• Asia pacific journal of information systems
• /
• v.19 no.2
• /
• pp.157-178
• /
• 2009

Corporate credit rating is a very important factor in the market for corporate debt. Information concerning corporate operations is often disseminated to market participants through the changes in credit ratings that are published by professional rating agencies, such as Standard and Poor's (S&P) and Moody's Investor Service. Since these agencies generally require a large fee for the service, and the periodically provided ratings sometimes do not reflect the default risk of the company at the time, it may be advantageous for bond-market participants to be able to classify credit ratings before the agencies actually publish them. As a result, it is very important for companies (especially, financial companies) to develop a proper model of credit rating. From a technical perspective, the credit rating constitutes a typical, multiclass, classification problem because rating agencies generally have ten or more categories of ratings. For example, S&P's ratings range from AAA for the highest-quality bonds to D for the lowest-quality bonds. The professional rating agencies emphasize the importance of analysts' subjective judgments in the determination of credit ratings. However, in practice, a mathematical model that uses the financial variables of companies plays an important role in determining credit ratings, since it is convenient to apply and cost efficient. These financial variables include the ratios that represent a company's leverage status, liquidity status, and profitability status. Several statistical and artificial intelligence (AI) techniques have been applied as tools for predicting credit ratings. Among them, artificial neural networks are most prevalent in the area of finance because of their broad applicability to many business problems and their preeminent ability to adapt. However, artificial neural networks also have many defects, including the difficulty in determining the values of the control parameters and the number of processing elements in the layer as well as the risk of over-fitting. Of late, because of their robustness and high accuracy, support vector machines (SVMs) have become popular as a solution for problems with generating accurate prediction. An SVM's solution may be globally optimal because SVMs seek to minimize structural risk. On the other hand, artificial neural network models may tend to find locally optimal solutions because they seek to minimize empirical risk. In addition, no parameters need to be tuned in SVMs, barring the upper bound for non-separable cases in linear SVMs. Since SVMs were originally devised for binary classification, however they are not intrinsically geared for multiclass classifications as in credit ratings. Thus, researchers have tried to extend the original SVM to multiclass classification. Hitherto, a variety of techniques to extend standard SVMs to multiclass SVMs (MSVMs) has been proposed in the literature Only a few types of MSVM are, however, tested using prior studies that apply MSVMs to credit ratings studies. In this study, we examined six different techniques of MSVMs: (1) One-Against-One, (2) One-Against-AIL (3) DAGSVM, (4) ECOC, (5) Method of Weston and Watkins, and (6) Method of Crammer and Singer. In addition, we examined the prediction accuracy of some modified version of conventional MSVM techniques. To find the most appropriate technique of MSVMs for corporate bond rating, we applied all the techniques of MSVMs to a real-world case of credit rating in Korea. The best application is in corporate bond rating, which is the most frequently studied area of credit rating for specific debt issues or other financial obligations. For our study the research data were collected from National Information and Credit Evaluation, Inc., a major bond-rating company in Korea. The data set is comprised of the bond-ratings for the year 2002 and various financial variables for 1,295 companies from the manufacturing industry in Korea. We compared the results of these techniques with one another, and with those of traditional methods for credit ratings, such as multiple discriminant analysis (MDA), multinomial logistic regression (MLOGIT), and artificial neural networks (ANNs). As a result, we found that DAGSVM with an ordered list was the best approach for the prediction of bond rating. In addition, we found that the modified version of ECOC approach can yield higher prediction accuracy for the cases showing clear patterns.

• Asian Oncology Nursing
• /
• v.3 no.1
• /
• pp.24-44
• /
• 2003

The purposes of this study was to identify and propose the expected role of the oncology nurse specialist by embodying role theory to oncology nurse specialist. The subjects of this study were 149 persons in 14 hospitals, who were classified to 4 groups, oncology nurse specialists(ONS) group, head nurses and charge nurses(HN & CN) group in hemato-oncology ward, registered nurses(RN) group in hemato-oncology ward, and hematologists & oncologists(H&O) group. The questionnaire which was consisted of 89 items for role of oncology nurse specialist, was made by researcher with a field study and literature review about role of oncology nurse specialist and verified by matrix delphi technique about content validity and construct validity. The data were collected from October 22, 2002 to November 5, 2002. All 4 groups proposed that ONS should perform an expert practitioner role first of all. But ONS group, RN group and H&O group proposed orderly expert practitioner, educator, researcher, consultant, and administrator & change agent, but HN & CN group did expert practitioner, educator, consultant, researcher, administrator & change agent. Expert practitioner had the most highest necessary degree in all groups and most highest performance degree in ONS group. That was consistent with results that all groups proposed role of expert practitioner at first. 4 items out of 20 items showed the meaningful differences between groups. For role of educator, oncology nurse specialist group proposed necessary degrees over 4.0 point out of 5.0 in all items. 4 items out of 18 items showed the meaningful differences between groups. For role of researcher, 3 nurses groups proposed a high necessary degree, but performance of ONS group was most lowest among 5 roles. 6 items out of 14 items showed the meaningful differences between groups. The role of consultant had high necessary degree in some items related to hematopoietic stem cell transplantation. 2 items out of 17 items showed the meaningful differences between groups. In nursing behaviors of administrator & change agent, those items about enacting principle, cost development and participation of professional academy had a high necessary degree. 4 items out of 18 items showed the meaningful differences between groups. Oncology nurse specialists group performed 5 roles orderly, expert practitioner, consultant, educator, administrator & change agent, researcher. This result was different from expected role of themselves as well as the other groups. There was a different necessary degree between role and embodied nursing behaviors of role. ONS group and RN group proposed orderly educator, researcher, administrator & change agent, expert practitioner, consultant, but the other groups did educator, expert practitioner, researcher, consultant, administrator & change agent. The expected standards of oncology nurse specialist in this study were usually master's degree, total career of 5-7 years, oncology career of 3-5 years and certification. But for the post, qualification and qualification institution, various opinions were suggested. In the conclusion, there was a different necessary degree between role and embodied nursing behaviors of role. All groups proposed expert practitioner at first in abstract role, but educator at first in embodied nursing behaviors of role. So we have to consider this difference carefully in the future research. ONS acted the role of expert practitioner first of all, but we should develope and expand the roles of researcher, and administrator & change agent. We should enact roles by role behaviors induced from mutual agreements in necessary degree and performance degree, and bargain the role behaviors that showed the meaningful differences between groups But, we should consider carefully which group's opinion we have to select. I suggested 36 items out of 89 items, in which ONS proposed necessary degree over 4.0 out of 5.0 and half of them performed as the nursing behaviors of oncology nurse specialist that did not induce role stress. For the future, We should role bargain the role with other groups based on these items.

• Journal of Distribution Research
• /
• v.16 no.1
• /
• pp.117-140
• /
• 2011

The sales promotion strategy can be classified into two categories: cost-down and value-added promotions. Although many studies have been performed on sales promotion in the past, little attention has been given on cost-down promotion, and in particular on its strategy. Also there has been a renewed interest on value-added promotion strategy as means of attracting more customers and delivering value by providing a greater benefit. This study examines the effect of value-added promotion on the evaluation of promotional products. The topic is important because the value-added promotions are not always perceived as a benefit by consumers and they depend on the situational factors. We postulate three hypotheses on the basis of related literature. We conduct two studies: one employing experiment (study 1) and the other, quasi-experiment (study 2) to investigate the effect of two situational factors, namely the promotional package type and retailer uncertainty on the valuation of promotional products. Study 1 employs 2 promotional package types (bundled with freebie type vs. one-plus-one type) by 3 evaluation targets (overall package, a focal product and freebie) in experimental design. Also it is found that consumers devaluate the promotional product when the level of retailer uncertainty is high (t=-4.70, p=.000) as shown in Table 2. As depicted in Figure 2, the interaction effect of retailer uncertainty and package types on the evaluation of promotional product as a whole does not appear to be significant. However, when the level of retailer uncertainty is high, the focal product suffers from lower valuation if it is included in a bundle with freebie type package. The purpose of Study 2 is to cross-check the results of Study 1. The results of Study 2 also show that the consumers devaluate the promotional products as a whole when the level of retailer uncertainty is high. Furthermore, Study2 examines the consumers' willingness to purchase. The willingness to purchase of 2 different consumer groups divided by their reservation prices before and after exposure to a promotional stimuli shows no significant differences(t=1.911, p=.057). The results suggest that the consumers' reservation prices before the exposure to the promotion can be a reference prices of their's. But after the promotional events, the promotional price would become their reference prices. Furthermore, when the level of retailer uncertainty is high, consumers devaluate the promotional product as a whole. Because the promotional offerings activate the persuasion knowledge of consumers and make them incorporate negative inference about the firm's motives into their valuation and consequently, consumers tend to hesitate to purchase.

• Journal of Distribution Research
• /
• v.16 no.1
• /
• pp.65-93
• /
• 2011

This paper aims to explore how relational constructs(supplier's idiosyncratic investment, retailer's dependence, and dyadic communication) impact the buyer's trust across the relationship lifecycle. Companies make great efforts to create trust and maintain long-term relationship with buyers. Buyer and supplier relationships evolve through distinct lifecycles. In each phase, buyers and suppliers exhibit different patterns of behaviour. Also, they have different attitudes toward, and expectation of, one another. In this study, the final outcome investigated is long-term orientation from the buyer's perspective. Data were collected from 296 survey questionnaires of retailers engaging in Crops Protected Material industry in Korea. Exploratory factor analysis and confirmatory factor analysis were used to purify the measurement scales. And structural equation analysis and Chow test were used to test the hypotheses. The result indicates that the effect of retailer's dependence on the trust is greatest. Especially, the effect of dependence is growing as the relationship has developed. Supplier's idiosyncratic investment increases trust in the early phase, while it is not found to affect trust in the maturity phase. In the early phase, the determination of trust begins to be increasingly influenced by partner's relationship investment only because other clues are not observed yet. Reciprocal communication has a positive effect on trust, and long-term orientation is directly influenced by retailer's trust. Communication's effect is greater in the build-up phase than in the maturity phase. A limitation of this study is the unequal sample size in each of the relationship phases. Literature has suggested that the effects of relational constructs on outcomes are different across relationship phases. This study shows that different relational constructs have different effects on trust depending upon the relationship phase.

• Journal of Nutrition and Health
• /
• v.47 no.5
• /
• pp.351-363
• /
• 2014

Purpose: The aim of this study was to develop a reliable measurement tool to assess the urban food environment, particularly focusing on food stores in Korea. Methods: The items for the measurement tool were selected through systematic literature reviews and adjusted to the Korean food environment. A total of 25 food stores in an urban area were recruited for the pilot test to evaluate the time required for completion of the survey, ease of response, willingness to participate, difficulties in conducting the survey, and appropriateness, and 34 food stores were recruited for assessment of the reliability of the tool using percent agreement and kappa value. Results: The measurement tool is composed of questions regarding food store characteristics, accessibility, and food availability. On average, 26 minutes was required to survey a single food store, and the subjects and interviewers answered that the process was not difficult for the survey. The percent agreement for the inter-rater and test-retest reliability ranged from 93.9~98.8% and 91.9~97.9, respectively. The kappa values ranged from 0.78 to 0.97, which was very high. Percent agreement and kappa value of food with healthy option were lower than those of the general food in the inter-rater reliability (p = 0.0027, p = 0.0095 respectively) as well as in the test-retest reliability (p = 0.0081, p = 0.0290, respectively), although they were still high enough (86.4~98.0% for percent agreement, 0.64~0.96 for kappa value). Conclusion: The newly developed measurement tool for assessment of food store environment appears to be feasible and reliable; therefore, it can be applied to research on the association between food environment and dietary behaviors as well as health outcomes.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.35 no.4
• /
• pp.456-463
• /
• 2006

The purposes of this study were to extract choice attributes of bakery cafe customers and to conduct important- performance analysis (IPA) of choice attributes perceived by bakery cafe customers. The questionnaire was developed through literature review and focus group interview, and modified after pilot test. The questionnaires for main survey were distributed to 320 males and females aged 20 and over in Seoul. A total of 275 questionnaires were used for analysis (85.9%) and the statistical analyses were conducted using SPSS Win (12.0) for descriptive analyses, exploratory factor analysis, reliability analysis, and correlation analyses. The main results were as follows. 'Products', 'convenience to use', 'services and price', 'interior environments' 'brand' and 'location' dimensions were extracted as choice attributes dimensions of bakery cafe customers and customers of bakery cafe regarded 'sanitation and cleanness', 'kindness of employees', 'quality of products', 'comfortable and pleasant facilities' and 'taste of bakery products' as more important than other attributes. In addition, the results of IPA showed that marketing managers of bakery cafes should focused on the dimension of 'services and price' in the reason that this dimension was low at performance although customers regarded it very important. Overall, researchers and managers of bakery cafes should understand unique choice attributes of bakery cafe customers, and make efforts to establish marketing strategies that meet bakery cafe customers' needs.

• Korean Journal of Construction Engineering and Management
• /
• v.16 no.3
• /
• pp.36-48
• /
• 2015

To complete various types of finish works with higher quality in much less time, TACT, which was mostly used for high-rise buildings, has been adapted to meet the needs for systematic schedule management in construction sites. However, the effectiveness of adapting TACT has not been shown as expected due to the different perspectives on productivity from both general contractor and subcontractors based on unforeseen conditions according to the types of site. Furthermore, not enough theoretical backgrounds, empirical data, and systematic approaches to solve the fundamental problems caused by each participants' different views on productivity has produced obstacles for establishing effective solutions. Therefore, this research aims to analyze the possible main reasons for having different point of views regarding productivity among various participants of residential building sites using TACT based on literature review, site survey, and interviews. Also, case study was conducted to propose obtainable productivity (OP) regression equation and productivity achievement ratio (PAR) with reduction factors (RFs) and actual productivity (AP) data from an actual construction site. The proposed outcome may assist general contractors converting output management with PPC to productivity management with actual data using PAR. On the other hand, subcontractors would be able to estimate theory-based maximum productivity of construction sites with TACT by using OP. The PAR will enhance the communication between general and sub-contractors for their decision making process. Finally, the main RFs derived from PAR could be used as essential keys for productivity management to increase the economical and operational effectiveness of the construction project.

• Journal of Korean Society of Environmental Engineers
• /
• v.29 no.6
• /
• pp.647-653
• /
• 2007

The ultimate objective of this study is to develop a reliable oxygen-enriched combustion techniques especially for the case of the flue gas recycling in order to reduce the $CO_2$ emissions from practical industrial boilers. To this end a systematic numerical investigation has been performed, as a first step, for the resolution of the combusting flame characteristics of lab-scale LNG combustor. One of the important parameters considered in this study is the level of flue gas recycling calculated in oxygen enriched environment. As a summary of flame characteristics, for the condition of 100% pure $O_2$ as oxidizer without any flue gas recycling, the flame appears as long and thin laminar-like shape with relatively high flame temperature. The feature of high peak of flame temperature is explained by the absence of dilution and heat loss effects due to the presence of $N_2$ inert gas. The same reasoning is also applicable to the laminarized thin flame one, which is attributed to the decrease of the turbulent mixing. These results are physically acceptable and consistent and further generally in good agreement with experimental results appeared in open literature. As the level of $CO_2$ recycling increases in the mixture of $O_2/CO_2$, the peak flame temperature moves near the burner region due to the enhanced turbulent mixing by the increased amount of flow rate of oxidizer stream. However, as might be expected, the flue gas temperature decreases due to presence of $CO_2$ gas together with the inherent feature of large specific heat of this gas. If the recycling ratio more than 80%, gas temperatures drop so significantly that a steady combustion flame can no longer sustain within the furnace. However, combustion in the condition of 30% $O_2/70% $ $CO_2$ can produce similar gas temperature profiles to those of conventional combustion in air oxidizer. An indepth analyses have been made for the change of flame characteristics in the aspect of turbulent intensity and heat balance.

• Journal of Korean Society of Environmental Engineers
• /
• v.35 no.6
• /
• pp.422-429
• /
• 2013

It is generally known that the increase of the Earth surface temperature due to the global warming together with the land desertification by rapid urban development has caused severe climate and weather change. In desert or desertification land, it is observed that there are always severe flooding phenomena, even if desert sand has the high porosity, which could be believed as the favorable condition of rain water infiltration into ground water. The high runoff feature causes possibly another heavy rain by quick evaporation with the depletion of underground water due to the lack of infiltration. The basic physics of desert flooding is reasonably assumed due to the thermal buoyancy of the higher temperature of the soil temperature than that of the rain drop. Considering the importance of this topic associated with water resource management and climate disaster prevention, no systematic investigation has, however, been reported in literature. In this study, therefore, a laboratory scale experiment together with the effort of numerical calculation have been performed to evaluate quantitatively the basic hypothesis of run-off mechanism caused by the increase of soil temperature. To this end, first, of all, a series of experiment has been made repeatedly with the change of soil temperature with well-sorted coarse sand having porosity of 35% and particle diameter, 2.0 mm. In specific, in case 1, the ground surface temperature was kept at $15^{\circ}C$, while in case 2 that was high enough at $70^{\circ}C$. The temperature of $70^{\circ}C$ was tested as this try since the informal measured surface temperature of black sand in California's Coachella Valley up to at 191 deg. $^{\circ}F$ ($88^{\circ}C$). Based on the experimental study, it is observed that the amount of runoff at $70^{\circ}C$ was higher more than 5% compared to that at $15^{\circ}C$. Further, the relative amount of infiltration by the decrease of the surface temperature from 70 to $15^{\circ}C$ is about more than 30%. The result of numerical calculation performed was well agreed with the experimental data, that is, the increase of runoff in calculation as 4.6%. Doing this successfully, a basic but important research could be made in the near future for the more complex and advanced topic for this topic.

• Journal of Nutrition and Health
• /
• v.50 no.2
• /
• pp.142-157
• /
• 2017

Purpose: The purpose of this study was to develop a nutrition quotient for adolescents (NQ-A) to assess overall dietary quality and food behavior of Korean adolescents. Methods: Development of the NQ-A was undertaken in three steps: item generation, item reduction, and validation. Candidate items of the NQ-A checklist were selected based on literature reviews, results of the fifth Korea National Health and Nutrition Examination Survey data, dietary guidelines for Korean adolescents, expert in-depth interviews, and national nutrition policies and recommendations. A total of 213 middle and high school students participated in a one-day dietary record survey and responded to 41 items in the food behavior checklist. Pearson's correlation coefficients between the responses to the checklist items along with nutritional status of the adolescents were calculated. Item reduction was performed, and 24 items were selected for the nation-wide survey. A total of 1,547 adolescents from 17 cities completed the checklist questionnaire. Exploratory factor and confirmatory factor analyses were performed to develop a final NQ-A model. Results: Nineteen items were finalized as the checklist items for the NQ-A. Checklist items were composed of five factors (balance, diversity, moderation, environment, and practice). The five-factor structure accounted for 47.2% of the total variance. Standardized path coefficients were used as weights of the items. The NQ-A and five-factor scores were calculated based on the obtained weights of the questionnaire items. Conclusion: Nutrition Quotient for adolescents (NQ-A) would be a useful instrument for evaluating dietary quality and food behavior of Korean adolescents. Further research on NQ-A is needed to reflect changes in adolescent's food behavior and environment.