• The KIPS Transactions:PartD
• /
• v.15D no.2
• /
• pp.223-234
• /
• 2008

Recently the web service area provides the efficient integrated environment of the internal and external of corporation and enterprise that wants the introduction of it is increasing. Also the web service develops and the new business model appears, the domestic enterprise environment and e-business environment are changing caused by web service. The web service which provides the similar function increases, most the method which searches the suitable service in demand of the user is more considered seriously. When it needs to choose one among the similar web services, service consumer generally needs quality information of web service. The problem, however, is that the advertised QoS information of a web service is not always trustworthy. A service provider may publish inaccurate QoS information to attract more customers, or the published QoS information may be out of date. Allowing current customers to rate the QoS they receive from a web service, and making these ratings public, can provide new customers with valuable information on how to rank services. This paper suggests the agent-based quality broker architecture which helps to find a service providing the optimum quality that the consumer needs in a position of service consumer. It is able to solve problem which modify quality requirements of the consumer from providing the architecture it selects a web service to consumer dynamically. Namely, the consumer is able to search the service which provides the optimal quality criteria through UDDI browser which is connected in quality broker server. To quality criteria value decision of each service the user intervention is excluded the maximum. In the existing selection architecture, the objective evaluation was difficult in subjective class of service selecting of the consumer. But the proposal architecture is able to secure an objectivity with the quality criteria value decision where the agent monitors binding information in consumer location. Namely, it solves QoS information of service which provider does not provide with QoS information sharing which is caused by with feedback of consumer side agents.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.35 no.7
• /
• pp.926-934
• /
• 2006

The base for preparing oyster hydrolysate-added yogurt was consisted of whole milk (1,000 mL), skim milk (44.05 to 42.05 g), enzymatic oyster hydrolysates powder (OHP, 0 to 2.0 g) and pectin. The yogurt base was fermented with 7 kinds of starter cultures (3% based on yogurt volume), such as Lactobacillus acidophilus, lactobacillus bulgaricus, lactobacillus casei, Lactobacillus fermentum, Lactobacillus pentosus, Streptcoccus thermophilus and the mixed starters (L. bulgaricus and S. thermophilus) at optimal temperature. Processing condition and quality characteristics of the yogurt were evaluated by analyzing pH, titratable acidity, viscosity, viable cell count, functional properties and sensory evaluation. The results suggested that the optimal conditions for preparing the good quality yogurt revealed the mixed starters (L. bulgaricus and S. thermophilus) for starter culture, 1.0 g of 3 kDa hydrolysate for amount, and 5.5 hrs for fermentation time. The good quality yogurt showed 4.31 for pH, 1.07% for titratable acidity, 469 cps for viscosity and $4.9{\times}10^8\;CFU/mL$ for viable cell count. The hydrolysate-added yogurt was 2 times higher in ACE inhibitory and antioxidant activities than commercial yogurt, and kept good quality during storage of 15 days at $5^{\circ}C$.

• Journal of Korean Society of Forest Science
• /
• v.73 no.1
• /
• pp.14-20
• /
• 1986

The processing times of the works, chainsaw felling, axe trimming and hand skidding to the corridor, by one-man-work method per single pole timber were investigated in the thinning young Japanese larch stand at the Training Forests of the Forest Work Tranining Center in Kangwon-do. The works were performed by a skilled worker with the craftman qualification and 69 trees were cut. Time was checked at intervals of 25/100 minute by the multimoment method and the worker's efficiency was evaluated for every cycle. Total working time was 8.11 hours of which 90% was for thinning work and 10% for cleaning work. Of the total working hours, 82.7% was net working time, 12.3% was general working time and 4.9% was non-valuated time. Of the net working time, 5.9hours, for only thinning, 20.9% was spent on moving to the feeling tree, 27.1% was spent on felling, 40.5% was spent on trimming and 11.5% was spent on skidding to corridor. Net chainsaw operating time was 0.94 hour which included 0.2 hour for cleaning work. Of the net chainsaw operating time, 0.94 hour, 66% was operating time and 34% was idle running time. The basic and general working times by DBH classes with application of 130% worker's efficiency calculated from regression equations were shown in table 1. For better practical using of this table, the simplified proposal was given in table 2.

• Journal of agriculture & life science
• /
• v.46 no.5
• /
• pp.111-125
• /
• 2012

In order to evaluate suitability by processing for non-GM soybean cultivars such as Gaechuck#2, Jinyangkong and CJ#1 lacking lipoxygenase (LOX) and kunitz trypsin inhibitor (KTI) protein, physicochemical characteristics and antioxidant activity of Kanjang made from above soybean were compared to Kanjang made from a conventional cultivar (Taekwangkong). Proximate compositions of soybeans were similar for the 4 kinds cultivars. Total phenol and flavonoid contents were significantly higher in cultivars lacking LOX and KTI protein than the Taekwangkong. In Kanjang, contents of total and reducing sugar were higher in Taekwangkong Kanjang than Kanjang from made cultivars lacking LOX and KTI protein. Contents of total and amino type nitrogen were the highest in the Jinyangkong Kanjang. Mineral contents were higher in the Jinyangkong and CJ#1 Kanjangs, amino acid contents were higher in the Kanjang made from 3 cultivars lacking LOX and KTI protein than the Taekwangkong. Taste of the Jinyangkong Kanjang with higher sweety and savory was also found to be superior to that of others in overall acceptability evaluation. Total phenol and flavonoid contents in Kanjang were significantly higher in the Kanjang made from cultivars lacking LOX and KTI protein than the Taekwangkong. Radical scavenging activity of Kanjang was increased in the total phenol contents dependent on. Reducing power by ferric-reducing antioxidant potential (FRAP) was significantly higher the Kanjang made from Gaechuck#2 and CJ#1 than the Taekwangkong. $Fe^{2+}$ chelating activity was higher in Taekwangkong Kanjang than the other cultivars, but its activity was similar to Jinyangkong Kanjang. Therefore, higher nutritional composition, total phenol and flavonoids contents and antioxidant activity in the Kanjang made from soybean cultivars lacking LOX and KTI protein might be provide better benefit for manufacture of another their products.

• FLOWER RESEARCH JOURNAL
• /
• v.18 no.2
• /
• pp.125-135
• /
• 2010

To discover the main characteristics of Korean traditional flower arrangement, they were compared with different articles and old paintings used in royal court ceremonies. The primary research involved principle of design. The times periods used were the Joseon Dynasty era of Korea, the Ming era of China, and the Edo eras of Japan. The result, which shows both the similarities and differences, of the research is summarized as follows. The similarities were that they all respect the features of nature, and their image expresses their creator's thinking. There was one technique, called 'Suje', in which a part of the stem is coming out from one branch. Also, each three eras preferred flowering trees and ornamental trees more than annuals or foliage plants. one of the differences was that korea used a simple number of materials. The work had volume and appeared mild by using a soft curved line which was repetitive and massive. The Joseon Dynasty era advanced a sense of beauty with artistic symmetry and balance. The work seemed soft and natural because of the little change in blank space, with almost no angle of line. The form had a characteristic preference of being taller than the typical Japanese arrangement. It appeared simple, calm, and rustic by using only one kind of material. In contrast, the Chinese style was gorgeous and displayed volume in a non-symmetrical tripodal form, which incorporated various colors and materials. Also, they avoided processing the materials in order to emphasize the original beauty of nature. Chinese flower arts did not become formalized because they did not consider the formality seriously the formal. The Japanese style was also gorgeous because it incorporated various materials and angles. It included an extreme technique in which an artificial line divided the blank space delicately. The line was both strong and delicate in an established form. The restriction of the main branch gave a light feeling, as well as more strain as in a balance sense. The Japanese eras emphasized more the use of line and a sense of blank space.

• Journal of agriculture & life science
• /
• v.45 no.2
• /
• pp.21-28
• /
• 2011

Hibiscus cannabinus L is a plant in the Malvaceae family. Kenaf was seeded at June 1st in 2010 and harvested at November 18th and dried and evaluated worth as a bulking agent for livestock composting. Harvested and dried Kenaf was divided into the bast, core and leaf. All materials were grinded by hammer mill and the moisture absorption, moisture evaporation, pH, volume weight and C/N ratio were measured. Kenaf was higher water absorption and evaporation ability than those of sawdust and chaff. The pH values of Kenaf were pH $2.8{\pm}0.01$ - $4.34{\pm}0.02$, which is lower pH value than those of sawdust (pH $5.28{\pm}0.01$) and chaff (pH $6.3{\pm}0.02$). The C/N ratio of Kenaf showed 649 of core, 204 of bast and 70 of leaf, which were lower than sawdust (789.1) but higher than chaff (132). In volume weight test, the materials were divided by particle size of Kenaf, named as group A(${\geq}4cm$), B(${\leq}4cm$, ${\geq}0.25cm$) and C(${\leq}0.25cm$). The volume of weight of group A and B for core, bast and leaf showed similar, but group C showed higher than those of sawdust and chaff. Especially, the volume weight of group C for leaf was 5 times higher than those of sawdust and chaff. This study suggested the possibility of using Kenaf as a bulk agent for composting of livestock manure. This is considered that strengthen the competitiveness of farmers through reducing the cost, prevention of environmental pollution caused by livestock manure and environmentally friendly processing of livestock manure.

• Journal of agriculture & life science
• /
• v.43 no.1
• /
• pp.51-59
• /
• 2009

The chemical components and antimicrobial activities of garlic from different area were investigated and analyzed to provide basic data for functional food materialization and processing. Hunter's values of garlic from different area were L 53.41~57.15, a -3.49~-4.38 and b 11.47~17.55. The moisture, crude protein, crude fat, nitrogen free extract, crude fiber and ash were 65.24~71.96, 6.24~9.35, 0.21~0.49, 19.01~22.72, 0.58~0.95 and 1.01~2.01%, respectively. The major minerals of garlic from different area were Na(27.22~112.03), Mg(18.17~32.56), K(242.16~569.28), Ca(28.60~63.93), P(117.72~265.21 mg%) and major free sugars were sucrose, glucose and fructose. The major amino acids of garlic from different area were proline, arglmne, glutamic acid and aspartic acid and content of total amino acid was 2,709.33~4,561.04 mg%. The ascorbic acid content of garlic from different area was 2.966~8.673 mg%. Composition of fatty acids of garlic from different area were linoleic acid, oleic acid and palmitic acid, unsaturated fatty acid and saturated fatty acid contents were 72.18~74.35 and 25.65~27.82%, respectively. Antimicrobial activities of garlic extracts as different area increased depends on concentration and showed the high antimicrobial activities against Gram(+) and Gram(-).

• Korean Journal of Remote Sensing
• /
• v.14 no.3
• /
• pp.213-222
• /
• 1998

Electro-Optical Camera(EOC) is the main payload of the KOrea Multi-Purpose SATellite(KOMPSAT) with the mission of cartography to build up a digital map of Korean territory including a Digital Terrain Elevation Map(DTEM). This instalment which comprises EOC Sensor Assembly and EOC Electronics Assembly produces the panchromatic images of 6.6 m GSD with a swath wider than 17 km by push-broom scanning and spacecraft body pointing in a visible range of wavelength, 510~730 nm. The high resolution panchromatic image is to be collected for 2 minutes during 98 minutes of orbit cycle covering about 800 km along ground track, over the mission lifetime of 3 years with the functions of programmable gain/offset and on-board image data storage. The image of 8 bit digitization, which is collected by a full reflective type F8.3 triplet without obscuration, is to be transmitted to Ground Station at a rate less than 25 Mbps. EOC was elaborated to have the performance which meets or surpasses its requirements of design phase. The spectral response, the modulation transfer function, and the uniformity of all the 2592 pixel of CCD of EOC are illustrated as they were measured for the convenience of end-user. The spectral response was measured with respect to each gain setup of EOC and this is expected to give the capability of generating more accurate panchromatic image to the users of EOC data. The modulation transfer function of EOC was measured as greater than 16 % at Nyquist frequency over the entire field of view, which exceeds its requirement of larger than 10 %. The uniformity that shows the relative response of each pixel of CCD was measured at every pixel of the Focal Plane Array of EOC and is illustrated for the data processing.

• Korean Journal of Remote Sensing
• /
• v.21 no.6
• /
• pp.445-456
• /
• 2005

The land surface parameters should be determined with sufficient accuracy, because these play an important role in climate change near the ground. As the surface reflectance presents strong anisotropy, off-nadir viewing results a strong dependency of observations on the Sun - target - sensor geometry. They contribute to the random noise which is produced by surface angular effects. The principal objective of the study is to provide a database of accurate surface reflectance eliminated the angular effects from MODIS 250m reflective channel data over Korea. The MODIS (Moderate Resolution Imaging Spectroradiometer) sensor has provided visible and near infrared channel reflectance at 250m resolution on a daily basis. The successive analytic processing steps were firstly performed on a per-pixel basis to remove cloudy pixels. And for the geometric distortion, the correction process were performed by the nearest neighbor resampling using 2nd-order polynomial obtained from the geolocation information of MODIS Data set. In order to correct the surface anisotropy effects, this paper attempted the semiempirical kernel-driven Bi- directional Reflectance Distribution Function(BRDF) model. The algorithm yields an inversion of the kernel-driven model to the angular components, such as viewing zenith angle, solar zenith angle, viewing azimuth angle, solar azimuth angle from reflectance observed by satellite. First we consider sets of the model observations comprised with a 31-day period to perform the BRDF model. In the next step, Nadir view reflectance normalization is carried out through the modification of the angular components, separated by BRDF model for each spectral band and each pixel. Modeled reflectance values show a good agreement with measured reflectance values and their RMSE(Root Mean Square Error) was totally about 0.01(maximum=0.03). Finally, we provide a normalized surface reflectance database consisted of 36 images for 2001 over Korea.