A Comparative Analysis of Fire Insurance Risk Classification: Linear Additative Models and Neural Network Approach

Gim, Gwangyong김광용 숭실대학교 경영경제전략연구소 1996 숭실경영경제연구 Vol.26 No.-

인터넷 설문조사의 방법론적인 문제점과 데이터마이닝 기법을 활용한 개인화된 인터넷설문조사 시스템의 구축

김광용,김기수 한국품질경영학회 2004 품질경영학회지 Vol.32 No.2

The purpose of this research is to summarize the methodological issues in internet survey and to suggest personalized internet survey system using data mining technique for enhancing the survey quality of internet survey as well as utilizing the benefit of interactive multimedia factors of internet survey. The data mining technique used in this paper is Case Based Reasoning for adopting individual design preference affecting survey quality. For achieving the research purpose, two surveys, pre & post survey, were performed. Pre survey was done for implementing CBR database to find individual index affecting survey quality and post survey was used for measuring the performance of personalized internet survey system. The result shows that the survey quality of personalized web survey system is better than generalized web survey system.

여러 가지 Inductive 방법에 대한 통합모델 개발과 그 실증적 유효성에 대한 연구

김광용 한국경영과학회 1998 韓國經營科學會誌 Vol.23 No.3

This research investigates computer generated hybrid second-order model of two numerically based approaches to risk classification: discriminant analysis and neural networks. The hybrid second-order models are derived by rule induction using the ID3 and tested in the several different kinds of data. This new hybrid approach is designed to combine the high prediction accuracy and robustness of DA or NN with perspicuity of ID3. The hybrid model also eliminates the problem of contradictory inputs of ID3. After doing empirical test for the validity of hybrid model using small and medium companies' bankrupt data, hybrid model shows high perspicuity, high prediction accuracy for bankrupt, and simplicity for rules. The hybrid model also shows high performance regardless the type of data such as numeric data, non-numeric data, and combined data.

홈네트워크의 특성이 아파트의 가치인식에 미치는 영향에 관한 실증적 연구

강형모(Hyoungmo Kang),김광용(Gwangyong Gim),김신곤(Shinkon Kim),김정곤(Jongkon kim) 한국IT서비스학회 2007 한국IT서비스학회지 Vol.6 No.3

  Korean homenetwork industry is now faced with problems like low profitability and unclear commercialilzation. In particular, low awareness of homenetwork usefulness is one of the main reasons why homenetwork industry is not developed.<BR>  Therefore, this study tried to find what kinds of characteristics of homenetwork affected the value of apartment and how. we conducted a survey of residents of apartment with homenetwork installed in it. Surveyed data was analyzed based on the model of extended Technology Acceptance Model(TAM) including perceived enjoyment and perceived trust as new factors.<BR>  The research result showed that all factors involved in suggested model had positive effects on the behavioral intention of using homenetwork and the value of apartment. Most importantly, homenetwork user"s behavioral intention increased perceived housing value of the apartment.<BR>  The research result can be used in explaining the advantage of homenetwork to the residence of apartment as well as in designing the homenwtwork install of apartment.

Service De나dml 서비스 품질이 조직의 업무 효율성 증대에 미치는 영향에 관한 연구

김동철(DongChul Kim),김광용(GwangYong Gim),임성택(SeongTaek Rim) 한국IT서비스학회 2009 한국IT서비스학회지 Vol.8 No.4

This study is to guide to estimate service quality on ITSM operation based on ITIL(IT Infrastructure Library) ITIL v2 and v3 provide how to integrate framework of IT service and business process. It is widely used as ITSM operation base. As a latest IT trend, ITSM covers Hardware, Software, SaaS, Network, Call center Helpdesk, ASP portal and IT operation. Servicedesk is selected as target area where valid sampling is addressable and service change is rapid. Traditional Helpdesk was focusing on technical support to solve internal IT issues passively. But it was evolved into Service Desk which focus in process and provide integrated service from customer's business view point preventively and proactively. Accordingly outsourcing types business are normally performed by group of professional capability. Service quality is measured under the SLA(Service Level Agreement). This study utilized SERVQUAL model as service quality measuring tool developed by Parasuraman, Zeitaml and Berry to find critical factors to satisfy customer. And test was processed regarding effectiveness of IT organization and customer view point thru sampling. Though valid parameters can be changed by ITSM area under the SERVQUAL models, they naturally can be accepted as a index of service quality measurement after sampling test with acceptable significance. And I recommend to follow this study as a preparation before official SLA

보험사기행동모형 개발에 관한 실증적 연구

이명진(Myungjin Lee),김광용(Gwangyong Gim) 한국IT서비스학회 2007 한국IT서비스학회지 Vol.6 No.2

Many researches have been done in insurance fraud as the amount and frequency of insurance fraud have been increasing continuously. In particular, the development of insurance fraud detection system using large database management techniques including data mining or link analysis based on visual method have been the main research topic in insurance fraud. However, this kinds of detection system were very ineffective to find unintentional insurance fraud happened by accident even though it was so good to find intentional and organized crime insurance fraud. Therefore, this research suggests insurance fraud as an ethical decision making and applies TPB(Theory of Planned Behavior) for the finding of reasons and prevention strategies of unintentional insurance fraud happened by accident. The results of research show that TPB is very appropriate model to explain the behavior of insurance fraud and that insurance agents force to do insurance fraud as affecting perceived behavior control. Therefore, education and pubic relations for insurance fraud are very effective for preventing insurance fraud and developing insurance service industry.

소비자 리뷰 특성이 모바일 커머스 성과에미치는 영향에 관한 연구

정상빈(Jeong, Sangbin),김광용(Gim, Gwangyong),임은택(Im, Euntack),김도연(Kim, Dohyeon) 글로벌경영학회 2020 글로벌경영학회지 Vol.17 No.6

모바일 기술과 네트워크 기술의 발전으로 소비자들은 언제 어디서든 자신이 구매하고자 하는 제품에 대하여 탐색하고 제품을 비교함으로써 최종적으로 구매하는 것이 가능하다. 이러한 유형의 상거래를 모바일 커머스(mobile commerce)라고 정의할 수 있으며 최근에는 스마트폰 이용률이 높아지면서 전통적인 전자상거래(e-commerce)의 성장세를 뛰어넘어 빠르게 성장하고 있다. 하지만 모바일 커머스는 PC 기반 인터넷 중심의 전통적인 전자상거래에 비해 작은 화면과 이동 중에 이뤄진다는 점 때문에 소비자는 자신의 제한적인 정보 획득 공간과 시간에서 정보를 습득하여 의사결정을 해야 한다. 따라서 모바일 커머스의 특성이 잘 갖춰진 환경에서 정보의 특성이 고객 의사결정에 어떠한 영향을 미치는지에 대한 연구가 필요하다. 본 논문은 싱가포르의 모바일 커머스 전문 플랫폼인 Shopee에서 판매 중인 남성용 티셔츠에 대한 계량적 리뷰와 텍스트 리뷰, 가격, 선호도와 판매량의 데이터를 HTMP parsing의 크롤링을 통해 수집하여 텍스트 리뷰는 빈도 분석과 LDA( Latent Dirichlet allocation) 토픽 모델링을 통해 특성과 키워드로 추출하고 Google language API를 통해 감성 분석함으로써 리뷰의 비계량적 특성을 측정하였다. 그 후 선형 회귀분석을 통해 선호도와 구매 의사결정에 리뷰 특성이 어떠한 영향을 미쳤는지 파악하였다. 분석 결과, 비계량적 리뷰 특성에서는 크기를 특성들이 선호도와 판매성과에 유의미한 영향을 미쳤다. 계량적 특성에서는 평균 평점이 선호도에는 유의미한 영향을 미치지 못하였으나 판매성과에 리뷰 총량과 함께 유의미한 영향을 미쳤다. 이를 통해 리뷰의 특성이 선호도와 판매성과에 유의미한 영향을 미치며 의사결정 목적에 따라 다른 의사결정 과정을 거친다는 것을 확인하였다. Since advances in mobile and network technology, consumers can purchase products anytime, anywhere by exploring and comparing products they want to buy. And this type of commerce can be defined as mobile commerce, and more recently, as smartphone usage has grown, surpassing the growth of traditional. However, mobile commerce is used in online virtual space on a smaller screen compared to traditional. Therefore, it is necessary to study how the characteristics of information provided by mobile commerce affect customer decision making. This paper collected data on quantitative reviews and text reviews, prices, favorites and sales volume through the crawling of HTML parsing from men s T-shirts on sale at Shopee, Singapore s mobile commerce specialized platform. From text review, characteristics and keywords were extracted through frequency analysis and LDA (Latent Dirichlet allocation) topic modeling. The sentiment attributes of the product review were analyzed based on keywords through the google language API. Finally, a linear regression analysis performed how the review characteristics affected the favorite and purchasing decision. As a result of the analysis, size characteristics, in the non-measuring review characteristics, don’t had a significant impact on favorites and sales. In quantitative characteristics, the average rating did not have a significant impact on preference, but it had a significant effect on sales performance with the total amount of reviews. This research confirms that the characters of the review have a significant impact on favorite and sales performance and is subject to different decision-making processes depending on the decision-making objective.

비즈니스 컨설팅 서비스의 품질과 편익이 고객과의 결속관계에 미치는 영향에 관한 연구

강형모(Hyoungmo Kang),김광용(Gwangyong Gim) 한국IT서비스학회 2008 한국IT서비스학회지 Vol.7 No.1

  A business consulting service combines company"s knowledge and information with external"s ones systematically. This combination achieve knowledge synergy and make added value for the business as a whole. Therefore, a business consulting service has been called Knowledge Ware Industry.<BR>  Consulting corporations and consultants must provide valuable services for running and growing their business successfully. A value of business consulting service could be identified offset quality and benefits from costs. This study conducted survey about business consulting service quality by SERVPERF model and benefit, customer commitment.<BR>  The research result showed that the quality of business consulting service effects on the benefit positively and benefit increased customer commitment. Finally, there is positive relationship between customer commitment and intention of continuing of the business consulting service.

공공부문의 지식활동 측정 지표에 관한 연구

강형모(Hyoungmo Kang),김광용(Gwangyong Gim) 한국IT서비스학회 2007 한국IT서비스학회지 Vol.6 No.1

  An objective and valid knowledge activity indicator (KAI) is essential for encouraging knowledge activity in public institution. However, knowledge activity of public institution is inherently different from one of private corporation and, therefore, new method to find aggregated knowledge mileage considering public organizational culture and structure is to be developed.<BR>  Therefore, this paper shows how to develop KAI of public institution based on three steps. First, knowledge activities were categorized as knowledge creation, knowledge sharing, and knowledge value increasing. Several knowledge indicators and hierarchical knowledge structure were developed for each knowledge categories. Second, personal knowledge mileage was calculated using AHP method considering whole organization’s opinion. Finally, these approaches to find KAI were applied to “P” public institution as case studies. The results shows that KAI developed in this paper was very effective in grouping knowledge workers based on knowledge activities.

오픈코스웨어의 지속적 사용의도에 영향을 미치는 요인에 관한 연구

이재모(Jaemo Lee),김광용(Gwangyong Gim) 한국IT서비스학회 2016 한국IT서비스학회지 Vol.15 No.2

Recently, Open Course Ware service area which offers free lecture materials has been steadily growing. This change comes from a paradigm shift of education service caused by the progress of ICT and a role change of consumers. Knowledge creation from higher educational institutions is slow compared to change of society. Therefore, consumers escape from one-sided education, and perform a new role of knowledge generation. Nevertheless, there is not much of academic research of Open Course Ware. Consequently, the paper is dedicated to empirical study on effect of perceived value, information system quality and brand trust on intention to continuous use. As implications for the research, as a practical point of view, the results-an economic side that offers contents for free of charge, a contents side that is fresh, various and substantial, a systematic side that makes customers to use the system easily, a service side which takes charge of customers support and brand trust affect satisfaction and intention to continuous use– can be used for establishment of executive strategies.