So Young Sohn

Support vector machines for default prediction of SMEs based on technology credit

In Korea, many forms of credit guarantees have been issued to fund small and medium enterprises (SMEs) with a high degree of growth potential in technology. However, a high default rate among funded SMEs has been reported. In order to effectively manage such governmental funds, it is important to develop an accurate scoring model for selecting promising SMEs. This paper provides a support vector machines (SVM) model to predict the default of funded SMEs, considering various input variables such as financial ratios, economic indicators, and technology evaluation factors. The results show that the accuracy performance of the SVM model is better than that of back-propagation neural networks (BPNs) and logistic regression. It is expected that the proposed model can be applied to a wide range of technology evaluation and loan or investment decisions for technology-based SMEs.

Structural equation model for predicting technology commercialization success index (TCSI)

Abstract Expecting high return, many firms try to invest on R&D of new technology. However, critical loss of assets would occur, when a firm fails to commercialize the developed technology. It would be of interest to provide the ideal environment for commercialization from the R&D stage. In this study, we use a structural equation model (SEM) to forecast the technology commercialization success index (TCSI) in relation to technology developer, technology receiver, technology transfer center, and environmental factors. The proposed SEM is fitted based on partial least square (PLS) estimation procedure. Individual TCSI is then found following the approach used for American customer satisfaction index (ACSI) for various combinations of characteristics of the type of technology, technology receiver, and technology developer. We expect that the proposed approach for TCSI can be used as guidance for an ideal match of technology with technology developer and technology receiver.

Segmentation of stock trading customers according to potential value

In this article, we use three clustering methods (K-means, self-organizing map, and fuzzy K-means) to find properly graded stock market brokerage commission rates based on the 3-month long total trades of two different transaction modes (representative assisted and online trading system). Stock traders for both modes are classified in terms of the amount of the total trade as well as the amount of trade of each transaction mode, respectively. Results of our empirical analysis indicate that fuzzy K-means cluster analysis is the most robust approach for segmentation of customers of both transaction modes. We then propose a decision tree based rule to classify three groups of customers and suggest different brokerage commission rates of 0.4, 0.45, and 0.5% for representative assisted mode and 0.06, 0.1, and 0.18% for online trading system, respectively.

Decision Tree based on data envelopment analysis for effective technology commercialization

Abstract In recent years, much governmental investment has been committed to R&D in the area of information technology industry. However, its commercialization rate is reported to be lower than expected. In order to prevent such waste, feedback information obtained from the rigorous evaluation of the finished R&D project needs to be utilized for future selection of new projects. The main purpose of our study is to provide the roadmap of the effective technology commercialization projects using the Decision Tree (DT) of data envelopment analysis (DEA) results when a company tries to develop or transfer its new technology. The environmental variables representing the characteristics of technology provider, receiver and technology itself are used as input variables for DT where the DEA results are used as a target variable. It is expected that our proposed approach can be effectively used to obtain the efficient scenario among the alternatives of several technology projects.

Structural equation model for effective CRM of digital content industry

Digital contents industry has been expanding its business based on broadband internet technology. But service level of contents provider has not reached the expected level of customers in the context of CRM yet. In this paper, we develop a structural equation model for customer satisfaction index (CSI) to measure the level of digital contents service quality reflecting the aspects of contents quality, service quality and provider quality. Our proposed model is applied to various types of contents such as on-line games, mobile contents, internet VOD services and e-music. The results give us the controllable feedback information to effectively improve customer satisfaction for each kind of digital content industry.

Cluster-based dynamic scoring model

Abstract Importance of early prediction of bad creditors has been increasing extensively. In this paper, we propose a behavioral scoring model which dynamically accommodates the changes of borrowers’ characteristics after the loans are made. To increase the prediction efficiency, the data set is segmented into several clusters and the observation period is fractionized. The computational results showed that the proposed model can replace the currently used static model to minimize the loss due to bad creditors. The results of this study will help the loan lenders to protect themselves from the potential borrowers with high default risks in a timely manner.

Predicting the financial performance index of technology fund for SME using structural equation model

As the technology credit fund is available to support SMEs which have intangible technology assets, many organizations are involved in technology evaluation with various factors. Technology evaluation has worth when the evaluation result can be related to financial performances of the recipient of credit funds. In this study, we propose a structural equation model (SEM) to analyze the relationship between technology evaluation factors and the financial performances by developing financial performance index (FPI). It is expected for the proposed model, which will be evaluated for the technology evaluation of enterprises, to be applied not only for the effective management of the technology credit funds for SMEs, but also will be used to evaluate financial performance of SMEs.

Cost of ownership model for the RFID logistics system applicable to u-city

Currently, popular areas of radio frequency identification (RFID) application include the domain of ubiquitous city (u-city), which would require very high initial infrastructure cost. This study proposes a cost of ownership (COO) model for RFID logistics system applicable to u-city in order to support the decision making process of infrastructure construction. We apply the proposed model to a case of RFID logistics system by establishing potential scenarios. The expected profit from the construction of RFID logistics system is evaluated for each scenario. We also conduct a sensitivity analysis to consider the effect of various parameter settings. This study is expected to help companies in selecting the most beneficial and most profitable RFID logistics system.

Managing loan customers using misclassification patterns of credit scoring model

A number of credit scoring models have been developed to evaluate credit risk of new loan applicants and existing loan customers, respectively. This study proposes a method to manage existing customers by using misclassification patterns of credit scoring model. We divide two groups of customers, the currently good and bad credit customers, into two subgroups, respectively, according to whether their credit status is misclassified or not by the neural network model. In addition, we infer the characteristics of each subgroup and propose management strategies corresponding to each subgroup.

Customer pattern search for after-sales service in manufacturing

Manufacturing corporations aim to sell their goods while they try to keep a sound customer relationship by providing high quality after-sales service (A/S). This is because while such services have always been important in marketing and sales industries, they are currently gaining importance in the manufacturing industry as well. Therefore, it is important to identify the needs of different customer groups and to provide respective A/S for each group accordingly. In this study, we propose a framework that consists of fuzzy clustering and an association rule to identify customer groups and their needs. We first carried out fuzzy clustering of customers in terms of indicators of CSI (customer satisfaction index). Next, the association rule is used to grasp the kind of A/S that customers consider important. Our results identified three groups of customers and their needs: Group 1 represents those who have a high degree of satisfaction, loyalty and high number of complaints. This group considers the home visiting service most important. Despite the fact that the Group 1 has high degree of complaint occurrence, they show a high degree of loyalty. Group 2 has very high degree of satisfaction and loyalty with a low level of complaints. This group considers important the A/S factors in all of service sections, including at the call center, the home visiting service, and claim handling. Group 3 has average satisfaction, number of complaints, and loyalty. Group 3 customers put weight on A/S factors dealing with the call center and the home visiting service. We expect that manufacturing firms can strengthen CRM (customer relationship management) by offering tailored A/S for each group accordingly.