data of GDP, world fleet, BCI, and BPI from 1999 to 2021. Applying the wavelet analysis and wavelet Granger causality test, the simultaneous examination of co-movement and causality between two factors and the shipping market in both the time and frequency domains is achieved. We find that co-movement and causality vary across time and frequencies, thereby existing dynamic relationships between variables. Second, compared to multiple coherencies using demand and supply factors together, partial coherencies indicate noticeable causalities. It implies that analyzing demand and supply factors separately is essential. Finally, shipping freight indices show a high correlation with the demand factor in a good market and with the supply factor in a bad market. Generally, GDP positively leads shipping freights in the recovery phase while the world fleet negatively leads shipping freights in the downturn. The research is meaningful in that the rarely-applied wavelet analysis is adopted in the shipping market and that it gives a reasonable ground to explain the role of supply and/or demand factors in different phases of the market cycle.

• Proceedings of the Korea Society of Environmental Toocicology Conference
• /
• 1997.12a
• /
• pp.91-99
• /
• 1997

과거에는 환경오염에 의한 피해의 유무, 즉 인과관계를 규명하는 것이 일차적인 관심 대상이었다. 그러나 구체적이고 정량적인 정보를 요구하는 현대사회에서는 오염피해의 정도와 심각성을 평가하여 주민에게 알려 주어야 하며, 어느 정도의 오염수준을 우리사회에서 받아들일 수 있는가의 판단이 매우 중요한 문제로 떠오르게 되었다. 또한 복잡다기화 되어가고 있는 사회 현상속에서 이해관계와 불확실성으로 얽혀 있는 환경문제를 풀어나가기 위해서는 과학적이고 합리적인 방법론이 요구되고 있다. 이를 위해 제시될 수 있는 방법론이 위해성 평가(decision-making) 수단이나 연구의 한 분야로 지난 30여년 동안 비교적 빠르게 발전되어져 왔다. 우리나라의 경우도 이미 위해성 평가에 대한 연구가 계속 발전중에 있으며, 특히 수계에서 검출 가능하고 잠재적인 위해성을 지니는 수질오염물질에 대한 전반적인 위해성을 평가하여 우리나라 수질관리정책에 유용한 기초자료들을 지시한 바 있다. 본문에서는 수질중 chloroform을 대상으로 확률분포를 이용한 위해성 평가 방법론과 대기중 benzene을 대상으로 노출 허용량(margine of exposure) 접근법을 소개하고자 한다.

• Review of Culture and Economy
• /
• v.20 no.1
• /
• pp.51-80
• /
• 2017

This study aims to analyze various issues and problems in the implementation process of cultural impact assessment in 2016. As a result of the analysis, it is found that the evaluation index of the cultural impact assessment is difficult to understand, and the referring data and materials in the evaluation planning phase in the process of the cultural impact assessment are deficient, and the evaluation planning is based on surveys rather than various methods. And there are some problems with evaluation scales and weighting, and they are focused on the evaluation itself rather than the feedback of the evaluation results. Therefore, this study suggests that, in the evaluation planning phase, the different methods of comparing the evaluation results according to the project execution, the project progress, and the completion time of the project should be adopted, and the project manager and the evaluation group participate developing the evaluation indicators together, and methodology for identifying causal relationship such as system dynamics will be adopted. In the stage of evaluation, it is necessary to introduce various methodology such as diverse residents participation, abundant case study, application of quantitative methodology, and the utilization of expert evaluation teams. In the evaluation feedback stage, it is necessary to provide thorough consulting based on the evaluation results.

• Journal of Gifted/Talented Education
• /
• v.26 no.2
• /
• pp.257-280
• /
• 2016

This study is focused on how the science gifted students can improve their system thinking by applying 3-steps mind map activities about astronomy contents. Through the pre-study two times, 9 students in 6th grade were participated in this study. It was analyzed the degree of improvement of systems thinking through the science gifted students' mind maps, causal maps and their interviews. The results of the study showed that the experimental group's system thinking who learned 3-steps mind map activities was more complex and diverse than the controlled group's. Therefore 3-steps mind map activities that have been developed in this study will be valuable as an alternative to improve their system thinking. Through this study, 3-steps mind map activities were confirmed that by inducing background knowledge activation and thinking systematization can be an alternative to improve the system thinking of science gifted students. So the 3-steps mind map activities can be helpful in cultivate human resources with creative problem solving ability based on their system thinking.

• The Journal of the Korea Contents Association
• /
• v.17 no.9
• /
• pp.406-417
• /
• 2017

The purpose of this study was to analyze an experience of sexual traffic victims and contribute for its formation of theory based on the results. A method of study was performed by using grounded theory method of Strauss and Corbin(1988). Through the results of study, one hundred and five concepts and thirty five subcategories and seventeen categories were deducted. 'destroy of mind and body' from the central phenomenon. The core category of experience of escaping from prostitution was presented by 'return of ordinary life recovered from destroy of life 'and the process of escaping from prostitution was presented in order of 'destroy of mind and body step', 'abandonment step', 'recovery of will step', 'challenge of adaption of ordianry life step'. In the study, it followed that an independent their own will of a sexually exploited women play an important role for a successful escaping from prostitution.