• Journal of the Korea Society for Simulation
• /
• v.15 no.3
• /
• pp.23-30
• /
• 2006

A recommender system for E-commerce site receives information from customers about which products they are interested in, and recommends products that are likely to fit their needs. In this paper, we investigate several methods for large-scale product purchase data for the purpose of producing useful recommendations to customers. We apply the traditional data mining techniques of cluster analysis and collaborative filtering(CF), and CF with reduction of product-dimensionality by use of latent semantic indexing(LSI). If reduced product-dimensionality obtained from LSI shows a similar latent trend of customers for buying products to that based on original customer-product purchase data, we expect less computational effort for obtaining the nearest-neighbor for target customer may improve the efficiency of recommendation performance. From simulation experiments on synthetic customer-product purchase data, CF-based method with reduction of product-dimensionality presents a better performance than the traditional CF methods with respect to the recall, precision and F1 measure. In general, the recommendation quality increases as the size of the neighborhood increases. However, our simulation results shows that, after a certain point, the improvement gain diminish. Also we find, as a number of products of recommendation increases, the precision becomes worse, but the improvement gain of recall is relatively small after a certain point. We consider these informations may be useful in applying recommender system.

• Journal of Global Scholars of Marketing Science
• /
• v.19 no.4
• /
• pp.57-63
• /
• 2009

With an exponential increase in electronic commerce (e-commerce), marketers are attempting to gain a competitive advantage by emphasizing service quality and post interaction service aspects, which leads to customer satisfaction or behavioral consequence. Particularly for apparel, service quality is one of the key determinants in encouraging customer e-loyalty, and hence the success of apparel retailing in the context of electronic commerce. Therefore, this study explores e-service quality (E-SERVQUAL) factors and their unique effects on e-loyalty for apparel online shopping based on Parasuraman et al' s (2005) framework. Specific objectives of this study are to identify underlying dimension of E-SERVQUAL, and analyze a structural model for examining the effect of E-SERVQUAL on e-loyalty for online apparel shopping. For the theoretical framework of service quality in the context of online shopping, literatures on traditional and electronic service quality factors were comparatively reviewed, and two aspects of core and recovery services were identified. This study hypothesized that E-SERVQUAL has an effect on e-loyalty; customer satisfaction has a positive effect on e-service loyalty for apparel online shopping; and customer satisfaction mediates in the effect of E-SERVQUAL on e-loyalty for apparel online shopping. A self-administered questionnaire was developed based on literatures. A total of 252 usable questionnaires were obtained from online consumers who had purchase experience with online shopping for apparel products and reside in standard metropolitan areas, in the United States. Factor analysis (e.g., exploratory, confirmatory) was conducted to assess the validity and reliability and the structural equation model including measurement and structural models was estimated via LISREL 8.8 program. Findings showed that the E-SERVQUAL of shopping websites for apparel consisted of five factors: Compensation, Fulfillment, Efficiency, System Availability, and Responsiveness. This supports Parasuraman (2005)'s E-S-QUAL encompassing two aspects of core service (e.g., fulfillment, efficiency, system availability) and recovery related service (e.g., compensation, responsiveness) in the context of apparel shopping online. In the structural equation model, there are five exogenous latent variables for e-SERVQUAL factors; and two endogenous latent variables (e.g., customer satisfaction, e-loyalty). For the measurement model, the factor loadings for each respective construct were statistically significant and were greater than .60 and internal consistency reliabilities ranged from .85 to .88. In the estimated structural model of the e-SERVEQUAL factors, the system availability was found to have direct and positive effect on e-loyalty, whereas efficiency had a negative effect on e-loyalty for apparel online shopping. However, fulfillment was not a significant predictor for explaining consequences of E-SERVQUAL for apparel online shopping. This finding implies that perceived service quality of system available was likely to increase customer satisfaction for apparel online shopping. However, it was not supported that e-loyalty was determined by service quality, because service quality has an indirect effect on e-loyalty (i.e., repurchase intention) by mediating effect of value or satisfaction in the context of online shopping for apparel. In addition, both compensation and responsiveness were found to have a significant impact on customer satisfaction, which influenced e-loyalty for apparel online shopping. Thus, there was significant indirect effect of compensation and responsiveness on e-loyalty. This suggests that the recovery-specific service factors play an important role in maximizing customer satisfaction levels and then maintaining customer loyalty to the online shopping site for apparel. The findings have both managerial and research implications. Fashion marketers can establish long-term relationship with their customers based on continuously measuring customer perceptions for recovery-related service quality, such as quick responses to problem and returns, and compensation for customers' problem after their purchases. In order to maintain e-loyalty, recovery services play an important role in the first choice websites for consumers to purchase clothing. Given that online consumers may shop anywhere, a marketing strategy for improving competitive advantages is to provide better service quality, maximize satisfaction, and turn to creating customers' e-loyalty for apparel online shopping. From a researcher's perspective, there are some limitations of this research that should be considered when interpreting its findings. For future research, findings provide a basis for the further study of this important topic along both theoretical and empirical dimensions. Based on the findings, more comprehensive models for predicting E-SERVQUAL's consequences can be developed and tested. For global fashion marketing, this study can expand to a cross-cultural approach into e-service quality for apparel by including multinational samples.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.34 no.5
• /
• pp.491-498
• /
• 2010

Humidification of PEM fuel cells is necessary for enhancing their performance and lifetime. In this study, a humidification system was designed and tested; the system includes an air-supply tube (inner diameter: 75 mm) through which a nozzle can be directly inserted and a cyclonic separator for the removal of water droplets. Three types of nozzles were employed to study the influence of injection pressure, air flow rate, and spray direction on the humidification performance. To evaluate the humidification performance, the concept of humidification efficiency was defined. In the absence of an external heat source, latent heat for evaporation will be supplied by the own enthalpies of water and air. Thus, the amount of water sprayed from the nozzle is the most critical factor affecting the humidification efficiency. Water droplets were efficiently removed by a cyclonic separator, but re-entrainment occurred at high air flow rates. The absolute humidity and humidification efficiency were $21.29\;kJ/kg_{da}$ and 86.57%, respectively, under the following conditions: nozzle type PJ24; spray direction angle $90^{\circ}$; injection pressure 1200 kPa; air flow rate 6000 Nlpm.

• Journal of Hydrogen and New Energy
• /
• v.31 no.1
• /
• pp.33-40
• /
• 2020

As hydrogen utilization becomes more active recently, a large amount of hydrogen should be supplied safely. Among the three supply methods, liquefied hydrogen, which is an optimal method of storage and transportation convenience and high safety, has a low temperature of -253℃, which is complicated by the liquefaction process and consumes a lot of electricity, resulting in high operating costs. In order to reduce the electrical energy required for liquefaction and to raise the efficiency, hydrogen is cooled by using a mixed refrigerant in a precooling step. The electricity required for the precooling process of the mixed refrigerant can be reduced by using the cold energy of LNG. Actually, LNG cold energy is used in refrigeration warehouse and air liquefaction separation process, and a lot of power reduction is achieved. The purpose of this study is to replace the electric power by using LNG cold energy instead of the electric air-cooler to lower the temperature of the hydrogen and refrigerant that are increased due to the compression in the hydrogen liquefaction process. The required energy was obtained by simulating mixed refrigerant (MR) hydrogen liquefaction system with LNG cold heat and electric system. In addition, the power replacement rate of the electric process were obtained with the pressure, the temperature of LNG, the rate of latent heat utilization, and the hydrogen liquefaction capacity, Therefore, optimization of the hydrogen liquefaction system using LNG cold energy was carried out.

• Journal of Biosystems Engineering
• /
• v.34 no.5
• /
• pp.351-357
• /
• 2009

This study was conducted to verify the simulation model through the drying test, and investigate effect of factors, such as temperature of drying air, airflow rate, and velocity of the airflow, on the drying. The low temperature drying simulation model was developed based on the circulation dry simulation model presented by Keum et al. (1987), and by modifying low temperature thin layer drying model, equilibrium moisture content model, latent heat of vaporization model, and crack ratio prediction model. The heat pump and experimental dryer with a capacity of 150kg were used for the test. The RMSE between the predicted and measured value was 0.27% (drying temperature), 0.15% (crack ratio), and 2.08% (relative humidity), so the relevance of the model was verified. In addition, the effect of drying temperature, airflow rate, and velocity of the airflow on the drying was examined. The experimental results showed that the crack ratio at drying temperature of $25{\sim}40^{\circ}C$ was allowable. Moreover, at below $30^{\circ}C$, variation of the crack ratio was slight, but drying time was delayed. Given these results, the drying temperature of over $30^{\circ}C$ was effective. As the airflow rate increased, required energy dramatically increased. Whereas drying rate slowly increased, so loss of drying efficiency was caused. Considering these results, the dryer needed to be designed and adjusted to lower than $30\;m^3/min{\cdot}ton$. As velocity of the airflow increased, required drying energy increased when the velocity of the airflow was over $5\;m^3$/hr, while crack ratio and drying rate showed little variation.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 2000.11c
• /
• pp.639-646
• /
• 2000

Hot air heater with light oil combustion is the most common heater for greenhouse heating in the winter season in Korea. However, since the heat efficiency of the heater is about 80%, considerable unused heat in the form of exhaust gas heat discharges to atmosphere. In order to capture this exhaust gas heat a heat recovery system for plant bed heating in the greenhouse was built and tested in the hot air heating system of greenhouse. The system consists of a heat exchanger made of copper pipes, ${\phi}\;12.7{\times}0.7t$ located inside the rectangular column of $330{\times}330{\times}900mm$, a water circulation pump, circulation plastic pipe and a water tame The total heat exchanger area is $1.5m^2$, calculated considering the heat exchange amount between flue gas and water circulated in the copper pipes. The system was attached to the exhaust gas path. The heat recovery system was designed as to even recapture the latent heat of flue gas when exposing to low temperature water in the heat exchanger. According to performance test it can recover 45,200 to 51,000kJ/hr depending on the water circulation rates of 330 to $690{\ell}$/hr from the waste heat discharged. The exhaust gas temperature left from the heat exchanger dropped to $100^{circ}C$ from $270^{circ}C$ by the heat exchange between the water and the flue gas, while water gained the difference and temperature increased to $38^{circ}C$ from $21^{circ}C$ at the water flow rate of $690{\ell}$/hr. And, the condensed water amount varies from 16 to $43m{\ell}$ at the same water circulation rates. This condensing heat recovery system can reduce boiler fuel consumption amount in a day by 34% according to the feasibility study of the actual mimitomato greenhouse. No combustion load was observed in the hot air heater.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.2
• /
• pp.578-584
• /
• 2013

Ice slurry that is a mixture of fine ice crystals and liquid water is a widely used working fluid in the ice thermal energy storage system due to its flowability and large latent heat of fusion. Generally ice slurry is made from supercooled water. But the excessive supercooling causes the water to freeze even worse to block the pipe. Additionally large degree of supercooling of water degrades the efficiency of the ice thermal energy storage system. Therefore the effective method to control the phase change from supercooled water to ice slurry is needed. In this paper we experimentally studied a novel method to generate the ice slurry from the supercooled water using the ultrasonic vibration. It was found that the cavitation impact of supercooled water by ultrasonic vibration can help the generation of ice slurry.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.37 no.5
• /
• pp.503-511
• /
• 2013

The humidifying supply gas is important in terms of the performance efficiency and membrane life improvement of a PEM fuel cell. A planar membrane humidifier is classified as a cross-flow and counter-flow type depending on the flow direction, and heat and mass transfer occur between the plate and the membrane. In this study, the changes in heat and mass transfer for various inlet temperatures and flow rates are compared according to the flow direction by using the sensible and latent ${\varepsilon}$-NTU method. The obtained results indicate that the counter flow shows higher heat and mass transfer performance than the cross flow at a low flow rate, and the difference in performance decreases as the flow rate increases. Furthermore, changes in the mass transfer performance decrease considerably with a nonlinear increase in the inlet temperature, and variations of the heat transfer performance are small.

• The Korean Journal of Applied Statistics
• /
• v.24 no.5
• /
• pp.951-961
• /
• 2011

In this paper, we propose a Bayesian inference using the Markov Chain Monte Carlo(MCMC) method for the zero inflated negative binomial(ZINB) regression model. The proposed model allows the regression model for zero inflation probability as well as the regression model for the mean of the dependent variable. This extends the work of Jang et al. (2010) to the fully defiend ZINB regression model. In addition, we apply the proposed method to a real data example, and compare the efficiency with the zero inflated Poisson model using the DIC. Since the DIC of the ZINB is smaller than that of the ZIP, the ZINB model shows superior performance over the ZIP model in zero inflated count data with overdispersion.

• Journal of the Korea Society of Computer and Information
• /
• v.10 no.5 s.37
• /
• pp.343-352
• /
• 2005

The existing CAI or web-based science learning system of elementary school has some disadvantages. For instance, it is composed of uniform courses designed by an instructor without considering the learner's characters, and the learner's opinions or questions raised during learning can not be delivered to the system. This structure has diminished the willingness or the motive of the learner and make an adverse effect on the learning efficiency. In this regards, Instruction-Learning System is needed to provide learning environment Pertinent to the learner's individual character and motivate the learner's active attendance and learning. This study is to design a character-based conversational Instruction-Learning System. This may induce the learner's active attendance through the communications between instructor and learner and furnish various learning materials to motivate the learners and attract their consistent interests in learning.