• Journal of Intelligence and Information Systems
• /
• v.24 no.2
• /
• pp.221-241
• /
• 2018

Deep learning is getting attention recently. The deep learning technique which had been applied in competitions of the International Conference on Image Recognition Technology(ILSVR) and AlphaGo is Convolution Neural Network(CNN). CNN is characterized in that the input image is divided into small sections to recognize the partial features and combine them to recognize as a whole. Deep learning technologies are expected to bring a lot of changes in our lives, but until now, its applications have been limited to image recognition and natural language processing. The use of deep learning techniques for business problems is still an early research stage. If their performance is proved, they can be applied to traditional business problems such as future marketing response prediction, fraud transaction detection, bankruptcy prediction, and so on. So, it is a very meaningful experiment to diagnose the possibility of solving business problems using deep learning technologies based on the case of online shopping companies which have big data, are relatively easy to identify customer behavior and has high utilization values. Especially, in online shopping companies, the competition environment is rapidly changing and becoming more intense. Therefore, analysis of customer behavior for maximizing profit is becoming more and more important for online shopping companies. In this study, we propose 'CNN model of Heterogeneous Information Integration' using CNN as a way to improve the predictive power of customer behavior in online shopping enterprises. In order to propose a model that optimizes the performance, which is a model that learns from the convolution neural network of the multi-layer perceptron structure by combining structured and unstructured information, this model uses 'heterogeneous information integration', 'unstructured information vector conversion', 'multi-layer perceptron design', and evaluate the performance of each architecture, and confirm the proposed model based on the results. In addition, the target variables for predicting customer behavior are defined as six binary classification problems: re-purchaser, churn, frequent shopper, frequent refund shopper, high amount shopper, high discount shopper. In order to verify the usefulness of the proposed model, we conducted experiments using actual data of domestic specific online shopping company. This experiment uses actual transactions, customers, and VOC data of specific online shopping company in Korea. Data extraction criteria are defined for 47,947 customers who registered at least one VOC in January 2011 (1 month). The customer profiles of these customers, as well as a total of 19 months of trading data from September 2010 to March 2012, and VOCs posted for a month are used. The experiment of this study is divided into two stages. In the first step, we evaluate three architectures that affect the performance of the proposed model and select optimal parameters. We evaluate the performance with the proposed model. Experimental results show that the proposed model, which combines both structured and unstructured information, is superior compared to NBC(Naïve Bayes classification), SVM(Support vector machine), and ANN(Artificial neural network). Therefore, it is significant that the use of unstructured information contributes to predict customer behavior, and that CNN can be applied to solve business problems as well as image recognition and natural language processing problems. It can be confirmed through experiments that CNN is more effective in understanding and interpreting the meaning of context in text VOC data. And it is significant that the empirical research based on the actual data of the e-commerce company can extract very meaningful information from the VOC data written in the text format directly by the customer in the prediction of the customer behavior. Finally, through various experiments, it is possible to say that the proposed model provides useful information for the future research related to the parameter selection and its performance.

• Korean Journal of Psychosomatic Medicine
• /
• v.8 no.1
• /
• pp.31-45
• /
• 2000

Objectives : The present study was performed to reveal differences between female and male cases of alopecia in their alopecia related variables such as patterns of hair loss, psychiatric characteristics, associate illnesses, and methods of treatment, and to use them as basic materials for proper management and early prevention of the alopecia prone cases. Methods : In order to analysis the gender difference in hair losses, the subjects were divided into two subgroups as the 51 cases of female alopecia and the 42 cases of male alopecia, who had visited to the department of psychiatry consulted from the department of dermatology, Yongsan hopital, ChungAng University, Seoul, Korea, from January 1998 to December 1998. In data analysis, the subjects were statistically assesed by chi-squre test and analysis of varaiance, through SPSS-$PC^+$ 9.0V. Results : 1) Female subjects were more likely showed lower socio-economical level including lower eonomical level, lower educational level, or lower occupational level in their parent's job, were more likely to have larger number of siblings and to have many sisters comparison to the male cases. 2) Female subjects were more likely visited to the department of dermatology, more history of alopecia in their female family members, lesser history of alopecia in their male family members, more loss of hairs in vertex or frontal region of scalp, lesser loss of hairs in occipital region, and lesser nail changes in comparison to the male cases. 3) Female subjects were more suffered from intra-familial conflicts and economical changes, or their introverted personality makeup, lesser likely suffered from changes of business and health changes, and showed lesser conflicts related with poorer adaptaion in their job life. 4) Female subjects were more likely diagnosed as depression or conversion disorders, more frequently complaint anxiety symptoms or depressive symptoms, higher level of anxiety index, lesser complaint somatization or obsessive compulsive symptoms, and lesser diagnosed as anxiety disorder in comparison to the male cases. 5) Female subjects were more likely tended to show personality makeup such as the introverted, the lie, the repressed, or the feminine trends than the male cases. 6) Female subjects were more significantly treated by antianxiety drug such as etizolam and dermatological therapies include tretinoin, and lesser treated by clotiazepam and prednicarbonate in comparison to the male cases. Conclusion : From the facts that The most important factors in developing hair loss in the female subjects in comparison to the male cases seems to be closely correlated with the serious psychopathology such as the presence of mental disorders including depression, the presence of complaining anxiety or depressive symptomatology, the presence of stressful life events such as intrafamilial life changes, and the presence of personality makeup such as the introverted, the lie, the repressed, or the feminine trends, the authors confirmed that dermatologists act as the primary care physician are in a unique position to recognize psychiatric comorbidity and execute meaningful intervention for female patients with the alopecia with psychiatrists.

• Journal of Animal Science and Technology
• /
• v.49 no.1
• /
• pp.137-146
• /
• 2007

This study was conducted to determine fertilizer nutrient and pollutant production of Holstein dairy cattle by estimating manure characteristics. The moisture content of feces was 83.9% and 95.1% for urine. The pH of feces and urine were in the ranges of 7.0～7.4 and 7.5～7.8, respectively. The average BOD5, COD, SS, T-N, T-P concentrations of the dairy feces were 18,294, 52,765, 102,889, 2,575, and 457mg/ℓ, respectively. Dairy urine showed lower levels of BOD5(5,455mg/ℓ), COD(8,089mg/ℓ), SS(593mg/ℓ), T-N(3,401mg/l), and T-P(13mg/ℓ) than feces. The total daily produced pollutant amounts of a dairy cow were 924.1g(Milking cow), 538.8g(Dry cow), 284.4g(Heifer) of BOD5, 2,336.5g (Milking cow), 1,651.8g(Dry cow), 734.1g(Heifer) of COD and 4,210.1g(Milking cow), 2,417.1g(Dry cow), 1,629.1g(Heifer) of SS and 194.8g(Milking cow), 96.4g(Dry cow), 58.3g(Heifer) of T-N and 24.0g(Milking cow), 10.2g(Dry cow), 6.1g(Heifer) of T-P. The calculated amount of pollutants produced by a 450kg dairy cow for one year were 181.3kg of BOD5, 492.5kg of COD, 899.9kg of SS, 36.0kg of T-N and 4.1kg of T-P. The total yearly estimated pollutant production from all head(497,261) of dairy cattle in Korea is 90,149 tons of BOD5, 244,890 tons of COD, 447,491 tons of SS, 17,898 tons of T-N and 2,008 tons of T-P. The fertilizer nutrient concentrations of dairy feces was 0.26% N, 0.1% P2O5 and 0.14% K2O. Urine was found to contain 0.34% N, 0.003% of P2O5 and 0.31% K2O. The total daily fertilizer nutrients produced by dairy cattle were 197.4g (Milking cow), 97.4g(Dry cow), and 57.9g(Heifer) of Nitrogen, 54.2g(Milking cow), 22.2g(Dry cow), and 14.2g(Heifer) of P2O5 and 110.8g(Milking cow), 80.4g (Dry cow), and 39.5g(Heifer) of K2O. The total yearly estimated fertilizer nutrient produced by a 450kg dairy animal is 36.2kg of N, 8.8kg of P2O5, 24.6kg of K2O. The estimated yearly fertilizer nutrient production from all dairy cattle in Korea is 18,000 tons of N, 4,397 tons of P2O5, 12,206 tons of K2O. Dairy manure contains useful trace minerals for crops, such as CaO and MgO, which are contained in similar levels to commercial compost being sold in the domestic market. Concentrations of harmful trace minerals, such as As, Cd, Hg, Pb, Cr, Cu, Ni, Zn, met the Korea compost standard regulations, with some of these minerals being in undetected amounts.

• Asia Marketing Journal
• /
• v.14 no.1
• /
• pp.99-130
• /
• 2012

Service failure and a poor service recovery may lead loyal customers to try to aggressively punish the service firm. We use perceived betrayal and desire for vengeance as the key constructs to understand customer retaliation. Perceived betrayal is defined as a customer's belief that a firm has intentionally violated what is normative in the context of their relationship. And the desire for vengeance is defined as the retaliatory feelings that consumers feel toward a firm, such as the desire to exert harm on the firm. The perceived betrayal and the desire for vengeance are key antecedents of retaliatory behaviors such as vindictive complaining, negative WOM and third-party complaining for publicity. The empirical results suggest that betrayal is a key motivational factor that lead customers to restore fairness by making use of all means, including retaliation. We also find that relationship quality has effect on a customer's response to a failure in service recovery. As the levels of relationship increases, a violation of the proper fairness has a stronger effect on the sense of betrayal experienced by customers. Considerable research has investigated consumer responses to dissatisfaction. But our study examine the response of outraged and highly frustrated consumers. We focus on emotional and behavioral processes that have not been covered by previous dissatisfaction researches and which are unique to outraged consumers caused by extremely dissatisfied purchase experience. It has recently been pointed out by various mass media that the customers not only have positive effects on the company performance but also put the company in crisis. It has often been reported that one customer's dissatisfaction, for example, never ends as it is, and it tends to grow for retaliating upon the company, depending on the level of seriousness of the dissatisfaction. This sometimes leads to a lawsuit against the company. Our study focuses on the customers' emotional and behavioral responses induced by their extreme dissatisfactions. We divided the customer groups into the customers with high relationship quality and the customers with low relationship quality, and the difference between two groups is examined. The objective of this study is to comprehend the causal relationship between the feeling of betrayal caused by the service failure and the retaliatory behavior triggered by the desire of revenge. Our study is divided into three parts. First, a causal relationship between perceived unfairness and the perceived betrayal and desire for revenge. Second, the effect of the perceived betrayal and desire for revenge on the retaliatory behavior is investigated. Finally, the moderating role of relationship quality in the causal relationship between the unfairness in service recovery and the perceived betrayal is analyzed. This study finds the following empirical results. The distributive unfairness, procedural unfairness and interactional unfairness had significant effects on the perceived betrayal. Especially, the perceived distributive unfairness results in the highest perceived betrayal. When the service company does not provide customers proper and sufficient compensation for the failure, they feel the strong sense of betrayal. And in the causal relationship between the perceived betrayal, desire for revenge and retaliatory behavior, the perceived betrayal has significant effects on e desire for revenge. In addition desire for revenge has significant effects on negative word of mouth, retaliatory complaining behavior and publicity of complaints through third group. Therefore the perceived unfairness has effects on retaliatory behavior through the mediation of the perceived betrayal and desire for revenge. Finally the moderating role of relationship quality was examined in the relationship between the unfairness and perceived betrayal. If the customers experienced the perceived unfairness in the process of service recovery, the customers with high relationship quality feel the stronger perceived betrayal than the customers with low relationship quality do. When they experience the double service failure, the customer group with high relationship quality accumulating the sense of trust feel the more perceived betrayal than the customer with low relationship quality who do not have strong trust. The contribution of this study is to find the effect of the service failure on the retaliatory behavior with the moderating roles of relationship quality. The dimensions of unfairness in service recovery is found to have differential effects on the perceived betrayal, desire for revenge. And these differential effect is moderated by the level of relationship quality.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.2 no.2
• /
• pp.93-120
• /
• 2007

To promote the commercialization of high technology based start-up company, it is essential to activate technology trading for the Innovative Small & Medium companies which eager to acquire technology competitiveness through technology trading and facilitate financial system which provide Small & Medium companies with technology commercialization money. This study focus on enhancing credibility of technology valuation solutions through the improvement of the existing technology valuation solutions. The existing technology valuation solutions in practice have deficiency in reasonable valuation because of subjective technology factor's calculation. And the commercialization risk of high technology based start-up companies cannot be fully reflected in the existing valuation solutions. The high risk of technology commercialization acts as a hurdle in the facilitation of providing money for the start-up companies. The improved new model will have more acceptable objectivity and reasonably reflect the commercialization risk of start-up companies. The new model has two distinctive features compared to existing solutions. The new model newly adopts commercialization success factor which reflects the risk of high technology based start-up company. And the new model excludes technology factor which is controversial among interested parties. The exclusion of technology factor will improve the objectivity of technology valuation and the adoption of commercialization success factor will solicit investor and capitalist who concern the high risk of technology based start-up companies. In conclusion, the improved new model is expected to activate technology trading and facilitate the money market through which high technology based start-up companies raise commercialization money.

• The Korean Journal of Pesticide Science
• /
• v.18 no.3
• /
• pp.202-209
• /
• 2014

RNA interference (RNAi) is the method which controls phenotypes of gene in live cells. Chitinase is the enzyme helping digestion and absorption of old cuticles during the ecdysis of insects. In order to investigate molting-inhibition effect with the chitinase related gene in Spodoptera litura, RNA was extracted from the $5^{th}$ instars. cDNA was synthesized and then we obtained about 700 bp size chitinase. After PCR products were cloned into a pGEM T-easy vector, colonies were picked. DNA was extracted from the colony cultures. EcoR I enzyme was used to check whether PCR products were inserted or not. And then we confirmed vector band of about 3 kb and insert band of about 700 bp. To synthesize the dsRNA, each DNA was cut with Spe I and Nco I enzymes (Circular DNA became lineared DNA). After synthesis of dsRNA, approximately 5 ul dsRNA was injected into the $3^{rd}$ abdominal segment of S. litura $4^{th}$ larvae. The concentration of dsRNA was about $10{\mu}g/{\mu}l$. We confirmed larval-larval molting : there were phenotypically abnormal individuals - for instance malformation, molting inhibition and change of integument color. Pupaadult molting : there were phenotypically abnormal individuals - for instance molting inhibition, change of wings and malformation. Also we could investigate the pupation, emergence and variation about noninjection, treated with DW and dsRNA. Each pupation was non-injection 83.3%, DW 78.3% and dsRNA 66.7%. Each emergence was non-injection 90.0%, DW 72.3% and dsRNA 65.0%. So we considered that chitinase dsRNA induced molting inhibition effect. But each variation was non-injection 8.9%, DW 2.9% and dsRNA 19.2%. Therefore dsRNA group showed the highest variation value. When 18 hours after injecting dsRNA, we could obtain abnormal individual.

• Journal of Korean Society of Environmental Engineers
• /
• v.27 no.7
• /
• pp.746-752
• /
• 2005

In this study, the effect of reaction parameters including reaction temperature, time, and pressure on sludge degradation and conversion to intermediates such as organic acids were investigated at low critical wet air oxidation(LC-WAO) conditions. Degradation pathways and a modified kinetic model in LC-WAO were proposed and the kinetics model predictions were compared with experimental data under various conditions. Results in the batch experiments showed that reaction temperature directly affected the thermal hydrolysis reaction rather than oxidation reaction. The efficiencies of sludge degradation and organic acid formation increased with the increase of the reaction temperature and time. The removal of SS at $180^{\circ}C$, $200^{\circ}C$, $220^{\circ}C$ and $240^{\circ}C$ of reaction temperatures and 10 min of reaction time were 52.6%, 68.3%, 72.6%, and 74.4%, respectively, indicating that most organic suspended solids were liquified at early stage of reaction. At $180^{\circ}C$, $200^{\circ}C$, $220^{\circ}C$ and $240^{\circ}C$ of reaction temperatures and 40 min of reaction time, the amounts of organic acids formed from 1 g of sludge were 93.5 mg/g SS, 116.4 mg/g SS, 113.6 mg/g SS, and 123.8 mg/g SS, respectively, and the amounts of acetic acid from 1 g of sludge were 24.5 mg/g SS, 65.5 mg/g SS, 88.1 mg/g SS, and 121.5 mg/g SS, respectively. This suggested that the formation of sludge to organic acids as well as the conversion of organic acids to acetic acid increased with reaction temperature. Based on the experimental results, a modified kinetic model was suggested for the liquefaction reaction of sludge and the formation of organic acids. The kinetic model predicted an increase in kinetic parameters $k_1$ (liquefaction of organic compounds), $k_2$ (formation of organic acids to intermediate), $k_3$ (final degradation of intermediate), and $k_4$ (final degradation of organic acids) with reaction temperature. This indicated that the liquefaction of organic solid materials and the formation of organic acids increase according to reaction temperature. The calculated activation energy for reaction kinetic constants were 20.7 kJ/mol, 12.3 kJ/mol, 28.4 kJ/mol, and 54.4 kJ/mol, respectively, leading to a conclusion that not thermal hydrolysis but oxidation reaction is the rate-limiting step.

• Journal of Korean Society of Environmental Engineers
• /
• v.22 no.4
• /
• pp.733-742
• /
• 2000

Fundamental materialistic characterization and adsorption/neutralization behavior of waste egg shell for heavy metal ion have been studied for its application to wastewater treatment. To investigate the structural change and thermal decomposition characteristics of egg shell. X-ray diffraction and FT-IR analysis were conducted for egg shell treated at $105^{\circ}C$ and $700^{\circ}C$, respectively. For the result of FT-IR analysis, the sample treated at $700^{\circ}C$ showed a reduced C-O absorption band compared with that of egg shell treated at $105^{\circ}C$, which may be due to the $CO_2$ release. Unlike to the result of FT-IR analysis, the XRD patterns of egg shell were almost similar for the cases of $105^{\circ}C$ and $700^{\circ}C$ treatment. however, characteristic diffraction pattern of CaO was observed for $850^{\circ}C$ treatment, at which $CaCO_3$ is known to be completely converted to CaO. TGA/DTA analysis showed a slow decline in weight loss up to $600^{\circ}C$ and, for $600{\sim}800^{\circ}C$ range, the weight loss became drastic by reason of $CO_2$ discharge, which was accompanied by an appearance of major endothermic peak. The ratio of practical breakthrough time to ideal one, total transfer unit, and mass transfer coefficient were observed to be increased as the adsorption was progressed in a multiple-column fixed-bed reactor using egg shell as an adsorbent, which signified the distribution effect of mass transfer for continuous adsorption reaction. The neutralization effect of egg shell for several types of acidic wastewater made of different mineral acids was not much different from each other except for the case of $H_2SO_4$, for which the neutralization reaction was thought to be retarded by the formation of gypsum.

• Korean Chemical Engineering Research
• /
• v.50 no.2
• /
• pp.251-256
• /
• 2012

Etherification of n-butanol to di-n-Butyl Ether was carried out over Keggin $H_{3+x}PW_{12-x}Nb_xO_{40}$ (x=0, 1, 2, 3) and $H_{6+x}P_2W_{18-x}Nb_xO_{62}$ (x=0, 1, 2, 3) Wells-Dawson heteropolyacid catalysts. Niobium-substituted Keggin and Wells-Dawson heteropolyacid catalysts with different niobium content were prepared. Successful preparation of the catalysts was confirmed by FT-IR, ICP-AES, and $^{31}P$ NMR analyses. Their acid properties were determined by $NH_3$-TPD (Temperature-Programmed Desorption) measurements. Heteropolyacid catalysts showed different acid properties depending on niobium content in both series. The correlation between acid properties of heteropolyacid catalysts and catalytic activity was then established. Acidity of Keggin and Wells-Dawson heteropolyacid catalysts decreased with increasing niobium content, and conversion of n-butanol and yield for di-n-butyl ether increased with increasing acidity of the catalysts, regardless of the identity of heteropolyacid catalysts (without heteropolyacid structural sensitivity). Thus, acidity of heteropolyacid catalysts served as an important factor determining the catalytic performance in the etherification of n-butanol to di-n-Butyl Ether.

• Journal of Bio-Environment Control
• /
• v.6 no.1
• /
• pp.1-14
• /
• 1997

This experiment was conducted to find out the compositions of nutrient solution for closed system in substrate culture of cucumber. Cucumber(Cucumis sativus L. cv. Eunsung baekdadagi) plants were grown in the substrates supplied with the nutrient solutions whose strengths were 1/2, 1, and 3/2 of the original concentration developed by National Horticultural Research Station in Japan. By increasing the nutrient concentrations, plant height decreased but leaf length, leaf width, and leaf number showed little differences. A number of marketable fruit and marketable yield were the highest in the concentration of 1 strength. The nutrient compositions of solution developed for closed system in cucumber substrate culture were N 11.4, P 3.3, K 6.0, Ca 4.5, and Mg 3.5 me.$\ell$$^{-1}$ during the vegetative growth period and N 10.4, P 3.3, K 5.0, Ca 4.5, and Mg 3,5 me.$\ell$$^{-1}$ during the reproductive growth period. To examine the suitability of nutrient solution developed in the above experiment, cucumber plants were grown in the substrates supplied with different solutions and concentrations - Yamasaki's nutrient solution(Yamasaki) of 1 S, nutrient solution of Research Station for Greenhouse Vegetable and Floriculture on the Netherlands(PTG) of 1 S, nutrient solution developed in the above experiment(SCU) of 1/2, 1, and 3/2 S. EC and pH in root zone changed little in the all treatments during growing period. As cucumber plants grew, the concentrations of N, P, and K in root zone decreased but Ca concentration increased. Net $CO_2$ assimilation rate of cucumber leaves was high in SCU of 1 and 3/2 S, and Yamasaki of 1 S. Growth of cucumber plants was the lowest in SCU of 1/2 S.