• 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 Global Scholars of Marketing Science
• /
• v.18 no.1
• /
• pp.115-139
• /
• 2008

Since financial crisis of IMF resulted in intensive competitiveness and adverse management environment, many hotel industries have responded it with restructuring. Since this restructuring is accompanied by reduction of employees, hence comes the recognition of justice in the procedure of restructuring. When the surviving employees in the restructuring process recognize unfairness in the procedure and practical operations, organization effectiveness can not be maintained due to losing trust of their employers. In this paper I will examine the relationship between validity of restructuring and compensatory programs for layoffs and surviving employees' trust of the employers. Also I will find out the relationship between remaining employees' trust of the employers and their job satisfaction and organization commitment. Through this relationship, we can prepare an alternative to reduce negative effect of restructuring. The hypotheses of this study are proposed as follows: H1: The higher surviving employees' recognition of procedural justice in restructuring process is, the higher their trust with a manager of the company is. H2: The higher surviving employees' recognition of distributive justice in restructuring process is, the higher their trust with a manager of the company is. H3: The higher surviving employees' recognition of procedural justice in restructuring process is, the higher their job satisfaction is. H4: The higher surviving employees' recognition of distributive justice in restructuring process is, the higher their job satisfaction is. H5: The higher surviving employees' recognition of procedural justice in restructuring process is, the higher their organization commitment is. H6: The higher surviving employees' recognition of distributive justice in restructuring process is, the higher their organization commitment is. H7: The higher surviving employees' trust with a manager of the company in restructuring process is, the higher their job satisfaction is. H8: The higher surviving employees' trust with a manager of the company in restructuring process is, the higher their organization commitment is. For the purposes of this study, employees working in luxury hotels located in Seoul were targeted. Self-administered questionnaires were distributed to those who consented with the investigation after explaining the purpose of the survey. A total of 500 questionnaires were distributed and 450 questionnaire were returned to the researcher for analysis. 430 of the returned questionnaires were used for analysis. As for the education for this survey, 250 junior college graduates or under (58.1%), 143 college graduates (33.3%) and 37 graduate school graduates (8.6%). As for the marital status, 315 persons (73.3%) are single and 115 are married (26.7%). As for the monthly income, 49 people (11.48%) are less than 2 million won, 148 (34.4%) are between 2 million and less than 2.5 million won, 153 (35.6%) are between 2.5 million to less than 3 million won, 80 (18.6%) are more than 3 million won. As for the workplace, 293 people (68.1%) work for the F&B department, 73 (17.0%) for rooms department, 41 (9.5%) for operation/ marketing department, 23 (5.3%) for account/ general affair department. As for the period of employment, 85 people (19.8%) are less than 5 years, 150 (34.9%) are between 6 to 9 years, 143 (33.3%) are between 10 to 14 years. and 52 (3.%) are more than 15 years. An exploratory factor analysis was used to survey validity and reliability of calculating tool on perceived values. This study used correlation between individual items and whole items and Cronbach's alpha value of multiple-item scale which is usually used to assess scale and reliability. Reliability of conceptual sub-dimension was assessed by basing on repeated procedure of correlation between individual items and whole items and factor loading. 1. Verification of correlation between validity of restructuring and trust This research showed that procedural and distributive justice of restructuring affects trust positively. The path coefficient between procedural justice of restructuring and trust is 0.719(t=10.135, p=0.000), and thereby the higher procedural justice results in higher trust. The path coefficient between distributive justice of restructuring and trust is 0.160(t=3.291, p=0.001), and thereby the higher distributive justice results in higher trust. Hence H1 and H2 are accepted. 2. Verification of correlation between validity of restructuring and job satisfaction The path coefficient between procedural justice of restructuring and job satisfaction is 0.179(t=2.202, p=0.028), and thereby the higher procedural justice results in higher job satisfaction. The path coefficient between distributive justice of restructuring and job satisfaction is 0.074(t=1.620, p=0.105), and thereby distributive justice of restructuring has no relationship with job satisfaction. Hence H3 is accepted, but H4 is removed. 3. Verification of correlation between validity of restructuring and organization commitment The path coefficient between procedural justice of restructuring and organization commitment is 0.188(t=2.466, p=0.014), and thereby the higher procedural justice results in higher organization commitment. The path coefficient between distributive justice of restructuring and organization commitment is 0.118(t=2.720, p=0.007), and thereby the higher distributive justice results in higher organization commitment. Hence H5 and H6 are accepted. 4. Verification of correlation between trust and job satisfaction The path coefficient between trust and job satisfaction is 0.610(t=6.736, p=0.000), and thereby the correlation has a meaningful result. Since the higher trust of the employer results in higher job satisfaction, H7 is accepted. 5. Verification of correlation between trust and organization commitment The path coefficient between procedural justice of restructuring and job satisfaction is 0.446(t=5.547 p=0.000), and thereby the higher trust of the employer results in higher organization commitment. Hence H8 is accepted. This research aimed to help the employers of hotel industries by analyzing the effects of validity of restructuring on employees' trust, job satisfaction and organization commitment. The research found that employer's validity of restructuring has significant affects on the degree of employee's trust with a manager, thereby reducing the negative effects of restructuring and enhancing organization commitment and job satisfaction. The principal purpose of this research is to confirm the correlation between employees' perceived validity of restructuring and their trust with a manager. Also whether this correlation results in competitive edge of the company is also investigated. It is also pointed out that employees had to participated the procedure of restructuring, sharing the philosophy and reason of restructuring. This participation and furthermore compensatory methods can reduce employees' anxiety of organization operations. Variable of trust appeared to have impact on intermediation effect between perceived variable of validity and job satisfaction, organization commitment, so that increase of trust with a manager plays an crucial role in increasing organization effectiveness. Since this research did not cover whole hotel industries which underwent restructuring, it showed a limit. Unlike previous studies which dealt with validity and trust of superior bosses, this research focussed on employers. Also the organization citizenship which is not considered in this study will be dealt with in the future study.

• Journal of KIISE:Computer Systems and Theory
• /
• v.34 no.8
• /
• pp.337-347
• /
• 2007

An embedded system is called a multi-mode embedded system if it performs multiple applications by dynamically reconfiguring the system functionality. Further, the embedded system is called a multi-mode multi-task embedded system if it additionally supports multiple tasks to be executed in a mode. In this Paper, we address a HW/SW partitioning problem, that is, HW/SW partitioning of multi-mode multi-task embedded applications with timing constraints of tasks. The objective of the optimization problem is to find a minimal total system cost of allocation/mapping of processing resources to functional modules in tasks together with a schedule that satisfies the timing constraints. The key success of solving the problem is closely related to the degree of the amount of utilization of the potential parallelism among the executions of modules. However, due to an inherently excessively large search space of the parallelism, and to make the task of schedulabilty analysis easy, the prior HW/SW partitioning methods have not been able to fully exploit the potential parallel execution of modules. To overcome the limitation, we propose a set of comprehensive HW/SW partitioning techniques which solve the three subproblems of the partitioning problem simultaneously: (1) allocation of processing resources, (2) mapping the processing resources to the modules in tasks, and (3) determining an execution schedule of modules. Specifically, based on a precise measurement on the parallel execution and schedulability of modules, we develop a stepwise refinement partitioning technique for single-mode multi-task applications. The proposed techniques is then extended to solve the HW/SW partitioning problem of multi-mode multi-task applications. From experiments with a set of real-life applications, it is shown that the proposed techniques are able to reduce the implementation cost by 19.0% and 17.0% for single- and multi-mode multi-task applications over that by the conventional method, respectively.

• Journal of KIISE:Databases
• /
• v.32 no.4
• /
• pp.345-361
• /
• 2005

A transform-space index indexes objects represented as points in the transform space An advantage of a transform-space index is that optimization of join algorithms using these indexes becomes relatively simple. However, the disadvantage is that these algorithms cannot be applied to original-space indexes such as the R-tree. As a way of overcoming this disadvantages, the authors earlier proposed the transform-space view join algorithm that joins two original- space indexes in the transform space through the notion of the transform-space view. A transform-space view is a virtual transform-space index that allows us to perform join in the transform space using original-space indexes. In a transform-space view join algorithm, the order of accessing disk pages -for which various space filling curves could be used -makes a significant impact on the performance of joins. In this paper, we Propose a new space filling curve called the adaptive row major order (ARM order). The ARM order adaptively controls the order of accessing pages and significantly reduces the one-pass buffer size (the minimum buffer size required for guaranteeing one disk access per page) and the number of disk accesses for a given buffer size. Through analysis and experiments, we verify the excellence of the ARM order when used with the transform-space view join. The transform-space view join with the ARM order always outperforms existing ones in terms of both measures used: the one-pass buffer size and the number of disk accesses for a given buffer size. Compared to other conventional space filling curves used with the transform-space view join, it reduces the one-pass buffer size by up to 21.3 times and the number of disk accesses by up to $74.6\%$. In addition, compared to existing spatial join algorithms that use R-trees in the original space, it reduces the one-pass buffer size by up to 15.7 times and the number of disk accesses by up to $65.3\%$.

• Korean Journal of Ichthyology
• /
• v.24 no.3
• /
• pp.201-219
• /
• 2012

The estuaries bank of Nakdong River, Korea, was constructed in 1987 and many arguments have been under discussion on the subject of ecological impact on ecosystem during 20 years. In this context, however, it was not focused on the role of fish-way and the way of improvement to promote fish movement to complete its life history. This paper was the result of investigation on fish fauna of upper and lower part of the bank and the analysis of comparison of the fishes with those using fish-way which were installed in each side of bank as ladder type and fish gate. The authors discussed the importance of fish gate and its management base on the result. The number of fishes collected in this study for upper part of estuaries bank was 31 species, among them 71% was the pure freshwater residents indicating the characteristics of the area as a freshwater ecosystem. It was observed that 9 species of coastal fishes were found in this region including Coilia ectens and Neosalanx sp. which is anadromous to spawn. Excluding 6 freshwater fishes, 30 fish species collected in lower part of estuaries bank were coastal fishes indicating that former brackish ecosystem was changed into marine one. The freshwater fishes found in this region were restricted the appearance only in the season of discharging freshwater into the sea. The number of fishes found in fish-way was 39 species, more than that of river and coast. But only 19 species founded in fish ladder was comparable with the result of investigation on fish gate including 32 species. It was considered that the fish gate has more important role as a fish passage from the fact that only it allowed to movement of numerous number of fry of Neosalanx sp. and Engraulis japonicus. These results indicates that two-way movement system is more efficient than downward only one in estuaries bank for fish migration.

• Journal of Internet Computing and Services
• /
• v.15 no.3
• /
• pp.79-90
• /
• 2014

User authentication based on ID and PW has been widely used. As the Internet has become a growing part of people' lives, input times of ID/PW have been increased for a variety of services. People have already learned enough to perform the authentication procedure and have entered ID/PW while ones are unconscious. This is referred to as the adaptive unconscious, a set of mental processes incoming information and producing judgements and behaviors without our conscious awareness and within a second. Most people have joined up for various websites with a small number of IDs/PWs, because they relied on their memory for managing IDs/PWs. Human memory decays with the passing of time and knowledges in human memory tend to interfere with each other. For that reason, there is the potential for people to enter an invalid ID/PW. Therefore, these characteristics above mentioned regarding of user authentication with ID/PW can lead to human vulnerabilities: people use a few PWs for various websites, manage IDs/PWs depending on their memory, and enter ID/PW unconsciously. Based on the vulnerability of human factors, a variety of information leakage attacks such as phishing and pharming attacks have been increasing exponentially. In the past, information leakage attacks exploited vulnerabilities of hardware, operating system, software and so on. However, most of current attacks tend to exploit the vulnerabilities of the human factors. These attacks based on the vulnerability of the human factor are called social-engineering attacks. Recently, malicious social-engineering technique such as phishing and pharming attacks is one of the biggest security problems. Phishing is an attack of attempting to obtain valuable information such as ID/PW and pharming is an attack intended to steal personal data by redirecting a website's traffic to a fraudulent copy of a legitimate website. Screens of fraudulent copies used for both phishing and pharming attacks are almost identical to those of legitimate websites, and even the pharming can include the deceptive URL address. Therefore, without the supports of prevention and detection techniques such as vaccines and reputation system, it is difficult for users to determine intuitively whether the site is the phishing and pharming sites or legitimate site. The previous researches in terms of phishing and pharming attacks have mainly studied on technical solutions. In this paper, we focus on human behaviour when users are confronted by phishing and pharming attacks without knowing them. We conducted an attack experiment in order to find out how many IDs/PWs are leaked from pharming and phishing attack. We firstly configured the experimental settings in the same condition of phishing and pharming attacks and build a phishing site for the experiment. We then recruited 64 voluntary participants and asked them to log in our experimental site. For each participant, we conducted a questionnaire survey with regard to the experiment. Through the attack experiment and survey, we observed whether their password are leaked out when logging in the experimental phishing site, and how many different passwords are leaked among the total number of passwords of each participant. Consequently, we found out that most participants unconsciously logged in the site and the ID/PW management dependent on human memory caused the leakage of multiple passwords. The user should actively utilize repudiation systems and the service provider with online site should support prevention techniques that the user can intuitively determined whether the site is phishing.

• Journal of Korean Ophthalmic Optics Society
• /
• v.11 no.3
• /
• pp.281-291
• /
• 2006

The aim of this study was performed the clinical test using Netspeg lens for good visual acuity on subjects with abnormal refraction status. The subjects of one hundreds adults (fifty males, fifty females, mean=21 years, range=19 to 24) were recorded. The subjects were researched the history including the systemic health, medication, genetics, allergy, systemic disease and ocular disease. The refraction test was recorded the monocular and binocular using objective method. Visual acuity was performed the binocular status using the Netspeg lens and CR-39. Stereopsis test was performed the titmus fly and TNO at near distance using Netspeg lens and CR-39. The P-VEP test was used the 16 pattern size(Bausch Lomb, production in USA) with three channels. Also Subjects viewed the p-vep stimulus with binocular vision through the corrected visual acuity using the Netspeg lens and CR-39. The contrast sensitivity test was performed the contrast sensitivity chart(pelli-Robertson, USA) at 1m distance using the Netspeg lens and CR-39. The ultrastructure of surface on the Netspeg lens and CR-39 was observed the SEM(JMS-5800, made in Japan). The results of this study was as follows: 1. In corrected visual acuity of abnormal refraction using the Netspeg lens and CR-39, the Netspeg lens wearer were acquired the good visual field and clear visual acuity comparative to CR-39 wearer in the subject vision test. however the comfort of visual acuity was similar results in the Netspeg lens and CR-39. Also the subjects of Netspeg lens wearer was good visual acuity more than CR-39 wearer and in the analysis of P-VEP, the amplitude of wave on Netspeg lens used appears to be better through the CR-39(p>0.5). Besides, on the contrast sensitivity, the Netspeg lens wearer was good results than CR-39. The value on stereopsis with TNO by Netspeg lens wearer was better than CR-39 in results. However, in the stereopsis test with Titmus, the Netspeg lens and CR-39 wearer was similar results. 2. The ultrastructure of Netspeg lens surface was the smooth and fine shape more than CR-39. Also, Netspeg lens have a fine line structure in ultrastructure. In conclusion, the results of this study conformed that the surface ultrastructure of Netspeg lens used is more specific pin hole design structure than CR-39. This study indicated that the vision of Netspeg lens used have a better than CR-39 in the corrected visual acuity for abnormal refraction eye. Therefore, In this paper, we suggested that the ultrastructure and line structure of Netspeg lens was related to good visual function. However the visual function of the aspheric Netspeg and ultra waterproof Netspeg lens was similar results.

• International Journal of Highway Engineering
• /
• v.13 no.3
• /
• pp.21-30
• /
• 2011

As well known, dowel bars are used to transfer traffic load acting on one edge to another edge of concrete slab in concrete pavement system. The dowel bars widely used in South Korea are round shape steel bar and they shows satisfactory performance under bending stress which is developed by repetitive traffic loading and environment loading. However, they are not invulnerable to erosion that may be caused by moisture from masonry joint or bottom of the pavement system. Especially, the erosion could rapidly progress with saline to prevent frost of snow in winter time. The problem under this circumstance is that the erosion not only drops strength of the steel dower bar but also comes with volume expansion of the steel dowel bar which can reduce load transferring efficiency of the steel dowel bar. To avoid this erosion problem in reasonable expenses, dowers bars with various materials are being developed. Fiber reinforced plastic(FRP) dower that is presented in this paper is suggested as an alternative of the steel dowel bar and it shows competitive resistance against erosion and tensile stress. The FRP dowel bar is developed in tube shape and is filled with high strength no shrinkage. Several slab thickness designs with the FRP dowel bars are performed by evaluating bearing stress between the dowel bar and concrete slab. To calculated the bearing stresses, theoretical formulation and finite element method(FEM) are utilized with material properties measured from laboratory tests. The results show that both FRP tube dowel bars with diameters of 32mm and 40mm satisfy bearing stress requirement for dowel bars. Also, with consideration that lean concrete is typical material to support concrete slab in South Korea, which means low load transfer efficiency and, therefore, low bearing stress, the FRP tube dowel bar can be used as a replacement of round shape steel bar.

• Horticultural Science & Technology
• /
• v.30 no.1
• /
• pp.27-33
• /
• 2012

This study aimed to investigate the responses of plant growth and blossom-end rot (BER) incidence to calcium (Ca) and its three kinds of antagonistic cations (K, Mg, and $NH_4$-N) with various ratios in nutrient solution for sweet paper (Capsicum annuum L. 'RZ208'). Both Ca to each cation and Ca to a series of cation combinations, such as potassium (K), ammonium nitrate ($NH_4$), or magnesium (Mg) were more influential to the fruit growth and quality than plant growth. Especially, the BER incidence was significantly influenced by the ratio treatments. For examples, when Ca:(K + Mg) or Ca:(K + Mg + $NH_4$) ratio was 1:2 the highest incident rate of BER about 70.3 or 86.3% was observed, lowering the marketable yield to 19 or 13.7% of the total yield, respectively. The correlation coefficiencies (= r) to relationships between the BER and K as well as BER and $NH_4$ were 0.82 (P < 0.05) and 0.65 (P < 0.05), respectively. Combination only with the Mg element was not correlated with the BER incidence. However, when both of the K and Mg concentrations were 0.65 (P < 0.05). The highest correlation coefficiency, 0.92 (P < 0.05), was found to a relationship between the BER and the tree elemental combination.

• Korean Journal of Environmental Agriculture
• /
• v.32 no.4
• /
• pp.355-358
• /
• 2013

BACKGROUND: Water-born pollution loads by agricultural non-point source (NPS) pollution are expected to become intensified due to ongoing precipitation change. Therefore, it is essential to develop a best management practice (BMP) that is suitable to agricultural environments in Korea. This study aimed to develop an environmental-friendly BMP to reduce NPS pollution load by agricultural activities. An eco-friendly way, small drainage pond, was suggested in this study to avoid direct drainage of agricultural runoffs and eventually reduce the amount of pollutants discharged into the surrounding aqua-environment. METHODS AND RESULTS: A small pond ($12m^2$) was constructed at the corner of a rice paddy field ($1,715m^2$) located in Suwon, Korea. Water was allowed to drain only via a small drainage pond. Sampling was repeatedly made at two locations, one from an entrance and the other from an exit of a pond, during the rice cultivation period (May to October, 2012). Generally, sampling was made only when runoff water drained through a pond, such as during and/or after rain (irrigation). The water quality analysis showed that all quality parameters (SS, $COD_{Mn}$, T-N, and T-P) were improved as water passed through the pond. The amount of runoff water was reduced by 96~100%. Suspended solids and COD concentrations was reduced by 79.3% and 45.6%, respectively. In case of T-N and T-P concentrations, the reduction rates were 52.2% and 60.5%, respectively and the amount of T-N and T-P were reduced by 16.3~73.0% and 15.4~70.1%, respectively. CONCLUSION(S): Our data implies that agricultural NPS pollution from rice paddy fields can be effectively managed when an appropriate drainage water management practice is imposed. In this paper, it was suggested that an installation of a small drainage pond can be effective to prevent not only the nutrient loss from rice fields but also pollutant discharge to surrounding water environments.