• Information Systems Review
• /
• v.16 no.3
• /
• pp.161-177
• /
• 2014

Corporate credit rating assessment consists of complicated processes in which various factors describing a company are taken into consideration. Such assessment is known to be very expensive since domain experts should be employed to assess the ratings. As a result, the data-driven corporate credit rating prediction using statistical and artificial intelligence (AI) techniques has received considerable attention from researchers and practitioners. In particular, statistical methods such as multiple discriminant analysis (MDA) and multinomial logistic regression analysis (MLOGIT), and AI methods including case-based reasoning (CBR), artificial neural network (ANN), and multiclass support vector machine (MSVM) have been applied to corporate credit rating.2) Among them, MSVM has recently become popular because of its robustness and high prediction accuracy. In this study, we propose a novel optimized MSVM model, and appy it to corporate credit rating prediction in order to enhance the accuracy. Our model, named 'GAMSVM (Genetic Algorithm-optimized Multiclass Support Vector Machine),' is designed to simultaneously optimize the kernel parameters and the feature subset selection. Prior studies like Lorena and de Carvalho (2008), and Chatterjee (2013) show that proper kernel parameters may improve the performance of MSVMs. Also, the results from the studies such as Shieh and Yang (2008) and Chatterjee (2013) imply that appropriate feature selection may lead to higher prediction accuracy. Based on these prior studies, we propose to apply GAMSVM to corporate credit rating prediction. As a tool for optimizing the kernel parameters and the feature subset selection, we suggest genetic algorithm (GA). GA is known as an efficient and effective search method that attempts to simulate the biological evolution phenomenon. By applying genetic operations such as selection, crossover, and mutation, it is designed to gradually improve the search results. Especially, mutation operator prevents GA from falling into the local optima, thus we can find the globally optimal or near-optimal solution using it. GA has popularly been applied to search optimal parameters or feature subset selections of AI techniques including MSVM. With these reasons, we also adopt GA as an optimization tool. To empirically validate the usefulness of GAMSVM, we applied it to a real-world case of credit rating in Korea. Our application is in bond rating, which is the most frequently studied area of credit rating for specific debt issues or other financial obligations. The experimental dataset was collected from a large credit rating company in South Korea. It contained 39 financial ratios of 1,295 companies in the manufacturing industry, and their credit ratings. Using various statistical methods including the one-way ANOVA and the stepwise MDA, we selected 14 financial ratios as the candidate independent variables. The dependent variable, i.e. credit rating, was labeled as four classes: 1(A1); 2(A2); 3(A3); 4(B and C). 80 percent of total data for each class was used for training, and remaining 20 percent was used for validation. And, to overcome small sample size, we applied five-fold cross validation to our dataset. In order to examine the competitiveness of the proposed model, we also experimented several comparative models including MDA, MLOGIT, CBR, ANN and MSVM. In case of MSVM, we adopted One-Against-One (OAO) and DAGSVM (Directed Acyclic Graph SVM) approaches because they are known to be the most accurate approaches among various MSVM approaches. GAMSVM was implemented using LIBSVM-an open-source software, and Evolver 5.5-a commercial software enables GA. Other comparative models were experimented using various statistical and AI packages such as SPSS for Windows, Neuroshell, and Microsoft Excel VBA (Visual Basic for Applications). Experimental results showed that the proposed model-GAMSVM-outperformed all the competitive models. In addition, the model was found to use less independent variables, but to show higher accuracy. In our experiments, five variables such as X7 (total debt), X9 (sales per employee), X13 (years after founded), X15 (accumulated earning to total asset), and X39 (the index related to the cash flows from operating activity) were found to be the most important factors in predicting the corporate credit ratings. However, the values of the finally selected kernel parameters were found to be almost same among the data subsets. To examine whether the predictive performance of GAMSVM was significantly greater than those of other models, we used the McNemar test. As a result, we found that GAMSVM was better than MDA, MLOGIT, CBR, and ANN at the 1% significance level, and better than OAO and DAGSVM at the 5% significance level.

• Asia pacific journal of information systems
• /
• v.20 no.3
• /
• pp.139-166
• /
• 2010

The supply chain management (SCM) systems have emerged as strong managerial tools for manufacturing firms in enhancing competitive strength. Despite of large investments in the SCM systems, many companies are not fully realizing the promised benefits from the systems. A review of literature on adoption, implementation and success factor of IOS (inter-organization systems), EDI (electronic data interchange) systems, shows that this issue has been examined from multiple theoretic perspectives. And many researchers have attempted to identify the factors which influence the success of system implementation. However, the existing studies have two drawbacks in revealing the determinants of systems implementation success. First, previous researches raise questions as to the appropriateness of research subjects selected. Most SCM systems are operating in the form of private industrial networks, where the participants of the systems consist of two distinct groups: focus companies and vendors. The focus companies are the primary actors in developing and operating the systems, while vendors are passive participants which are connected to the system in order to supply raw materials and parts to the focus companies. Under the circumstance, there are three ways in selecting the research subjects; focus companies only, vendors only, or two parties grouped together. It is hard to find researches that use the focus companies exclusively as the subjects probably due to the insufficient sample size for statistic analysis. Most researches have been conducted using the data collected from both groups. We argue that the SCM success factors cannot be correctly indentified in this case. The focus companies and the vendors are in different positions in many areas regarding the system implementation: firm size, managerial resources, bargaining power, organizational maturity, and etc. There are no obvious reasons to believe that the success factors of the two groups are identical. Grouping the two groups also raises questions on measuring the system success. The benefits from utilizing the systems may not be commonly distributed to the two groups. One group's benefits might be realized at the expenses of the other group considering the situation where vendors participating in SCM systems are under continuous pressures from the focus companies with respect to prices, quality, and delivery time. Therefore, by combining the system outcomes of both groups we cannot measure the system benefits obtained by each group correctly. Second, the measures of system success adopted in the previous researches have shortcoming in measuring the SCM success. User satisfaction, system utilization, and user attitudes toward the systems are most commonly used success measures in the existing studies. These measures have been developed as proxy variables in the studies of decision support systems (DSS) where the contribution of the systems to the organization performance is very difficult to measure. Unlike the DSS, the SCM systems have more specific goals, such as cost saving, inventory reduction, quality improvement, rapid time, and higher customer service. We maintain that more specific measures can be developed instead of proxy variables in order to measure the system benefits correctly. The purpose of this study is to find the determinants of SCM systems success in the perspective of vendor companies. In developing the research model, we have focused on selecting the success factors appropriate for the vendors through reviewing past researches and on developing more accurate success measures. The variables can be classified into following: technological, organizational, and environmental factors on the basis of TOE (Technology-Organization-Environment) framework. The model consists of three independent variables (competition intensity, top management support, and information system maturity), one mediating variable (collaboration), one moderating variable (government support), and a dependent variable (system success). The systems success measures have been developed to reflect the operational benefits of the SCM systems; improvement in planning and analysis capabilities, faster throughput, cost reduction, task integration, and improved product and customer service. The model has been validated using the survey data collected from 122 vendors participating in the SCM systems in Korea. To test for mediation, one should estimate the hierarchical regression analysis on the collaboration. And moderating effect analysis should estimate the moderated multiple regression, examines the effect of the government support. The result shows that information system maturity and top management support are the most important determinants of SCM system success. Supply chain technologies that standardize data formats and enhance information sharing may be adopted by supply chain leader organization because of the influence of focal company in the private industrial networks in order to streamline transactions and improve inter-organization communication. Specially, the need to develop and sustain an information system maturity will provide the focus and purpose to successfully overcome information system obstacles and resistance to innovation diffusion within the supply chain network organization. The support of top management will help focus efforts toward the realization of inter-organizational benefits and lend credibility to functional managers responsible for its implementation. The active involvement, vision, and direction of high level executives provide the impetus needed to sustain the implementation of SCM. The quality of collaboration relationships also is positively related to outcome variable. Collaboration variable is found to have a mediation effect between on influencing factors and implementation success. Higher levels of inter-organizational collaboration behaviors such as shared planning and flexibility in coordinating activities were found to be strongly linked to the vendors trust in the supply chain network. Government support moderates the effect of the IS maturity, competitive intensity, top management support on collaboration and implementation success of SCM. In general, the vendor companies face substantially greater risks in SCM implementation than the larger companies do because of severe constraints on financial and human resources and limited education on SCM systems. Besides resources, Vendors generally lack computer experience and do not have sufficient internal SCM expertise. For these reasons, government supports may establish requirements for firms doing business with the government or provide incentives to adopt, implementation SCM or practices. Government support provides significant improvements in implementation success of SCM when IS maturity, competitive intensity, top management support and collaboration are low. The environmental characteristic of competition intensity has no direct effect on vendor perspective of SCM system success. But, vendors facing above average competition intensity will have a greater need for changing technology. This suggests that companies trying to implement SCM systems should set up compatible supply chain networks and a high-quality collaboration relationship for implementation and performance.

• Journal of Korean Academy of Nursing
• /
• v.5 no.2
• /
• pp.74-91
• /
• 1975

A purpose of this study was to compare the breast massage and manual expression of the breast before engagement after delivery with the time of engagement, the throbbing pain in breast, the first amount of breast milk and involution of the uterus. The subjects selected for this study were 138 women (experimental group;69, control group :69) who were admitted to the Dept. of Obtest. and Gyneco. of Kyung Hee University Hospital from Jan. 5 to June 5, 1975. The results of study were as follows; 1 The average age of the women 26.9 years old in the experimental group and 27.6 years old in the control group. As to religion, the number of those who had no religion was 58.0 percent and 62.4 percent respectively. Classified according to occupation, there were 87.0 percent in house wives of the booths group. Educational background; 87.0 percent of high school graduates or above, 78.3 percent respectively. The occupation of husband 53.7 percent of company employees stood highest and 42.0 percent respectively. In according to the grade of wards, 55.1 percent and 52.2 percent of four-men room stood highest. 2. Physical condition: Body weight before this Pregnancy(T=0.4962, N.S.), the size of breast(X²df2 = 0.1728, N.S.), the shape of nipple(X²df3 =1.3804, N.S.), hemoglobin's level of the first day after delivery(T=1.2572, N.S.), the above were showed non significant between the experimental group and control group The investigator found any difference between the two groups of the health condition during the pregnancy, 3. The rate of no experience of breast massage during pregnancy was 85 percent and 75.4 percent (X²df1=2.2562, N.S.). 4. As to the meal during hospitalization after delivery: The booth of the groups in ordinary food took usually of meyer soup and milk(X²df8=2.5957, N.S.). 5. The relation between the first step of breast massage, second step of the manual expression of breast before engagement after delivery and time of engagement : average time of engagement in the experimental group (2.1 days±0.8) was shortened than the control group (3.3 days±1.2). (T=-6.9045, P< 0.005). It toot less time in the experimental group of primipara(2.2days±0.7) than in the control group (3.1day±1.2) and it also took less time in the experimental group of multipara (2.0 days±0.9) than in the control group (3.5days±1.4). (Primipara T=-3.9266, 0< 0.005. Multipara T= 5.2356, P<0.005). 6. The relationship between the first step of the massage and second step of manual expression and the throbbing pain at the time of engagement: The experimental group showed less effect than control group (X²df4= 27.3342 P<0.005). The separate study of primipara and multipara showed remarkable difference in the group of primipara)X²df4=20.7285, p<0.005) and little difference in the multipara group (X²df4=8.8351, p< 0. 10). 7. The relationship between the first step of the breast massage, second stop of the manual expression and first amount of breast milk: The average amount of breast milk increased more conspicuously in the experimental group (33.8㎖±23.4) than in the control(29.8㎖±25.3) (T=0.8262, N.S.). No remarkable difference was found in the respective groups that investigated in the groups of primipara and of multipara. (Primipara T=1.1467, N.S., Multipara T=-0.0354, N.S.). 8. The relationship between the first step of breast massage and second step of manual expression of breast and involution of uterus : Average time needed for uttering involution was sooner in the experimental group of primipasa(-3.3 F.B.±1.1), than the control group of primipara (-2.5F. B.±1.2), and it was sooner in the experimental group of muitipara (-3.0 F. B.±l..3), than the control group of multipara(-2.3 F.B±0.9). Primipara T=-2.9272, p< 0.005, Multipara T=2.5557, p< 0.01).

• Journal of Internet Computing and Services
• /
• v.16 no.3
• /
• pp.79-93
• /
• 2015

With the rapid development of information technology, the ways of usage have changed drastically. The ways and efficiency of traditional service application to data processing already could not satisfy the requirements of modern users. Nowadays, users have already understood the importance of data. Therefore, the processing and saving of big data have become the main research of the Internet service company. In China, with the rise and explosion of 115 Cloud leads to other technology companies have began to join the battle of cloud services market. Although currently Chinese cloud services are still mainly dominated by cloud storage service, the series of service contents based on cloud storage service have been affirmed by users, and users willing to try these new ways of services. Thus, how to let users to keep using cloud services has become a topic that worth for exploring and researching. The academia often uses the TAM model with statistical analysis to analyze and check the attitude of users in using the system. However, the basic TAM model obviously already could not satisfy the increasing scale of system. Therefore, the appropriate expansion and adjustment to the TAM model (i. e. TAM2 or TAM3) are very necessary. This study has used the status of Chinese internet users and the related researches in other areas in order to expand and improve the TAM model by adding the brand influence, hardware environment and external environments to fulfill the purpose of this study. Based on the research model, the questionnaires were developed and online survey was conducted targeting the cloud services users of four Chinese main cities. Data were obtained from 210 respondents were used for analysis to validate the research model. The analysis results show that the external factors which are service contents, and brand influence have a positive influence to perceived usefulness and perceived ease of use. However, the external factor hardware environment only has a positive influence to the factor of perceived ease of use. Furthermore, the perceived security factor that is influenced by brand influence has a positive influence persistent intention to use. Persistent intention to use also was influenced by the perceived usefulness and persistent intention to use was influenced by the perceived ease of use. Finally, this research analyzed external variables' attributes using other perspective and tried to explain the attributes. It presents Chinese cloud service users are more interested in fundamental cloud services than extended services. In private cloud services, both of increased user size and cooperation among companies are important in the study. This study presents useful opinions for the purpose of strengthening attitude for private cloud service users can use this service persistently. Overall, it can be summarized by considering the all three external factors could make Chinese users keep using the personal could services. In addition, the results of this study can provide strong references to technology companies including cloud service provider, internet service provider, and smart phone service provider which are main clients are Chinese users.

• Journal of Intelligence and Information Systems
• /
• v.21 no.4
• /
• pp.93-110
• /
• 2015

The recommender system is a system which recommends products to the customers who are likely to be interested in. Based on automated information filtering technology, various recommender systems have been developed. Collaborative filtering (CF), one of the most successful recommendation algorithms, has been applied in a number of different domains such as recommending Web pages, books, movies, music and products. But, it has been known that CF has a critical shortcoming. CF finds neighbors whose preferences are like those of the target customer and recommends products those customers have most liked. Thus, CF works properly only when there's a sufficient number of ratings on common product from customers. When there's a shortage of customer ratings, CF makes the formation of a neighborhood inaccurate, thereby resulting in poor recommendations. To improve the performance of CF based recommender systems, most of the related studies have been focused on the development of novel algorithms under the assumption of using a single profile, which is created from user's rating information for items, purchase transactions, or Web access logs. With the advent of big data, companies got to collect more data and to use a variety of information with big size. So, many companies recognize it very importantly to utilize big data because it makes companies to improve their competitiveness and to create new value. In particular, on the rise is the issue of utilizing personal big data in the recommender system. It is why personal big data facilitate more accurate identification of the preferences or behaviors of users. The proposed recommendation methodology is as follows: First, multimodal user profiles are created from personal big data in order to grasp the preferences and behavior of users from various viewpoints. We derive five user profiles based on the personal information such as rating, site preference, demographic, Internet usage, and topic in text. Next, the similarity between users is calculated based on the profiles and then neighbors of users are found from the results. One of three ensemble approaches is applied to calculate the similarity. Each ensemble approach uses the similarity of combined profile, the average similarity of each profile, and the weighted average similarity of each profile, respectively. Finally, the products that people among the neighborhood prefer most to are recommended to the target users. For the experiments, we used the demographic data and a very large volume of Web log transaction for 5,000 panel users of a company that is specialized to analyzing ranks of Web sites. R and SAS E-miner was used to implement the proposed recommender system and to conduct the topic analysis using the keyword search, respectively. To evaluate the recommendation performance, we used 60% of data for training and 40% of data for test. The 5-fold cross validation was also conducted to enhance the reliability of our experiments. A widely used combination metric called F1 metric that gives equal weight to both recall and precision was employed for our evaluation. As the results of evaluation, the proposed methodology achieved the significant improvement over the single profile based CF algorithm. In particular, the ensemble approach using weighted average similarity shows the highest performance. That is, the rate of improvement in F1 is 16.9 percent for the ensemble approach using weighted average similarity and 8.1 percent for the ensemble approach using average similarity of each profile. From these results, we conclude that the multimodal profile ensemble approach is a viable solution to the problems encountered when there's a shortage of customer ratings. This study has significance in suggesting what kind of information could we use to create profile in the environment of big data and how could we combine and utilize them effectively. However, our methodology should be further studied to consider for its real-world application. We need to compare the differences in recommendation accuracy by applying the proposed method to different recommendation algorithms and then to identify which combination of them would show the best performance.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.32 no.12
• /
• pp.1903-1914
• /
• 2008

The purpose of this study was to propose an efficient and feasible global market entry strategy for the Korean apparel industry by analyzing the Italian fashion industry. In particular, the study investigated the role of foreign exhibitions and showrooms supported and organized by Italian fashion organizations. The methodology for this study was to analyze industrial reports, review previous studies and conduct in-depth interviews with 23 industry experts in Italy, Korea and LA. The results indicated that the most prominent factor in the Italian fashion industry was the fashion cluster, which is a strong and organic network of diverse fashion related areas No matter the size of the enterprise, firms can get practical, prompt and efficient support from diverse associations. The network operated by the associations provides strong support to each firm by organizing collections and exhibitions, and providing promotional activities. Showrooms and agents are another supportive "gate keeper", directly related to an enterprise's sales. However, Korean fashion firms did not have enough information or knowledge for foreign exhibitions, nor did they make aggressive promotional efforts in the global market. Despite the many fashion-related associations exist in Korea, their programs are too focused on visible accomplishments and are too oriented on "big company" and "big voice", rather than many "small firms". In conclusion, the Korean fashion industry-particularly the fashion industry in Seoul-has strong potential to become the center of the global fashion market in the future. However, the fashion support system that can act as the channel to promote firms and to meet global buyers needs to be supplemented. To feasibly create this system, government or industry associations should develop a strong and generous support system and network, and they must recognize the need for small firms to exist.

• Asia pacific journal of information systems
• /
• v.24 no.2
• /
• pp.233-253
• /
• 2014

Can the stock market really be predicted? Stock market prediction has attracted much attention from many fields including business, economics, statistics, and mathematics. Early research on stock market prediction was based on random walk theory (RWT) and the efficient market hypothesis (EMH). According to the EMH, stock market are largely driven by new information rather than present and past prices. Since it is unpredictable, stock market will follow a random walk. Even though these theories, Schumaker [2010] asserted that people keep trying to predict the stock market by using artificial intelligence, statistical estimates, and mathematical models. Mathematical approaches include Percolation Methods, Log-Periodic Oscillations and Wavelet Transforms to model future prices. Examples of artificial intelligence approaches that deals with optimization and machine learning are Genetic Algorithms, Support Vector Machines (SVM) and Neural Networks. Statistical approaches typically predicts the future by using past stock market data. Recently, financial engineers have started to predict the stock prices movement pattern by using the SNS data. SNS is the place where peoples opinions and ideas are freely flow and affect others' beliefs on certain things. Through word-of-mouth in SNS, people share product usage experiences, subjective feelings, and commonly accompanying sentiment or mood with others. An increasing number of empirical analyses of sentiment and mood are based on textual collections of public user generated data on the web. The Opinion mining is one domain of the data mining fields extracting public opinions exposed in SNS by utilizing data mining. There have been many studies on the issues of opinion mining from Web sources such as product reviews, forum posts and blogs. In relation to this literatures, we are trying to understand the effects of SNS exposures of firms on stock prices in Korea. Similarly to Bollen et al. [2011], we empirically analyze the impact of SNS exposures on stock return rates. We use Social Metrics by Daum Soft, an SNS big data analysis company in Korea. Social Metrics provides trends and public opinions in Twitter and blogs by using natural language process and analysis tools. It collects the sentences circulated in the Twitter in real time, and breaks down these sentences into the word units and then extracts keywords. In this study, we classify firms' exposures in SNS into two groups: positive and negative. To test the correlation and causation relationship between SNS exposures and stock price returns, we first collect 252 firms' stock prices and KRX100 index in the Korea Stock Exchange (KRX) from May 25, 2012 to September 1, 2012. We also gather the public attitudes (positive, negative) about these firms from Social Metrics over the same period of time. We conduct regression analysis between stock prices and the number of SNS exposures. Having checked the correlation between the two variables, we perform Granger causality test to see the causation direction between the two variables. The research result is that the number of total SNS exposures is positively related with stock market returns. The number of positive mentions of has also positive relationship with stock market returns. Contrarily, the number of negative mentions has negative relationship with stock market returns, but this relationship is statistically not significant. This means that the impact of positive mentions is statistically bigger than the impact of negative mentions. We also investigate whether the impacts are moderated by industry type and firm's size. We find that the SNS exposures impacts are bigger for IT firms than for non-IT firms, and bigger for small sized firms than for large sized firms. The results of Granger causality test shows change of stock price return is caused by SNS exposures, while the causation of the other way round is not significant. Therefore the correlation relationship between SNS exposures and stock prices has uni-direction causality. The more a firm is exposed in SNS, the more is the stock price likely to increase, while stock price changes may not cause more SNS mentions.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.8 no.2
• /
• pp.242-253
• /
• 1998

South Korea has been producing asbestos over 60 years. The use of asbestos was over 50 years for production of asbestos slate and 27 years for asbestos friction materials including asbestos textile and brake-lining. Thus, it can be supposed that asbestos related diseases such as asbestosis, lung cancer and mesothelioma could be found in the vulnerable workers exposed to asbestos in 1955-1975, given the average latency period of 10-30 years. Asbestos was produced primarily by Japanese during World War II In Korea. The production of chrysotile peaked to 4,815 tons in 1944. From 1978 to 1984, 10,000 tons of asbestos were produced annually. However, the production was interrupted by raising labor costs and extinction of mine reserves, and finally they had to depend on import for the need of asbestos. In 1945, there were 16 asbestos mines, in total, with the addition of new asbestos mines in South Korea. Imports of asbestos was increased from 74,000 tons to 95,000 tons during the period of 1976 - 1992. But the imports was reduced to 88,000 tons in 1995. Since, in addition to the import of asbestos itself, the imports of asbestos products were increased as well and the accumulation of asbestos reached to 30,000 tons during the period of 1964 to 1993. In 1965, there was only one asbestos company with 207 employees. But the size of asbestos industry has been expanded so much that 118 asbestos companies could be found in 1993 with 1,476 workers. However, there was no record on the survey of asbestos concentration to which workers were exposed in any companies in 1983. The record of the air-borne concentration of the asbestos in textile working places in 1984 showed 6.7 fibers/cc by geometric mean(GM), but it was reduced to 1.2 fibers/cc in 1993. GMs of asbestos in working places for construction materials and asbestos textiles were also decreased from 1.7 fibers/cc to 0.55 fibers/cc during the period of 1984 - 1996.

• Journal of the Korean Applied Science and Technology
• /
• v.34 no.3
• /
• pp.545-554
• /
• 2017

This study is concerned with the stability of liquid crystal forming emulsifier with localized depend on change of pH using liquid crystal forming agent of advanced company. The liquid crystal emulsifying agent was localized using Sugar Crystal-LC (bio-tech Co., Ltd., Korea), and comparative samples were measured by using Nikkomulese-LC (Nikko Camicarls, Japan) and Alacel-LC (Croda Camicarls, UK). Liquid crystal formation was confirmed microscopically to show the formation of liquid crystal structure at acidic (pH=4.2), neutral (pH=7.0) and alkaline (pH=11.7). The particles of the liquid crystal were observed with a polarizing microscope according to the stirring speed. The stirring time was all the same for 3 minutes with a homo-mixer, and the stirring speed was increased to 2500 rpm, 3500 rpm and 4500 rpm to observe the liquid crystal state. As a result, it was found that the Korean surfactant was more stable and clear liquid crystal structure was formed than the two foreign acids. In the case of the UK in acid zone, the emulsion particle size was uniform and unstable. In the case of Japanese surfactant, it has similar structure and performance to those of localized Korean. It was found that Korean surfactant had superior emulsifying performance in acid zone compared with foreign products. It is possible to develop various formulations such as liquid crystal cream, lotion, eye cream, etc. using Sugar Crystal-LC emulsifier as an application cosmetic field, and it is expected that it can be widely applied as emulsifying technology for skin care external application in the pharmaceutical industry and the pharmaceutical industry as well as the cosmetics industry.

• Journal of Intelligence and Information Systems
• /
• v.17 no.4
• /
• pp.157-173
• /
• 2011

As the Internet use explodes recently, the malicious attacks and hacking for a system connected to network occur frequently. This means the fatal damage can be caused by these intrusions in the government agency, public office, and company operating various systems. For such reasons, there are growing interests and demand about the intrusion detection systems (IDS)-the security systems for detecting, identifying and responding to unauthorized or abnormal activities appropriately. The intrusion detection models that have been applied in conventional IDS are generally designed by modeling the experts' implicit knowledge on the network intrusions or the hackers' abnormal behaviors. These kinds of intrusion detection models perform well under the normal situations. However, they show poor performance when they meet a new or unknown pattern of the network attacks. For this reason, several recent studies try to adopt various artificial intelligence techniques, which can proactively respond to the unknown threats. Especially, artificial neural networks (ANNs) have popularly been applied in the prior studies because of its superior prediction accuracy. However, ANNs have some intrinsic limitations such as the risk of overfitting, the requirement of the large sample size, and the lack of understanding the prediction process (i.e. black box theory). As a result, the most recent studies on IDS have started to adopt support vector machine (SVM), the classification technique that is more stable and powerful compared to ANNs. SVM is known as a relatively high predictive power and generalization capability. Under this background, this study proposes a novel intelligent intrusion detection model that uses SVM as the classification model in order to improve the predictive ability of IDS. Also, our model is designed to consider the asymmetric error cost by optimizing the classification threshold. Generally, there are two common forms of errors in intrusion detection. The first error type is the False-Positive Error (FPE). In the case of FPE, the wrong judgment on it may result in the unnecessary fixation. The second error type is the False-Negative Error (FNE) that mainly misjudges the malware of the program as normal. Compared to FPE, FNE is more fatal. Thus, when considering total cost of misclassification in IDS, it is more reasonable to assign heavier weights on FNE rather than FPE. Therefore, we designed our proposed intrusion detection model to optimize the classification threshold in order to minimize the total misclassification cost. In this case, conventional SVM cannot be applied because it is designed to generate discrete output (i.e. a class). To resolve this problem, we used the revised SVM technique proposed by Platt(2000), which is able to generate the probability estimate. To validate the practical applicability of our model, we applied it to the real-world dataset for network intrusion detection. The experimental dataset was collected from the IDS sensor of an official institution in Korea from January to June 2010. We collected 15,000 log data in total, and selected 1,000 samples from them by using random sampling method. In addition, the SVM model was compared with the logistic regression (LOGIT), decision trees (DT), and ANN to confirm the superiority of the proposed model. LOGIT and DT was experimented using PASW Statistics v18.0, and ANN was experimented using Neuroshell 4.0. For SVM, LIBSVM v2.90-a freeware for training SVM classifier-was used. Empirical results showed that our proposed model based on SVM outperformed all the other comparative models in detecting network intrusions from the accuracy perspective. They also showed that our model reduced the total misclassification cost compared to the ANN-based intrusion detection model. As a result, it is expected that the intrusion detection model proposed in this paper would not only enhance the performance of IDS, but also lead to better management of FNE.