• Journal of the Korea Society of Computer and Information
• /
• v.14 no.3
• /
• pp.121-131
• /
• 2009

BcN provides a wide variety of high-quality multimedia services such as broadcasting and communication convergence services. But the quality degeneration is observed in BcN when we use broadcasting and communication convergence service via more than one network of different internet service providers. In this paper, a QoS management system which is able to measure and maintain objectively the quality-related information in overall networks is proposed. The proposed QoS management system is tested on the pilot networks of BcN consortiums by measuring the quality of voice and video experienced by the actual users of the commercial video phone services. The result of the experiment shows that it is possible to figure out service qualify between a user and a service provider by analyzing the information from agents. The per-service traffic information collected by probes is proved to be useful to pinpoint the party responsible for the loss of the service qualify in case of the services including different service providers. As the result of the experiment, it is shown that the proposed QoS management system would play a key role of resolving the quality dispute, which is one of the important issues of QoS-guaranteed BcN.

• Journal of Broadcast Engineering
• /
• v.27 no.1
• /
• pp.133-145
• /
• 2022

Compared to the continuously increasing dog population and industry size in Korea, systematic analysis of related data and research on breed classification methods are very insufficient. In this paper, an automatic breed classification method is proposed using deep learning technology for 14 major dog breeds domestically raised. To do this, dog images are collected for deep learning training and a dataset is built, and a breed classification algorithm is created by performing transfer learning based on VGG-16 and Resnet-34 as backbone networks. In order to check the transfer learning effect of the two models on dog images, we compared the use of pre-trained weights and the experiment of updating the weights. When fine tuning was performed based on VGG-16 backbone network, in the final model, the accuracy of Top 1 was about 89% and that of Top 3 was about 94%, respectively. The domestic dog breed classification method and data construction proposed in this paper have the potential to be used for various application purposes, such as classification of abandoned and lost dog breeds in animal protection centers or utilization in pet-feed industry.

• 재활복지
• /
• v.15 no.4
• /
• pp.27-47
• /
• 2011

The purpose of this research is to develop transition process and programmes for gently converting people with disabilities to local communities, through their empowerments. The methods used in the research were literature research, case studies on related programs, and expert interviews. The following shows the results. Firstly, the transition process for people with developmental disorders is exceedingly limited, and the process does not consider special attention needed for people with developmental disorders. Secondly, the postsecondary education programs for people with developmental disorders are being started as transition processes, however these programs need more connection with secondary schools. Thirdly, the primary factors which must be contemplated for University programs(for people with developmental disorders, after secondary education) are self-cognition and self-understanding. According to these results, we can suggest transition processes and programs within Universities for people with developmental disorders. (In Korea) Firstly, the postsecondary transition for people with developmental disorders requires a consideration of their characteristics, and mainstreaming education is also needed for their self-cognition and self-understanding. Considering this, the transition process within Universities can be an alternative. Secondly, tertiary educations within Universities for people with developmental disorders need to consider the following factors: ① The programs in Universities need to set up a support network. ② The door to program should be open to all students with developmental disorders within the Universities. ③ Organization of human resources must be systematized and connected, so that specialists within the Universities can participate in the postsecondary education programs for people with developmental disorders. ④ Inclusive learning activities must be under-taken, allowing students without disabilities to participate in the programs for students with developmental disorders.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.10
• /
• pp.67-76
• /
• 2023

Although deep learning models are making innovative achievements in the field of computer vision, the problem of vulnerability to adversarial examples continues to be raised. Adversarial examples are attack methods that inject fine noise into images to induce misclassification, which can pose a serious threat to the application of deep learning models in the real world. In this paper, we propose a model that detects adversarial examples using differences in predictive values between edge-learned classification models and underlying classification models. The simple process of extracting the edges of the objects and reflecting them in learning can increase the robustness of the classification model, and economical and efficient detection is possible by detecting adversarial examples through differences in predictions between models. In our experiments, the general model showed accuracy of {49.9%, 29.84%, 18.46%, 4.95%, 3.36%} for adversarial examples (eps={0.02, 0.05, 0.1, 0.2, 0.3}), whereas the Canny edge model showed accuracy of {82.58%, 65.96%, 46.71%, 24.94%, 13.41%} and other edge models showed a similar level of accuracy also, indicating that the edge model was more robust against adversarial examples. In addition, adversarial example detection using differences in predictions between models revealed detection rates of {85.47%, 84.64%, 91.44%, 95.47%, and 87.61%} for each epsilon-specific adversarial example. It is expected that this study will contribute to improving the reliability of deep learning models in related research and application industries such as medical, autonomous driving, security, and national defense.

• Journal of Korean Society of Disaster and Security
• /
• v.16 no.2
• /
• pp.85-98
• /
• 2023

Flood is one of the most frequent natural disasters worldwide. In Korea, the probability of urban flooding is greatly increasing due to complex factors such as global warming, an increase in impervious areas, and limitations in expanding water supply facilities in existing urban areas. However, large-scale civil engineering works to prevent urban inundation are socially and economically difficult to obtain national consent. Recently the importance of resilience, which is the ability to return to the original state after a disaster through rapid recovery while preparing for natural disasters to a level that the local community can afford socially and economically, is increasing. Accordingly, various studies on urban resilience have been conducted, but the resilience measurement method related to the lifeline that provides essential services of the city is insufficient. However, among lifelines, road networks are important facilities for the transportation of recovery resources and rapid recovery in the event of a natural disaster, so road networks are a major factor that must be considered when measuring the degree of recovery of a city in the field of natural disasters. Therefore, this study proposes a recovery evaluation method considering the characteristics of resilience and road networks in the urban flooding field and analyzes the effect of road networks on urban resilience.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.52 no.3
• /
• pp.89-102
• /
• 2024

This study aims to examine the characteristics of the user experience of New Town neighborhood parks and explore issues that diversify the experience of the parks. In order to quantitatively analyze a large amount of park visitors' experiences, text-based Naver blog reviews were collected and analyzed. Among the Phase 1 and 2 New Towns, the parks with the highest user experience postings were selected for each city as the target of analysis. Blog text data was collected from May 20, 2003, to May 31, 2022, and analysis was conducted targeting Ilsan Lake Park, Bundang Yuldong Park, Gwanggyo Lake Park, and Dongtan Lake Park. The findings revealed that all four parks were used for everyday relaxation and recreation. Second, the analysis underscores park's diverse user groups. Third, the programs for parks nearby were also related to park usage. Fourth, the words within the top 20 rankings represented distinctive park elements or content/programs specific to each park. Lastly, the results of the network analysis delineated four overarching types of park users and the networks of four park user types appeared differently depending on the park. This study provides two implications. First, in addition to the naturalistic characteristics, the differentiation of each park's unique facilities and programs greatly improves public awareness and enriches the individual park experience. Second, if analysis of the context surrounding the park based on spatial information is performed in addition to text analysis, the accuracy of interpretation of text data analysis results could be improved. The results of this study can be used in the planning and designing of parks and greenspaces in the Phase 3 New Towns currently in progress.

• Information Systems Review
• /
• v.19 no.4
• /
• pp.43-62
• /
• 2017

This study investigates the effect of bonding and bridging social capital on the technological innovation performance in the Pangyo Techno Valley. In particular, we consider the information technology (IT) usage in industrial cluster as an antecedent of social capital. IT instigates the intra and extra communication and information sharing between employees, thereby promoting the formation of a network of various members. Results show that the IT usage factor positively affects both bridging and bonding social capital, but an evident difference exists among the effects of social capital on the technological innovation performance. In case of Pangyo industrial cluster, bridging social capital exerts significant effect on the technological innovation performance, whereas bonding social capital shows insignificance. Bridging social capital is composed of the interactions of various networks. Bonding social capital is based on the strong tie from trust and internal cooperation. Results are related with the characteristics of Pangyo Techno Valley, where various IT ventures need active communication and information sharing with other organizations for technological innovation performance.

• Korean Journal of Ecology and Environment
• /
• v.47 no.4
• /
• pp.282-291
• /
• 2014

This study was conducted to make distribution model of the Red-crowned Crane and the White-naped Crane according to the anthropogenic and natural factors affecting distribution of crane in Cheorwon, Korea. It was investigated that the impact power and its range of the indirect anthropogenic influence to feeding flock density in cranes from the road, residential area, military facilities, civilian control zone (CCZ), greenhouse and natural influence such as roosting site and available feeding area. Available feeding area is the most important factor for the crane's feeding site selection. The feeding flock density of the cranes near the residential area was lower than that of area far from the area, and tended to increase within 2.5 km distance. The increasing tendencies of feeding flock density from military facilities and high traffic volume road were similar, but the density in military facilities increased within 800 m, and the density from high traffic volume road increased within 2 km. These results suggested that residential area, military facilities and the road with high traffic volume had significant effect on foraging densities to the certain range. As the distance from the road with low traffic volume and roosting site increased, feeding flock density tended to decrease. The density of Red-crowned crane and White-naped crane inside the CCZ were respectively higher than those of outside the CCZ, especially for the Red-crowned crane. As a result, density of Red crowned cranes inside the CCZ was 5.2 times higher than that of outside, while that of white-naped cranes was 2.2 times bigger. If the density of greenhouse is lower than $40km^{-2}$, crane's feeding flock density in the low greenhouses density area was higher than that of high greenhouses density area. However, there was no difference in the feeding flock density if the density of the green houses is higher than $40km^{-2}$. The model for the Red-crowned Crane was related with available feeding area, distance from residential area, civilian control zone and distance from high traffic road. The model of the White-naped Crane was related with available feeding area, distance from roosting site and distance from lake. Finally, the estimated feeding flock density of cranes significantly correlated with density model according to the natural and anthropogenic factors.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.1
• /
• pp.301-311
• /
• 2013

This study aims to analyze characteristics, health behaviors and health management level related to private health examination recipients in one general hospital. To achieve this, we analyzed 150,501 cases of private health examination data for 11 years from 2001 to 2011 for 20,696 participants in 2011 in a Dae-Jeon general hospital health examination center. The cluster analysis for classify private health examination group is used z-score standardization of K-means clustering method. The logistic regression analysis, decision tree and neural network analysis are used to periodic/non-periodic private health examination classification model. 1,000 people were selected as a customer management business group that has high probability to be non-periodic private health examination patients in new private health examination. According to results of this study, private health examination group was categorized by new, periodic and non-periodic group. New participants in private health examination were more 30~39 years old person than other age groups and more patients suspected of having renal disease. Periodic participants in private health examination were more male participants and more patients suspected of having hyperlipidemia. Non-periodic participants in private health examination were more smoking and sitting person and more patients suspected of having anemia and diabetes mellitus. As a result of decision tree, variables related to non-periodic participants in private health examination were sex, age, residence, exercise, anemia, hyperlipidemia, diabetes mellitus, obesity and liver disease. In particular, 71.4% of non-periodic participants were female, non-anemic, non-exercise, and suspicious obesity person. To operation of customized customer management business for private health examination will contribute to efficiency in health examination center.

• Korean Journal of Ecology and Environment
• /
• v.48 no.3
• /
• pp.176-187
• /
• 2015

In this study, we hypothesized that the size of wintering crane population would change due to the climate factors. We assumed that wintering population size would differ by climate values in January, which is the coldest period in year. Especially, White-naped cranes were able to choose wintering site between Cheorwon and other alternative place where snow coverage had low influence, differing from Red crowned cranes. For this reason, we predicted the population size of White-naped cranes would fluctuate according to the extent of snow coverage in Cheorwon. Therefore we used snow coverage data based on MODIS and climate data from KMA (Korea Meteorological Administration) that are generally used. We analyzed the crane's population size in Cheorwon in January from 2002 to 2014. The temperature in the Cheorwon increased from 2002 to wintering period in 2007~ 2008 and went down, showing the lowest temperature in 2011~ 2012. With this phenomenon, warmth index showed the similar pattern with temperature. Amount of newly accumulated snow (the amount of snow that fallen from 0:01 am to 11:29 pm in a day) was low after 2002, but rapidly increased in 2010~ 2011 and 2011~ 2012. The area of snow coverage rapidly declined from 2002 to 2005~ 2006 but suddenly expanded in wintering period in 2009~ 2010 and 2010~ 2011. Wintering population size of the White-naped cranes decreased as snow coverage area increased in January and the highest correlation was found between them, compared to the other climatic factors. However, the number of individuals of Red crowned cranes had little relationship with general climate factors including snow cover range. Therefore it seems that population size of the Red crowned crane varied by factors related with habitat selection such as secure roosting site and area of foraging place, not by climatic factors. In multiple regression analysis, wintering population of White-naped cranes showed significant relationship with logarithmic value of snow cover range and its period. Therefore, it suggests that the population size of the White-naped crane was affected by snow cover range n wintering period and this was because it was hard for them to find out rice grains which are their main food items, buried in snow cover. The population size variation in White-naped cranes was caused by some individuals which left Cheorwon for Izumi where snow cover had little influence on them. The wintering population in Izumi and Cheorwon had negative correlation, implying they were mutually related.