• 보험의학회지
• /
• 제32권2호
• /
• pp.28-32
• /
• 2013

We studied trends of cardiac valve surgery using insurance data. 368 persons were included our study. We studied whether there are frauds or not. Only 4 cases were done at less than 1year from an insurance contract. We reviewed medical records of all persons. We could find the type of valve disease in 211 cases. The findings are atrial valve 40.1%, mitral valve 34.6% and others 25.3%. When we divided by materials of surgery, mechanical valves were used in 68.8% of men and 70.6% of woman. The main causes of valve disease were infection(55.1%). And degenerative valve disease 32% and congenital valve disease were 13%. We cannot find definite evidence of insurance frauds in the cardiac valve surgery. But there are some limitation in data analysis.

• 정보처리학회논문지:컴퓨터 및 통신 시스템
• /
• 제9권4호
• /
• pp.89-100
• /
• 2020

최근 보험 처리과정에서 소용되는 시간과 비용을 절감하고자, 자동차 보험에 블록체인 스마트 컨트랙트 기술을 도입하는 연구들이 활발하다. 그러나 기존의 연구들은 사고를 입증하기에 미흡한 수준의 교통 사고관련 데이터의 활용으로 악의적인 보험자의 사고 위조, 손상 확대 등의 보험사기 위협에 노출되어 있다. 이를 해결하고자, 본 논문에서는 자동차에 탑재된 센서, RSU, IoT 기기 등을 통한 다양한 종류의 데이터와 차량용 스마트 컨트랙트를 이용하여 사고데이터 기반 신뢰도 산정 모델을 제안한다. 특히 교통사고 데이터의 종류 및 상태에 따라 가중치를 달리하고, 다양한 사고 상황에 따라 학습되는 신뢰도 산정 모델을 고려하여 회귀모델을 적용했다. 제안 모델은 보험 처리과정의 투명성, 보험 처리 과정의 간소화와 같은 기존 장점을 유지하며 효과적인 보험사기 차단, 보험 소송의 감소의 효과를 보일 것으로 기대된다.

• 보험의학회지
• /
• 제28권1_2호
• /
• pp.31-35
• /
• 2009

Background : Many of internists have been working for insurance industry. Insurance medicine is use of medical knowledge for insurance industry. There is social role of insurance medicine in terms of soundness of insurance administration. Recently social role of internists also have been being watched. Although theme of insurance medicine is medical risk selection, insurance claims administration also needs medical experts'opinion. There are not any corroborative study of medical consulting for insurance claims. Among insurance industry, someone called this medical review of insurance claims as 'medical claims review'. Aim : To investigate usefulness of medical review of insurance claims. Design : Questionnaire survey with claim staffs in one of insurance claim adjustment company in Korea. Methods : 265 claim staffs were divided into 4 groups and conducted survey using a questionnaire of 20 questions. Utility score, job satisfaction score, and difficult factors of claims administration were measured. Results : Utility score and job satisfaction score are highest in medical claims review group. The most difficult in claim administration to claim staffs was demonstrated to medical knowledge. Conclusion : Medical review of insurance claims is proved to be worthy. Document-based consulting method, namely medical claims review, is more useful than telephone-based simple query among claim staffs...Subjects of the medical claims review are medical record and it's principle is independent medical examination with evidence-based approach, it also has role of protecting fraud of insurance claims. Two main question types of medical claims review are verification and advice.

• International Journal of Computer Science & Network Security
• /
• 제21권9호
• /
• pp.125-131
• /
• 2021

Detecting fraudulent insurance claims is difficult due to small and unbalanced data. Some research has been carried out to better cope with various types of fraudulent claims. Nowadays, technology for detecting fraudulent insurance claims has been increasingly utilized in insurance and technology fields, thanks to the use of artificial intelligence (AI) methods in addition to traditional statistical detection and rule-based methods. This study obtained meaningful results for a fraudulent insurance claim detection model based on machine learning (ML) and deep learning (DL) technologies, using fraudulent insurance claim data from previous research. In our search for a method to enhance the detection of fraudulent insurance claims, we investigated the reinforcement learning (RL) method. We examined how we could apply the RL method to the detection of fraudulent insurance claims. There are limited previous cases of applying the RL method. Thus, we first had to define the RL essential elements based on previous research on detecting anomalies. We applied the deep Q-network (DQN) and double deep Q-network (DDQN) in the learning fraudulent insurance claim detection model. By doing so, we confirmed that our model demonstrated better performance than previous machine learning models.

• 산경연구논집
• /
• 제8권6호
• /
• pp.25-31
• /
• 2017

Purpose - This study examines car accidents that occurred in South Korea territory, and analyzes criminal liability of the offender and certain issues of driver's insurance, but a civil liability to the injured is excluded as civil liability belongs to auto insurance. Research design, data, and methodology - With carrying out this research, case study of driver's liability and literature review were adopted throughout. For this, car accidents that occurred in South Korean territory were examined and then criminal liability of the offender and certain issues of driver's insurance were analyzed. Results - From this case study on driver's liability it was found that the offender cannot receive insurance money from the insurer irrespective of the valid drive insurance, if there is no 'bill of agreement of criminal consensus'. This study suggests some ideas, offers suggestions of convenience and assistance of qualified claim staff to overcome a hurdle of drive insurance. Conclusions - As long as the accident is not a fraud and scam by the parties concerned, advance payment of agreement of criminal consensus is required to the insured, the policy holder within the limit of liability of driver insurance, on condition that the drive insurance is valid.

• 디지털융복합연구
• /
• 제11권10호
• /
• pp.593-608
• /
• 2013

상해상병으로 청구되는 건수가 증가함에 따라 조사 대상을 보다 정교하게 선정하여 상해요인 조사 대상을 줄이면서 환수율 및 환수금액을 올릴 수 있는 방안을 마련할 필요가 있다. 이를 위해서 2006~2011년까지의 상해요인 조사자료를 수집하여 의사결정나무 모형을 활용하여 지역가입자 상해상병 진료건에 대한 부당환수 조사대상 선정모형을 개발하였다. 최종 개발된 모형결과에 따르면, 조사대상 유형은 18개로 분류되었고, 이러한 분류결과는 실제 조사가 시행될 시, 모형을 적용하지 않았을 때 보다 최고 12.8배 높은 부당환수결정율을 나타낼 수 있을 것으로 분석되었다. 또한, 본 연구에서 개발된 조사 대상자 선정 모형을 실제 업무에 적용하기 위해서는 조사물량 대비 국민건강보험공단의 조사인력 및 운영 계획을 보다 면밀히 검토해야만 모형 적용의 효과성이 극대화 될 수 있을 것으로 판단된다.

• 융합보안논문지
• /
• 제16권6_2호
• /
• pp.53-60
• /
• 2016

현재의 보험범죄에 대한 보안대책은 보험금 지급심사단계에서 적발중심적인 사후대응에 치중하고 있다. 그러나 사후 처벌만으로는 치유할 수 없는 피해를 남기기 때문에 보험범죄를 사전에 예방할 수 있는 대책을 강구할 필요가 있다. 따라서 이 논문에서는 보험범죄의 특성과 사례분석을 통하여 문제점을 규명해보고, 이에 대한 대안을 제시해보고자 한다. 현행 보험제도의 문제점은 첫째, 보험계약 체결시에 보험계약자의 신용상태, 중복가입여부, 자발적 가입여부 등에 대한 확인이 허술하다는 것이다. 둘째, 보험사고 발생시 철저한 조사나 형사고발이 제대로 이루어지지 않고 있다. 셋째, 악성보험계약자에 대한 정보교환이나 관리가 미흡하다. 따라서 보험범죄부터 국민을 지키기 위해서는 첫째, 보험계약시에 보험계약자들에 대한 신용상태, 사고경력 등에 대한 심사를 강화해야 할 것이다. 둘째, 보험사기인지시스템과 사회관계망의 지속적인 업그레이드를 통하여 보험범죄를 사전에 차단할 수 있도록 해야 할 것이다. 셋째, 보험사간에 정보를 공유하여 보험범죄에 대한 감시시스템을 강화하여야 할 필요가 있다.

• 한국IT서비스학회:학술대회논문집
• /
• 한국IT서비스학회 2009년도 춘계학술대회
• /
• pp.357-360
• /
• 2009

본 연구는 데이터마이닝 기법을 이용하여 건강보험청구료에 있어서 이상정도가 심한 요양기관을 탐지하고, 실제 의료영역에 적용하기 위한 시스템 개발을 목적으로 한다. 현재 건강보험 심사평가원의 이상탐지시스템은 평가대상이 되는 항목을 개별적으로 평가하고, 탐지된 기관의 선정 이유에 대한 근거제시가 부족한 단점을 가지고 있다. 따라서 본 연구에서는 항목을 종합적으로 평가할 수 있는 정량적 지표를 설계하고, 항목들의 상대적 중요도를 파악할 수 있도록 항목들에 대한 가중치 부여한다. 또한 지표에서 얻어진 값으로 등급을 구분하고, 의사결정나무기법(decision tree)를 이용하여 해석력을 높이는 방법을 제시한다.

• 보건행정학회지
• /
• 제19권3호
• /
• pp.109-124
• /
• 2009

Background: This study aimed to examine whether cases of punishing false claimants threat general physicians to check their medical cost claims with care to avoid being suspected, and identify empirically general deterrence effects of the on-site investigation system in the Korean National Health Insurance. Methods: 800 clinics were selected among a total of 15,443 clinics that had no experience of on-site investigation until June 2007 using a stratified proportional systematic sampling method. We conducted logistic multiple regression to examine the association between factors related to provider's perception of on-site investigation and high level of perceived deterrence referring to fear of punishment after adjusting provider's service experiences and general characteristics. Results: The probability of high perceived deterrence was higher 1.7 times (CI: 1.13-2.56), 2.73 times (CI: 1.68-4.45) each among clinics exchanging the information once or more per year or once or more for 2-3 months than among clinics no exchanging the information about on-site investigation. Also, the probability of high perceived deterrence was higher 2.27 times (CI: 1.28-4.45) among clinics that knows more than 3 health care institutions having experienced an on-site investigation than among clinics knowing no case. Conclusion: A clinic knowing more punishment cases by onsite investigation and exchanging more frequently information about on-site investigation is likely to present high perceived deterrence. This result will provide important information to enlarge preventive effects of on-site investigation on fraud and abuse claims.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 2001년도 The Pacific Aisan Confrence On Intelligent Systems 2001
• /
• pp.97-108
• /
• 2001

Frauds detection is a difficult problem, requiring huge computer resources and complicated search activities Researchers have struggled with the problem. Even though a fee research approaches have claimed that their solution is much better than others, research community has not found 'the best solution'well fitting every fraud. Because of the evolving nature of the frauds. a novel and self-adapting method should be devised. In this research a new approach is suggested to solving frauds in insurance claims credit card transaction. Based on evolutionary computing approach, the method is itself self-adjusting and evolving enough to generate a new self of decision-makin rules. We believe that this new approach will provide a promising alternative to conventional ones, in terms of computation performance and classification accuracy.