• Convergence Security Journal
• /
• v.24 no.3
• /
• pp.165-171
• /
• 2024

Due to the IT convergence trend, many industrial domains are developing their own IoT services. With batteries and lightweight devices, IoT could expand into various fields including smart farms, smart environments, and smart energy. Many battery-powered IoT devices are passive in enforcing security techniques to maintain service time. This is because security technologies such as cryptographic operations consume a lot of power, so applying them reduces service maintenance time. This vulnerable IoT device security environment is not stable. In order to provide safe IoT services, security techniques considering battery power consumption are required. In this study, we propose an IoT device authentication technology that minimizes power consumption. The proposed technology is a device authentication function based on the Diffie-Hell man algorithm, and has the advantage that malicious attackers cannot masquerade the device even if salt is leaked during the transmission section. The battery power consumption of the authentication technology proposed in this study and the ID/PW-based authentication technology was compared. As a result, it was confirmed that the authentication technique proposed in this study consumes relatively little power. If the authentication technique proposed in this study is applied to IoT devices, it is expected that a safer IoT security environment can be secured.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.

• Journal of Intelligence and Information Systems
• /
• v.19 no.3
• /
• pp.1-23
• /
• 2013

To discover significant social issues such as unemployment, economy crisis, social welfare etc. that are urgent issues to be solved in a modern society, in the existing approach, researchers usually collect opinions from professional experts and scholars through either online or offline surveys. However, such a method does not seem to be effective from time to time. As usual, due to the problem of expense, a large number of survey replies are seldom gathered. In some cases, it is also hard to find out professional persons dealing with specific social issues. Thus, the sample set is often small and may have some bias. Furthermore, regarding a social issue, several experts may make totally different conclusions because each expert has his subjective point of view and different background. In this case, it is considerably hard to figure out what current social issues are and which social issues are really important. To surmount the shortcomings of the current approach, in this paper, we develop a prototype system that semi-automatically detects social issue keywords representing social issues and problems from about 1.3 million news articles issued by about 10 major domestic presses in Korea from June 2009 until July 2012. Our proposed system consists of (1) collecting and extracting texts from the collected news articles, (2) identifying only news articles related to social issues, (3) analyzing the lexical items of Korean sentences, (4) finding a set of topics regarding social keywords over time based on probabilistic topic modeling, (5) matching relevant paragraphs to a given topic, and (6) visualizing social keywords for easy understanding. In particular, we propose a novel matching algorithm relying on generative models. The goal of our proposed matching algorithm is to best match paragraphs to each topic. Technically, using a topic model such as Latent Dirichlet Allocation (LDA), we can obtain a set of topics, each of which has relevant terms and their probability values. In our problem, given a set of text documents (e.g., news articles), LDA shows a set of topic clusters, and then each topic cluster is labeled by human annotators, where each topic label stands for a social keyword. For example, suppose there is a topic (e.g., Topic1 = {(unemployment, 0.4), (layoff, 0.3), (business, 0.3)}) and then a human annotator labels "Unemployment Problem" on Topic1. In this example, it is non-trivial to understand what happened to the unemployment problem in our society. In other words, taking a look at only social keywords, we have no idea of the detailed events occurring in our society. To tackle this matter, we develop the matching algorithm that computes the probability value of a paragraph given a topic, relying on (i) topic terms and (ii) their probability values. For instance, given a set of text documents, we segment each text document to paragraphs. In the meantime, using LDA, we can extract a set of topics from the text documents. Based on our matching process, each paragraph is assigned to a topic, indicating that the paragraph best matches the topic. Finally, each topic has several best matched paragraphs. Furthermore, assuming there are a topic (e.g., Unemployment Problem) and the best matched paragraph (e.g., Up to 300 workers lost their jobs in XXX company at Seoul). In this case, we can grasp the detailed information of the social keyword such as "300 workers", "unemployment", "XXX company", and "Seoul". In addition, our system visualizes social keywords over time. Therefore, through our matching process and keyword visualization, most researchers will be able to detect social issues easily and quickly. Through this prototype system, we have detected various social issues appearing in our society and also showed effectiveness of our proposed methods according to our experimental results. Note that you can also use our proof-of-concept system in http://dslab.snu.ac.kr/demo.html.

• Journal of Applied Biological Chemistry
• /
• v.57 no.3
• /
• pp.267-277
• /
• 2014

Objective of this paper is to assess the available data on the pesticide uses and regulations in the golf courses, and provide the nationwide systematic management options. Numbers of golf courses in Korea are rapidly increasing from 2000s and reached at 421 sites by the end of 2011. Accordingly pesticide usage has been increased with years in direct proportion to the increasing number of golf courses. Amounts of pesticide applied in 2011 were 118,669 kg as of an active ingredient and were in the orders of fungicides (54.9%) > insecticides (24.4%) > herbicides (13.3%) > growth regulators (0.1%). Average pesticide usages in 2011 were 280.9 kg per golf course and $5.4kg\;ha^{-1}$. Frequencies of the residual pesticide detections in green and turf were higher than those in fairway and soil, respectively. Residue of highly toxic pesticides was not detected in golf courses. Ministry of Environment in 2010 has developed the 'golf course pesticide monitoring and management system' which is the advanced online registry for kind and amount of pesticides applied in each golf course. This system is intended for monitoring of the pesticide uses and residual levels and protecting the environmental pollution from pesticides in the golf course. In 2009, management of pesticides in the golf courses became the task of Ministry of Environment, being merged from many federal agency and ministries. The protocol for the site-specific best management practices, on which to base results from the risk assessment, should be set for pesticides in the golf to minimize the environmental impacts.

• The Korean Journal of Pesticide Science
• /
• v.5 no.3
• /
• pp.1-11
• /
• 2001

New technologies will have a large impact on the discovery of new herbicide site of action. Genomics, combinatorial chemistry, and bioinformatics help take advantage of serendipity through tile sequencing of huge numbers of genes or the synthesis of large numbers of chemical compounds. There are approximately $10^{30}\;to\;10^{50}$ possible molecules in molecular space of which only a fraction have been synthesized. Combining this potential with having access to 50,000 plant genes in the future elevates tile probability of discovering flew herbicidal site of actions. If 0.1, 1.0 or 10% of total genes in a typical plant are valid for herbicide target, a plant with 50,000 genes would provide about 50, 500, and 5,000 targets, respectively. However, only 11 herbicide targets have been identified and commercialized. The successful design of novel herbicides depends on careful consideration of a number of factors including target enzyme selections and validations, inhibitor designs, and the metabolic fates. Biochemical information can be used to identify enzymes which produce lethal phenotypes. The identification of a lethal target site is an important step to this approach. An examination of the characteristics of known targets provides of crucial insight as to the definition of a lethal target. Recently, antisense RNA suppression of an enzyme translation has been used to determine the genes required for toxicity and offers a strategy for identifying lethal target sites. After the identification of a lethal target, detailed knowledge such as the enzyme kinetics and the protein structure may be used to design potent inhibitors. Various types of inhibitors may be designed for a given enzyme. Strategies for the selection of new enzyme targets giving the desired physiological response upon partial inhibition include identification of chemical leads, lethal mutants and the use of antisense technology. Enzyme inhibitors having agrochemical utility can be categorized into six major groups: ground-state analogues, group specific reagents, affinity labels, suicide substrates, reaction intermediate analogues, and extraneous site inhibitors. In this review, examples of each category, and their advantages and disadvantages, will be discussed. The target identification and construction of a potent inhibitor, in itself, may not lead to develop an effective herbicide. The desired in vivo activity, uptake and translocation, and metabolism of the inhibitor should be studied in detail to assess the full potential of the target. Strategies for delivery of the compound to the target enzyme and avoidance of premature detoxification may include a proherbicidal approach, especially when inhibitors are highly charged or when selective detoxification or activation can be exploited. Utilization of differences in detoxification or activation between weeds and crops may lead to enhance selectivity. Without a full appreciation of each of these facets of herbicide design, the chances for success with the target or enzyme-driven approach are reduced.

• Journal of Hospice and Palliative Care
• /
• v.9 no.2
• /
• pp.93-100
• /
• 2006

Purpose: This study was designed to understand the doctor's attitude toward hospice and palliative care for terminal cancer patients. Methods: Specialists who work at general hospital were surveyed with questionnaires about hospice and palliative care for terminal cancer patients. The questionnaires comprise 17 items. The data were statistically analyzed. Results: Eighty one doctors responded. Their median age was 35 years old. Thirty six doctors (44.4%) were from internal medicine. The median of specialist's experience was 4 years. Forty three respondents (53.2%) have rarely examined and treated cancer patients even a week. Thirty seven respondents (45.6%) knew the exact definition of hospice and palliative care. Eighty respondents (98.8%) felt that hospice and palliative care is necessary, and 91.2% of them responded the necessity of palliative medicine specialist. As to the question 'Do you positively referred terminal cancer patient to hospice and palliative care?', 55 respondents (67.9%) responded 'Yes' and 22 (27.2%) 'No'. Among the 'Yes' respondents 17 (30.9%) had an experience of hesitation for referring patients to hospice and palliative care; the most common reason was the disagreement of family members (6, 35.3%). As for the reasons of responding 'No', 6 doctors (27.2%) did so because of their 'feeling of abandoning the patients' and the other f for the 'lack of information on the referral procedure for hospice and palliative care'. Thirty seven specialists (45.7%) thought it is most desirable for the patients to have hospice and palliative care for 3 months before death. Fifty eight specialists (71.6%) responded that hospice and palliative care help controlling the patient's psychological symptoms before all. Conclusion: While most doctors recognize the need of hospice and palliative care for patients with terminal ranter, their attitude toward hospice and palliative care was rather reserved. We suggest that continuing education, information and promotion for hospice and palliative care should be provided for doctors.

• Journal of Korean Home Economics Education Association
• /
• v.27 no.4
• /
• pp.121-140
• /
• 2015

The purposes of this study were to analyze the duplicated contents of 'Consumer life' area of Technology & Home Economics and other subject textbooks for the middle and high school students. It focused on textbooks compiled following the 2009 revised curriculum. To achieve the purposes of this study, "Technology & Home Economics I II", "Social studies I II", and "Ethics I II"textbooks for middle school and "Technology & Home Economics", "Social studies", and "Life & Ethics" textbooks for high school were analyzed based on the criteria for analyses of 'Consumer life' area. The results were as follows. First, the analysis of duplicated contents in Technology & Home Economics and other subjects (Ethics, Social studies) for middle school revealed that Technology & Home Economics textbook had the most proportion of 'Consumer Life' area, followed by Social studies and Ethics. The duplicated content elements in Technology & Home Economics, Ethics, and Social studies textbooks for middle school were 'consumer decision making', 'consumer information', 'economic impact of consumption', 'food life and sustainability', and 'consumption and sustainability'. Secondly, as a result of the content analysis of textbooks for high school Technology & Home Economics, Social studies, and Life & Ethics according to the criteria of analysis, it was found that Technology & Home Economics textbook had the most proportion of 'Consumer Life' area, followed by Life & Ethics and Social studies. The "content elements" 'food life management and consumption environment', 'desire of consumption', 'economic impact of consumption', 'changing factors and characteristics of consumer culture', and 'consumption and sustainability' were commonly found in all three textbooks. In this way, the 'Consumer life' area of Technology & Home Economics is thought to play a central role in teaching the 'Consumer Life' area because of its strength that contains detailed contents about consumer life for adolescent consumers who will apply it to everyday life. Based on the result of this research, it is needed to consider articulation of 'Consumer life' area of secondary schools for the future curriculum development of Technology & Home Economics to reduce the duplicated contents and to help the adolescents develop the ability to solve consumption problems they may encounter in real life and grow up to be rational adult consumers.

• Culinary science and hospitality research
• /
• v.22 no.3
• /
• pp.66-78
• /
• 2016

The purpose of this study was to provide hotel bakery organizations with useful information for human resources management based on a substantial clarification of the relationship and correlation of hotel bakery employees' self-leadership, self-efficacy, and organizational commitment. Resources were gathered from June 20, 2015 to July 10, 2015 by distributing a total of 500 surveys, from which 377 were collected. Excluding 23 survey forms not suitable for the analysis, 354 survyes were processed through factor analysis, reliability test, and multivariant structural analysis using SPSS 18.0 and AMOS 18.0 to verify the hypotheses. The findings of the analysis can be summarized as follows: first, behavior-centered strategies, natural compensation, and constructive thinking strategies had a significantly positive impact on self-efficacy. Second, in the analysis of impact of self-efficacy on organizational commitment, it was significant for emotional immersion, but did not have a significantly positive impact on normative immersion. Third, in the relationship between self-leadership and organizational commitment, behavior-centered strategies and natural compensation did not have a significant impact on emotional immersion. However, constructive thinking strategies had a significant impact. The following implications can be derived based on the above findings: this study implies the possibility of future studies on the variables of self-efficacy as it set behavior-centered strategies, natural compensation, and constructive thinking strategies as the preliminary factors under hotel bakery employees' self-leadership; and it analyzed the causality of each factor with emotional immersion and normative immersion as the subordinate factors of self-efficacy and organizational commitment to show that self-leadership and self-efficacy of hotel bakery employees based on emotional immersion and normative immersion can stably improve the organization of hotel bakeries.

• The Journal of Korean Society for Radiation Therapy
• /
• v.25 no.2
• /
• pp.167-173
• /
• 2013

Purpose: This study is to evaluate the efficacy of the CBCT and EXACTRAC the image on the spine stereotactic body radiation treatment. Materials and Methods: The study compared the accuracy of the dose distribution for changes in the real QA phantom for The shape of the body of the phantom was performed. Novalis treatment artificially set up at the center and to the right, on the Plan 1 mm, 2 mm, 3 mm in front 1 mm, 2 mm, 3 mm and upwards 1 mm, 2 mm, 3 mm and $0.5^{\circ}$ by moving side to side Exactrac error correction and error values of CBCT and plan changes on the dose distribution were recorded and analyzed. Results: Cubic Phantom of the experimental error, the error correction Exactrac X-ray 6D Translation in the direction of the 0.18 mm, Rotation direction was $0.07^{\circ}$. Translation in the direction of the 3D CBCT 0.15 mm Rotation direction was $0.04^{\circ}$. DVH dose distribution using the results of the AP evaluate the change in the direction of change was greatest when moving. Conclusion: ExacTrac image-guided radiation therapy with a common easy and fast to get pictures from all angles, from the advantage of CBCT showed a potential alternative. But every accurate information compared with CT treatment planning and treatment of patients with more accurate than the CBCT ExacTrac the location provided. Changes in the dose distribution in the experiment results show that the treatment of spinal SBRT set up some image correction due to errors at the target and enter the spinal cord dose showed that significant differences appear.

• Pediatric Gastroenterology, Hepatology & Nutrition
• /
• v.6 no.1
• /
• pp.24-31
• /
• 2003

Purpose: As diagnostic tools for Hirschsprung's disease (HD), barium enema and rectal biopsy have radiation exposure and invasiveness respectively; however anorectal manometry does not have these disadvantages. We therefore performed this study to evaluate the diagnostic efficacy of anorectal manometry. Methods: We reviewed medical records of infants with one or two symptoms of vomiting, abdominal distension, chronic diarrhea or constipation who had a anorectal manometry followed by barium enema and/or biopsy from July 1995 to May 2002. We evaluated the sensitivity, specificity and predictive value of anorectal manometry and barium enema for diagnosis of HD. We also measured sphincter length, median value of balloon volume at which rectoanal inhibitory reflex (RAIR) occurred. Results: All 61 patients received anorectal manometry, 33 of 61 received barium enema. 18 of 61 were diagnosed as HD according to histology and 43 of 61 were evaluated as a control. The sensitivity, specificity, positive predictive value, negative predictive value of anorectal manometry and barium enema for diagnosis of HD were 1.00, 0.91, 0.82, 1.00 and 0.93, 0.67, 0.70, 0.92 respectively. The mean value of sphincter length in control was $1.68{\pm}0.67$ cm and correlated with age, weight and significantly longitudinal length. The median value of balloon volume at which RAIR occurred was 10 mL and did not correlated with age, weight and longitudinal length. Conclusion: This study suggests that anorectal manometry is an excellent initial screening test for Hirschsprung's disease because of its safety and accuracy.