• Journal of Korean Society of Forest Science
• /
• v.54 no.1
• /
• pp.1-24
• /
• 1981

Kangweon-Do is rich in sightseeing resources. There are three sightseeing areas;first, mountain area including Seolak and Ohdae National Parks, and chiak Provincial Park; second eastern coastal area; third lake area including the watersheds of North Han River. In this paper, several methods of forestation were studied for landscaping the North Han River watersheds centering around Chounchon. In Chunchon lake complex, there are four lakes; Uiam, Chunchon, Soyang and Paro from down to upper stream. The total surface area of the above four lakes is $14.4km^2$ the total pondage of them 4,155 million $m^3$, the total generation of electric power of them 410 thousand Kw, and the total forest area bordering on them $1,208km^2$. The bordering forest consists of planned management forest ($745km^2$) and non-planned management forest ($463km^2$). The latter is divided into green belt zone, natural conservation area, and protection forest. The forest in green belt amounts to $177km^2$ and centers around the 10km radios from Chunchon. The forest in natural conservation area amounts to $165km^2$, which is established within 2km sight range from the Soyang-lake sides. Protection forest surrounding the lakes is $121km^2$ There are many scenic places, recreation gardens, cultural goods and ruins in this lake complex, which are the same good tourist resources as lakes and forest. The forest encirelng the lakes has the poor average growing stock of $15m^3/ha$, because 70% of the forest consists of the young plantation of 1 to 2 age class. The ration of the needle-leaved forest, the broad-leaved forest and the mixed forest in 35:37:28. From the standpoint of ownership, the forest consists of national forest (36%), provincial forest (14%), Gun forest (5%) and private forest(45%). The greater part of the forest soil, originated from granite and gneiss, is much liable to weathering. Because the surface soil is mostly sterile, the fertilization for improving the soil quality is strongly urged. Considering the above-mentioned, the forestation methods for improving landscape of the North Han River Watersheds are suggested as follows: 1) The mature-stage forest should be induced by means of fertilizing and tendering, as the forest in this area is the young plantation with poor soil. 2) The bare land should be afforested by planting the rapid growing species, such as rigida pine, alder, and etc. 3) The bare land in the canyon with moderate moist and comparatively rich soil should be planted with Korean-pine, larch, ro fir. 4) Japaness-pine stand should be changed into Korean-pine, fir, spruce or hemlock stand from ravine to top gradually, because the Japanese-pine has poor capacity of water conservation and great liability to pine gall midge. 5) Present hard-wood forest, consisting of miscellaneous trees comparatively less valuable from the point of wood quality and scenerity, should be change into oak, maple, fraxinus-rhynchophylla, birch or juglan stand which is comparatively more valuable. 6) In the mountain foot within the sight-range, stands should be established with such species as cherry, weeping willow, white poplar, machilus, maiden-hair tree, juniper, chestnut or apricot. 7) The regeneration of some broad-leaved forests should be induced to the middle forest type, leading to the harmonious arrangement of the two storied forest and the coppice. 8) For the preservation of scenery, the reproduction of the soft-wood forest should be done under the selection method or the shelter-wood system. 9) Mixed forest should be regenerated under the middle forest system with upper needle-leaved forest and lower broad-leaved forest. In brief, the nature's mysteriousness should be conserved by combining the womanly elegance of the lakes and the manly grandeur of the forest.

• The Korean Journal of Pesticide Science
• /
• v.15 no.4
• /
• pp.545-572
• /
• 2011

The genetic diagnosis methods by RT-PCR and Virion capture (VC)/RT-PCR against Rice stripe virus (RSV) were developed. Three diagnosis methods of seedling test, ELISA and RT-PCR were compared in virus detection sensitivity (VDS) for RSV. The VDS of ELISA for RSV viruliferous small brown plant hopper (SBPH) was higher with 40.5% than that of seedling test. The VDS of RT-PCR was higher with 21% than that of ELISA. The VDS of ELISA and VC/RT-PCR was same with 9.2% in average on the SBPH collected from fields at the areas of Gimpo, Pyungtaeg and Sihueng, Gyeonggi province in 2009. The specific primers of RSV for SBPH and rice plant were developed for the diagnosis by Real time PCR. The RQ value of Real time PCR for the viruliferous and non viruliferous SBPH was 1 for 50 heads of non viruliferous SBPH, 96.5 for 50 heads of viruliferous SBPH, 23.1 for 10 heads of viruliferous SBPH + 40 heads of non viruliferous SBPH, and 75.6 for 30 heads of viruliferous SBPH + 20 heads of non viruliferous SBPH. The RQ value was increased positively by the ratio of viruliferous SBPH. Full sequences of 4 genomes of RSV RNA1, RNA2, RNA3 and RNA4 were analysed for the 13 RSV isolates from rice plants collected from different areas. Genetic relationships among the RSV isolates of Korea, Japan and China were classified as China + Korea, and China + Korea + Japan by phylogenetic analysis for RSV RNA1 and RNA2. In case of RNA3 involved in pathogenicity, genetic relationship of RSV among the three countries was grouped into 3 as China, China + Korea, and Korea + Japan. According to the genetic relationships in RSV RNA4, RSV isolates were grouped into 4 as China, Korea, China + Korea + Japan, and Korea + Japan. Viruliferous insect rate (VIR) of RSV in average increased in each year from 2008 to 2010, and the rates were 4.3%, 6.1%, and 7.2%, respectively, at the 28 major rice production areas in 7 provinces including Gyeonggido. The highest VIR in each year was 11.3% of Gyeonggido in 2008, 20.1% of Jellanamdo in 2009 and 14.2% of Chungcheongbukdo in 2010. The highest VIR depending upon the investigated areas was 22.1% at Buan of Jellabukdo in 2008, 36% at Wando and Jindo of Jellanamdo in 2009, and 30.0% at Boeun of Chungcheongbukdo in 2010. Average population density (APD) of overwintered SBPH was 13.1 heads in 2008, 13.9 heads in 2009 and 5.6 heads in 2010. The highest APD was 39.1 and 60.4 heads at Buan of Jellabukdo in 2008 and 2009, respectively, and 14.0 heads at Pyungtaeg of Gyeonggido. The acreage of RSV occurred fields was 869 ha in the western and southern parts, mainly at Jindo and Wando areas, of Jellanamdo in 2008. In 2009, RSV occurred in the acreage of 21,541 ha covered whole country, especially, partial and whole plant death were occurred with infection rate of 55.2% at 3,025 plots in 53 Li, 39 Eup/Myun, 19 Si/Gun of Gyeonggido, Incheonsi, Chungcheongnamdo, Jeollabukdo and Jeollanamdo. Seasonal development of overwintered SBPH was investigated at Buan, Jeollabukdo, and Jindo, Jeollanamdo for 3 years from 2008. Most SBPH developed to the 3rd and 4th instar on the periods of May 20 to June 10, and they developed to the adult stage for the 1st generation on Mid and Late June. In 2009, all SBPH trapped by sky net trap were adult on May 31 to June 1 at Mid-western aeas of Taean, Seosan and Buan, and South-western areas of Sinan and Jindo. The population density of adult SBPH was 963 heads at Taean, 919 at Seocheon and 819 at Sinan area. The origin of these higher population of adult SBPH were verified from the population of non-overwintered SBPH but immigrant SBPH. From Mid May to Mid June in 2010, adult SBPH could not be counted as immigrant insects by sky net trap. The variation of RSV VIR was high with 2.1% to 9.5% for immigrant adult SBPH trapped by sky net trap at Hongsung of Chungcheongbukdo, Buan of Jeollabukdo and so forth in 2009. The highest VIR for the immigrant adult SBPH was 9.5% at Boryung of Chungcheongnamdo, followed by 7.9% at Hongsung of Chungcheongnamdo, 6.5% at Younggwang of Jeollanamdo, and 6.4% at Taean of Cheongcheongnamdo. The infection rate of RSV on rice plants induced by the immigrant adult SBPH cultivated near sky net trap after about 10 days from immigration on June 12 in 2009 was 84.6% at Taean, 65.4% at Buan and 92.9% at Jindo, and 81% in average through genetic diagnosis of RT-PCR. Barley known as a overwintering host plant of RSV had very low infection rate of 0.2% from 530 specimens collected at 10 areas covering whole country including Pyungtaeg of Gyeonggido. Twenty nine plant species were newly recorded as natural hosts of RSV. In winter annual plant species, 11 plants including Vulpia myuros showed RSV infection rate of 24.9%. The plant species in summer annual ecotype were 13 including Digitaria ciliaris with 44.9%, Echinochloa crusgalli var. echinata with 95.2% and Setaria faberi with 65.5% in infection rate of RSV. Five perennial plants including Miscanths sacchariflorus with infection rate of 33.3% were recorded as hosts of RSV. Rice cultivars, 8 susceptible cultivars including Donggin1 and 17 resistant ones including Samgwang, were screened in field conditions at 3 different areas of Buan, Iksan and Ginje in 2009. All the susceptible cultivars were showed typical symptom of mosaic and wilt. In 17 genetic resistant cultivar, 12 cultivars were susceptible, however, 5 cultivars were field-resistant plus genetic resistant to RSV as non symptom expression. When RSV was artificially inoculated at seedling stage to 4 cultivars known as genetic resistant and 3 cultivars known as genetic susceptible, the symptom expression in resistant cultivars was lower as 19.3% in average than that of 53.3% in susceptible ones. In comparison of symptom expression rate and viral infection rate using resistant Nampyung and susceptible Heugnam cultivars by artificial inoculation of RSV at seedling stage, the symptom expression of Heugnam was higher as 28% than 12% of Nampyung. However, virion infection of resistant Nampyung cultivar was higher as 12% reversely than 85% of susceptible Heugnam. Yield loss of rice was investigated by the artificial inoculation of RSV at the seedling stage of resistant cultivars of Nampyung and Onnuri, and susceptible cultivars of Donggin1 and Ungwang for 3 years from 2008. The average yield per plant was 7.8 g, 8.5 g and 13.8 g on rice plants inoculated at seedling stage, tillering stage and maximum tillering stage, respectively. The yield loss rate was increased by earlier infection of RSV with 51% at seedling stage, 46% at tillering stage and 13% at maximum tillering stage. In resistant rice cultivars, there was no statistically significant relation between infection time and yield loss. In natural fields on susceptible rice cultivar of Ungwang at Taean and Jindo areas in 2009, the yield loss rate was increased with same tendency to the infection hill rate having the corelation coefficient of 0.94 when the viral infection was over 23.4%.

• Journal of Korean Society of Forest Science
• /
• v.16 no.1
• /
• pp.1-32
• /
• 1972

By means of the interspecific hybridization in the Sub-genus Diploxylon of the Genus Pinus, $F_1$ hybrids of Pinus rigida${\times}$elliottii, Pinus rigida${\times}$radiata, P. rigida${\times}$serotina and P. densiflora${\times}$thunbergii had been produced. And on the basis of the crossabilities of these hybrids the taxonomic affinities of these pines were examined. And the needle characteristics of these hybrid and the occurence of phenolic substances in these $F_1$ hybrid were also investigated to see the potential usefulness of these characteristics for the diagnosis of the taxonomic affinity. And, the growth performances of the $F_1$ hybrids have also been compared with those of parental species. In order to contribute to the establishment of the hybrid seed orchard the introgression phenomena between P. densiflora and P. thunbergii in the eastern coastal area have also been investigated along with the investigation of the heterozygosity of plus trees of P. densiflora growing in the clone bank in Suwon. And the results were summarized as follows. 1. On the basis of crossabilities as well as on the taxonomic affinities according to the systems of Shaw, Pilger and Duffield, it has been proven that the parental species of those hybrids are of close affinities and range of the fertile hybrid seed production rate was as high as 28-58% in the best hybrid combination (Table 13). 2. Among those hybrids, the ${\times}$ Pinus, rigiserotina hybrid seemed to be most promising in the growth performance exhibiting 109-155% more volume growth compared to the seed parent with the statistic significance of 1% level (Tables 16 and 17). 3. Notwithstanding the fact that the all of the pollen parents are cold tender, all hybrids exhibit cold hardiness as much as their seed parent and it seems to suggest that the characteristics of cold hardiness were transmitted from the seed parent. 4. Though a striking difference in needle length was observed between the parental species of each hybrid, it was difficult to distinguish each hybrid from their seed parent by the needle length except ${\times}$P. rigiserotina which is characterized by long needle which is 65% more longer than the needle of the seed parent (Table 21). 5. With regard to the anatomical characteristics of needle, the hypoderm is apparently thicker in most of the $F_1$ hybrid pines and the characteristics of resin canals are dominated by medial in most $F_1$ hybrid. And, the fibrovascular bundles were apart as were in their seed parent. Therefore it was found to be possible to distinguish the hybrids pines from their parents by the needle characteristics. And, it is to be noticed that the ${\times}$P. densithunbergii was more close to the pollen parent having RDI value of 0.73 (Fig.l, Table 22). 6. It has been demonstrated that ${\times}$P. rigielliottii, ${\times}$P. rigiradiata and ${\times}$P. rigitaeda have a phenolic substance (No.7) of light yellow at Rf-0.46, same as their seed parent, but no trace of phenolic substance was observed in their pollen parent. This fact will serve as an important criteria for early identification of hybridity in progeny testing. However, the fact that both of ${\times}$P. rigiserotina and ${\times}$P. densithunbergii exhibit the same reactions of phenolic substances as well their parental species seems to indicate the close affinities between the parental species of the respective hybrid (Fig.2, Table 23). 7. The separation and the reaction of phenolic substance developed on TLC were found to be same in the same species showing no variations between the individuals, and no variations due to tree part of sampling, tree age or pollen sources. And the reaction was also observed regardless of the not varied by the kind of developing solvent whether it is Aceton-Chloroform (3:7 v/v) or Benzene-Methanol-Acetic acid (90:16:8 v/v). 8. The introgression phenomena of natural Pinus densifiora stand in both east and west coastal area indicates that the major part of the red pines investigated are all heterozygous and the heterozygosity of pines are higher in the west coast than in the east coast(Tables 24 and 25). 9. Based on the RDI, among the plus trees of Pinus densiflora selected in Korea and Japan as well, no pure P. densiflora has been found. Since all of the sample trees of Pinus densiflora were found to be as heterozygous bearing part of the characteristics of P. thunbergii, those red pines were considered to be natural heterotic hybrid pines(Figs. 3 and 4. Tables 26 and 27).

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.53 no.spc
• /
• pp.96-102
• /
• 2008

This study was carried out to examine agronomic characteristics and seed qualities of 827 Korean soybean landraces according to the collecting areas. Agronomic characteristics and seed qualities was examined according to the localities by weather characteristics. Days to flowering was the longest (mean 70.9 days) in the northwest seashore, and the most short in the south inland (mean 63.8 days). Growth days was the longest in the Gyeongbuk inland (mean 136.0 days), and the most short in the mountain-high (mean 132.8 days). The stem height was the most short in the north-west seashore (mean 81.3 cm), and in the south seashore (mean 58.8 cm). The seed weight of 100 seeds was the lightest in the north-west seashore (mean 17.1 g), and the heaviest in the south-west seashore (mean 21.5 g). The crude protein contents was the lowest in the north-west seashore (mean 40.1%), the highest in the south seashore (mean 42.8%), and vice versa in the crude oil. The contents of sucrose and oligo-saccharide was the highest 5.8% and 8.6%, respectively, in the mountain-high, and the lowest 4.8% and 7.6% in the Gyeongbuk inland and south seashore, respectively. Among the iso-flavone contests, the contents of genistein and total was the highest, $635.6\;{\mu}g/g$ and $1258.1\;{\mu}g/g$, respectively, in the mountain inland, and the lowest $319.2\;{\mu}g/g$ and $835.3\;{\mu}g/g$ in the south-west seashore. The contents of glycitein was the highest $124.3\;{\mu}g/g$ in the south seashore.

• Asia pacific journal of information systems
• /
• v.22 no.3
• /
• pp.99-124
• /
• 2012

Our society has long been talking about necessity for innovation. Since companies in particular need to carry out business innovation in their overall processes, they have attempted to apply many innovation factors on sites and become to pay more attention to their innovation. In order to achieve this goal, companies has applied various information technologies (IT) on sites as a means of innovation, and consequently IT have been greatly developed. It is natural for the field of IT to have faced another revolution which is called cloud computing, which is expected to result in innovative changes in software application via the Internet, data storing, the use of devices, and their operations. As a vehicle of innovation, cloud computing is expected to lead the changes and advancement of our society and the business world. Although many scholars have researched on a variety of topics regarding the innovation via IT, few studies have dealt with the issue of could computing as IT. Thus, the purpose of this paper is to set the variables of innovation attributes based on the previous articles as the characteristic variables and clarify how these variables affect "Performance Expectancy" of companies and the intention of using cloud computing. The result from the analysis of data collected in this study is as follows. The study utilized a research model developed on the innovation diffusion theory to identify influences on the adaptation and spreading IT for cloud computing services. Second, this study summarized the characteristics of cloud computing services as a new concept that introduces innovation at its early stage of adaptation for companies. Third, a theoretical model is provided that relates to the future innovation by suggesting variables for innovation characteristics to adopt cloud computing services. Finally, this study identified the factors affecting expectation and the intention to use the cloud computing service for the companies that consider adopting the cloud computing service. As the parameter and dependent variable respectively, the study deploys the independent variables that are aligned with the characteristics of the cloud computing services based on the innovation diffusion model, and utilizes the expectation for performance and Intention to Use based on the UTAUT theory. Independent variables for the research model include Relative Advantage, Complexity, Compatibility, Cost Saving, Trialability, and Observability. In addition, 'Acceptance for Adaptation' is applied as an adjustment variable to verify the influences on the expected performances from the cloud computing service. The validity of the research model was secured by performing factor analysis and reliability analysis. After confirmatory factor analysis is conducted using AMOS 7.0, the 20 hypotheses are verified through the analysis of the structural equation model, accepting 12 hypotheses among 20. For example, Relative Advantage turned out to have the positive effect both on Individual Performance and on Strategic Performance from the verification of hypothesis, while it showed meaningful correlation to affect Intention to Use directly. This indicates that many articles on the diffusion related Relative Advantage as the most important factor to predict the rate to accept innovation. From the viewpoint of the influence on Performance Expectancy among Compatibility and Cost Saving, Compatibility has the positive effect on both Individual Performance and on Strategic Performance, while it showed meaningful correlation with Intention to Use. However, the topic of the cloud computing service has become a strategic issue for adoption in companies, Cost Saving turns out to affect Individual Performance without a significant influence on Intention to Use. This indicates that companies expect practical performances such as time and cost saving and financial improvements through the adoption of the cloud computing service in the environment of the budget squeezing from the global economic crisis from 2008. Likewise, this positively affects the strategic performance in companies. In terms of effects, Trialability is proved to give no effects on Performance Expectancy. This indicates that the participants of the survey are willing to afford the risk from the high uncertainty caused by innovation, because they positively pursue information about new ideas as innovators and early adopter. In addition, they believe it is unnecessary to test the cloud computing service before the adoption, because there are various types of the cloud computing service. However, Observability positively affected both Individual Performance and Strategic Performance. It also showed meaningful correlation with Intention to Use. From the analysis of the direct effects on Intention to Use by innovative characteristics for the cloud computing service except the parameters, the innovative characteristics for the cloud computing service showed the positive influence on Relative Advantage, Compatibility and Observability while Complexity, Cost saving and the likelihood for the attempt did not affect Intention to Use. While the practical verification that was believed to be the most important factor on Performance Expectancy by characteristics for cloud computing service, Relative Advantage, Compatibility and Observability showed significant correlation with the various causes and effect analysis. Cost Saving showed a significant relation with Strategic Performance in companies, which indicates that the cost to build and operate IT is the burden of the management. Thus, the cloud computing service reflected the expectation as an alternative to reduce the investment and operational cost for IT infrastructure due to the recent economic crisis. The cloud computing service is not pervasive in the business world, but it is rapidly spreading all over the world, because of its inherited merits and benefits. Moreover, results of this research regarding the diffusion innovation are more or less different from those of the existing articles. This seems to be caused by the fact that the cloud computing service has a strong innovative factor that results in a new paradigm shift while most IT that are based on the theory of innovation diffusion are limited to companies and organizations. In addition, the participants in this study are believed to play an important role as innovators and early adapters to introduce the cloud computing service and to have competency to afford higher uncertainty for innovation. In conclusion, the introduction of the cloud computing service is a critical issue in the business world.

• Journal of Intelligence and Information Systems
• /
• v.22 no.3
• /
• pp.45-69
• /
• 2016

Recently, sentiment analysis using open Internet data is actively performed for various purposes. As online Internet communication channels become popular, companies try to capture public sentiment of them from online open information sources. This research is conducted for the purpose of analyzing pulbic sentiment of Korean Top-10 companies using a multi-categorical sentiment lexicon. Whereas existing researches related to public sentiment measurement based on big data approach classify sentiment into dimensions, this research classifies public sentiment into multiple categories. Dimensional sentiment structure has been commonly applied in sentiment analysis of various applications, because it is academically proven, and has a clear advantage of capturing degree of sentiment and interrelation of each dimension. However, the dimensional structure is not effective when measuring public sentiment because human sentiment is too complex to be divided into few dimensions. In addition, special training is needed for ordinary people to express their feeling into dimensional structure. People do not divide their sentiment into dimensions, nor do they need psychological training when they feel. People would not express their feeling in the way of dimensional structure like positive/negative or active/passive; rather they express theirs in the way of categorical sentiment like sadness, rage, happiness and so on. That is, categorial approach of sentiment analysis is more natural than dimensional approach. Accordingly, this research suggests multi-categorical sentiment structure as an alternative way to measure social sentiment from the point of the public. Multi-categorical sentiment structure classifies sentiments following the way that ordinary people do although there are possibility to contain some subjectiveness. In this research, nine categories: 'Sadness', 'Anger', 'Happiness', 'Disgust', 'Surprise', 'Fear', 'Interest', 'Boredom' and 'Pain' are used as multi-categorical sentiment structure. To capture public sentiment of Korean Top-10 companies, Internet news data of the companies are collected over the past 25 months from a representative Korean portal site. Based on the sentiment words extracted from previous researches, we have created a sentiment lexicon, and analyzed the frequency of the words coming up within the news data. The frequency of each sentiment category was calculated as a ratio out of the total sentiment words to make ranks of distributions. Sentiment comparison among top-4 companies, which are 'Samsung', 'Hyundai', 'SK', and 'LG', were separately visualized. As a next step, the research tested hypothesis to prove the usefulness of the multi-categorical sentiment lexicon. It tested how effective categorial sentiment can be used as relative comparison index in cross sectional and time series analysis. To test the effectiveness of the sentiment lexicon as cross sectional comparison index, pair-wise t-test and Duncan test were conducted. Two pairs of companies, 'Samsung' and 'Hanjin', 'SK' and 'Hanjin' were chosen to compare whether each categorical sentiment is significantly different in pair-wise t-test. Since category 'Sadness' has the largest vocabularies, it is chosen to figure out whether the subgroups of the companies are significantly different in Duncan test. It is proved that five sentiment categories of Samsung and Hanjin and four sentiment categories of SK and Hanjin are different significantly. In category 'Sadness', it has been figured out that there were six subgroups that are significantly different. To test the effectiveness of the sentiment lexicon as time series comparison index, 'nut rage' incident of Hanjin is selected as an example case. Term frequency of sentiment words of the month when the incident happened and term frequency of the one month before the event are compared. Sentiment categories was redivided into positive/negative sentiment, and it is tried to figure out whether the event actually has some negative impact on public sentiment of the company. The difference in each category was visualized, moreover the variation of word list of sentiment 'Rage' was shown to be more concrete. As a result, there was huge before-and-after difference of sentiment that ordinary people feel to the company. Both hypotheses have turned out to be statistically significant, and therefore sentiment analysis in business area using multi-categorical sentiment lexicons has persuasive power. This research implies that categorical sentiment analysis can be used as an alternative method to supplement dimensional sentiment analysis when figuring out public sentiment in business environment.

• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.71-88
• /
• 2017

Language models were originally developed for speech recognition and language processing. Using a set of example sentences, a language model predicts the next word or character based on sequential input data. N-gram models have been widely used but this model cannot model the correlation between the input units efficiently since it is a probabilistic model which are based on the frequency of each unit in the training set. Recently, as the deep learning algorithm has been developed, a recurrent neural network (RNN) model and a long short-term memory (LSTM) model have been widely used for the neural language model (Ahn, 2016; Kim et al., 2016; Lee et al., 2016). These models can reflect dependency between the objects that are entered sequentially into the model (Gers and Schmidhuber, 2001; Mikolov et al., 2010; Sundermeyer et al., 2012). In order to learning the neural language model, texts need to be decomposed into words or morphemes. Since, however, a training set of sentences includes a huge number of words or morphemes in general, the size of dictionary is very large and so it increases model complexity. In addition, word-level or morpheme-level models are able to generate vocabularies only which are contained in the training set. Furthermore, with highly morphological languages such as Turkish, Hungarian, Russian, Finnish or Korean, morpheme analyzers have more chance to cause errors in decomposition process (Lankinen et al., 2016). Therefore, this paper proposes a phoneme-level language model for Korean language based on LSTM models. A phoneme such as a vowel or a consonant is the smallest unit that comprises Korean texts. We construct the language model using three or four LSTM layers. Each model was trained using Stochastic Gradient Algorithm and more advanced optimization algorithms such as Adagrad, RMSprop, Adadelta, Adam, Adamax, and Nadam. Simulation study was done with Old Testament texts using a deep learning package Keras based the Theano. After pre-processing the texts, the dataset included 74 of unique characters including vowels, consonants, and punctuation marks. Then we constructed an input vector with 20 consecutive characters and an output with a following 21st character. Finally, total 1,023,411 sets of input-output vectors were included in the dataset and we divided them into training, validation, testsets with proportion 70:15:15. All the simulation were conducted on a system equipped with an Intel Xeon CPU (16 cores) and a NVIDIA GeForce GTX 1080 GPU. We compared the loss function evaluated for the validation set, the perplexity evaluated for the test set, and the time to be taken for training each model. As a result, all the optimization algorithms but the stochastic gradient algorithm showed similar validation loss and perplexity, which are clearly superior to those of the stochastic gradient algorithm. The stochastic gradient algorithm took the longest time to be trained for both 3- and 4-LSTM models. On average, the 4-LSTM layer model took 69% longer training time than the 3-LSTM layer model. However, the validation loss and perplexity were not improved significantly or became even worse for specific conditions. On the other hand, when comparing the automatically generated sentences, the 4-LSTM layer model tended to generate the sentences which are closer to the natural language than the 3-LSTM model. Although there were slight differences in the completeness of the generated sentences between the models, the sentence generation performance was quite satisfactory in any simulation conditions: they generated only legitimate Korean letters and the use of postposition and the conjugation of verbs were almost perfect in the sense of grammar. The results of this study are expected to be widely used for the processing of Korean language in the field of language processing and speech recognition, which are the basis of artificial intelligence systems.

• Journal of Intelligence and Information Systems
• /
• v.23 no.1
• /
• pp.95-108
• /
• 2017

Recently, AlphaGo which is Bakuk (Go) artificial intelligence program by Google DeepMind, had a huge victory against Lee Sedol. Many people thought that machines would not be able to win a man in Go games because the number of paths to make a one move is more than the number of atoms in the universe unlike chess, but the result was the opposite to what people predicted. After the match, artificial intelligence technology was focused as a core technology of the fourth industrial revolution and attracted attentions from various application domains. Especially, deep learning technique have been attracted as a core artificial intelligence technology used in the AlphaGo algorithm. The deep learning technique is already being applied to many problems. Especially, it shows good performance in image recognition field. In addition, it shows good performance in high dimensional data area such as voice, image and natural language, which was difficult to get good performance using existing machine learning techniques. However, in contrast, it is difficult to find deep leaning researches on traditional business data and structured data analysis. In this study, we tried to find out whether the deep learning techniques have been studied so far can be used not only for the recognition of high dimensional data but also for the binary classification problem of traditional business data analysis such as customer churn analysis, marketing response prediction, and default prediction. And we compare the performance of the deep learning techniques with that of traditional artificial neural network models. The experimental data in the paper is the telemarketing response data of a bank in Portugal. It has input variables such as age, occupation, loan status, and the number of previous telemarketing and has a binary target variable that records whether the customer intends to open an account or not. In this study, to evaluate the possibility of utilization of deep learning algorithms and techniques in binary classification problem, we compared the performance of various models using CNN, LSTM algorithm and dropout, which are widely used algorithms and techniques in deep learning, with that of MLP models which is a traditional artificial neural network model. However, since all the network design alternatives can not be tested due to the nature of the artificial neural network, the experiment was conducted based on restricted settings on the number of hidden layers, the number of neurons in the hidden layer, the number of output data (filters), and the application conditions of the dropout technique. The F1 Score was used to evaluate the performance of models to show how well the models work to classify the interesting class instead of the overall accuracy. The detail methods for applying each deep learning technique in the experiment is as follows. The CNN algorithm is a method that reads adjacent values from a specific value and recognizes the features, but it does not matter how close the distance of each business data field is because each field is usually independent. In this experiment, we set the filter size of the CNN algorithm as the number of fields to learn the whole characteristics of the data at once, and added a hidden layer to make decision based on the additional features. For the model having two LSTM layers, the input direction of the second layer is put in reversed position with first layer in order to reduce the influence from the position of each field. In the case of the dropout technique, we set the neurons to disappear with a probability of 0.5 for each hidden layer. The experimental results show that the predicted model with the highest F1 score was the CNN model using the dropout technique, and the next best model was the MLP model with two hidden layers using the dropout technique. In this study, we were able to get some findings as the experiment had proceeded. First, models using dropout techniques have a slightly more conservative prediction than those without dropout techniques, and it generally shows better performance in classification. Second, CNN models show better classification performance than MLP models. This is interesting because it has shown good performance in binary classification problems which it rarely have been applied to, as well as in the fields where it's effectiveness has been proven. Third, the LSTM algorithm seems to be unsuitable for binary classification problems because the training time is too long compared to the performance improvement. From these results, we can confirm that some of the deep learning algorithms can be applied to solve business binary classification problems.

• KDI Journal of Economic Policy
• /
• v.12 no.3
• /
• pp.71-94
• /
• 1990

This paper reviews the rationale of monetary and banking intervention by an outside authority, either the government or the central bank, and seeks to delineate clearly the optimal limits to the monetary and banking deregulation currently underway in Korea as well as on a global scale. Furthermore, this paper seeks to establish an objective and balanced view on the role of the central bank, especially in light of the current discussion on the restructuring of Korea's central bank, which has been severely contaminated by interest-group politics. The discussion begins with the recognition that the modern free banking school and the new monetary economics are becoming formidable challenges to the traditional role of the government or the central bank in the monetary and banking sector. The paper reviews six arguments that have traditionally been presented to support intervention: (1) the possibility of an over-issue of bank notes under free banking instead of central banking; (2) externalities in and the public good nature of the use of money; (3) economies of scale and natural monopoly in producing money; (4) the need for macro stabilization policy due to the instability of the real sector; (5) the external effects of bank failure due to the inherent instability of the existing banking system; and (6) protection for small banknote users and depositors. Based on an analysis of the above arguments, the paper speculates on the optimal role of the government or central bank in the monetary and banking system and the optimal degree of monetary and banking deregulation. By contrast to the arguments for free banking or laissez-faire monetary systems, which become fashionable in recent years, monopoly and intervention by the government or central bank in the outside money system can be both necessary and optimal. In this case, of course, an over-issue of fiat money may be possible due to political considerations, but this issue is beyond the scope of this paper. On the other hand, the issue of inside monies based on outside money could indeed be provided for optimally under market competition by private institutions. A competitive system in issuing inside monies would help realize, to the maxim urn extent possible, external economies generated by using a single outside money. According to this reasoning, free banking activities will prevail in the inside money system, while a government monopoly will prevail in the outside money system. This speculation, then, also implies that the monetary and banking deregulation currently underway should and most likely will be limited to the inside money system, which could be liberalized to the fullest degree. It is also implied that it will be impractical to deregulate the outside money system and to allow market competition to provide outside money, in accordance with the arguments of the free banking school and the new monetary economics. Furthermore, the role of the government or central bank in this new environment will not be significantly different from their current roles. As far as the supply of fiat money continues to be monopolized by the government, the control of the supply of base money and such related responsibilities as monetary policy (argument(4)) and the lender of the last resort (argument (5)) will naturally be assigned to the outside money supplier. However, a mechanism for controlling an over-issue of fiat money by a monopolistic supplier will definitely be called for (argument(1)). A monetary policy based on a certain policy rule could be one possibility. More importantly, the deregulation of the inside money system would further increase the systemic risk inherent in the current fractional banking system, while enhancing the efficiency of the system (argument (5)). In this context, the role of the lender of the last resort would again become an instrument of paramount importance in alleviating liquidity crises in the early stages, thereby disallowing the possibility of a widespread bank run. Similarly, prudential banking supervision would also help maintain the safety and soundness of the fully deregulated banking system. These functions would also help protect depositors from losses due to bank failures (argument (6)). Finally, these speculations suggest that government or central bank authorities have probably been too conservative on the issue of the deregulation of the financial system, beyond the caution necessary to preserve system safety. Rather, only the fullest deregulation of the inside money system seems to guarantee the maximum enjoyment of external economies in the single outside money system.

• Journal of The Korean Society of Grassland and Forage Science
• /
• v.27 no.3
• /
• pp.209-218
• /
• 2007

This experiment was conducted to investigate effect of plowing frequency and sowing dates on the agronomic characteristics, feed value, weed yield and palatability of silage corn. Treatments were a basal treatment(C: May 5 seeding, plowing once, weeding control once), T1(May 12 seeding, plowing twice, weeding control 0 time), T2(May 19 seeding, plowing three times, weeding control 0 time, T3(May 26 seeding, plowing four times, weeding control 0 time) and T4(June 2 seeding, plowing five times, weeding control 0 time). The experiment was performed at the College of Life and Natural Sciences of Sangju University in Sangju in 2006. The plant height and ear height showed highly in order to C > T1 > T2 > T3 > T4 treatment, leaf length was the highest at T2 (96.0cm). Leaf width and number of dead leaf were the highest at C and T3 (11.2cm), C, C and T1 (4.6), respectively. Stem diameter was the highest at T3 as 31mm, while T2 was the lowest as 25mm (p<0.05). Ear circle showed highly in order of C > T1 > T4 > T3 > T4 (p<0.05), and tip filling degree was the highest at C treatment as 8.8, while T4 treatment was the lowest as 6.0 (p<0.05). The stem hardness and grain hardness were C < T1 < T2 < T3 < T4 (p<0.05). Stem saccharinity was T1(6.1%) was the highest, while T2(3.0%) was the lowest(P<0.05). Fresh yield of weed was the lowest at C treatment as 500 kg/ha, but T1 treatment was the highest as 44,100 kg/ha (p<0.05). Weed coverage rate showed highly in order of T1 > T2 > T3 > T4 > C treatment (p<0.05). Fresh yield of corn was the highest at C treatment as T3,550 kg/ha, but T4 treatment was the lowest as 65,500 kg/ha (p<0.05). Dry matter yield of corn showed highly in order of C(26,978 kg/ha) > T1(26,130 kg/ha) > T2(20,255 kg/ha) > T3(20,255 kg/ha) > T4(17,508 kg/ha) treatment (p<0.05). Crude protein content was T1(7.69%) > T4(7.42%) > T2(6.34%) > T3(5.99%) > C(5.91%) treatment (p<0.05), and Crude fat content showed highly in order of C (2.13%) > T1(2.04%) > T2(1.96%) > T3(1.95%) > T4(1.84%) treatment. Relative palatability of Holstein, Korean native goat and spotted deer was the highest at C treatment, but Korean native cattle was the highest at T1 treatment.