A health insurance system is an organization that provide health care services to meet the health needs of target population. Social health insurance is one of the primary methods of funding health insurance systems. Fraud is very rampant in today’s society and results in huge loss for the health insurance system. Intentional deception and misrepresentation are involved in frauds, so it will result in unauthorized benefits. Data mining is the technique applied to detect the fraudulent claims in the health insurance system. Data mining is basically a filtering through a large amount of data to get useful information. Elimination of fake claims is necessary to make health insurance industry free fraud.
Magna A : Student MESCE, Kuttippuram.
Dept. of Computer Science
MESCE, Kuttippuram.
data mining, SVM, unsupervised, supervised
Fraud becomes more diverse, as the amount of data grows. Reduction of fraud can result by elimination of fake claims. Data mining finds useful hidden patterns to present the required knowledge. After considering the advantages and disadvantages of most of the classification and clustering techniques. ECM is considered as the clustering technique and SVM as the classification technique. A health insurance claim prepared with the intention to deceive, conceal or distort relevant information that eventually accounts for health care benefits for an individual or a particular group is defined as fraudulent health insurance claim.Frauds by health insurance companies or its employees include preparation of bogus claims by fake physicians, billing for products or services not rendered, exaggerated claims submission, billing prepared for higher level of services, modifications or alterations made in submission of health insurance claims, change in diagnosis of the patient, fake documentation, and fraud committed by the employees of a hospital or any other healthcare product or service provider in order to make a quick buck.Fraudulent and dishonest health insurance claims are a major morale and moral hazard not only for the health insurance industry but even for the entire nation’s economy. Concrete proof as evidence including documentation, statements made by the policyholder and his family members and even neighbors are taken into consideration. The essential components of fraud include intention to deceive, derive benefits from the health insurance industry, preparation of exaggerated or inflated claims or medical bills and an intention to induce the firm to pay more than it otherwise would. Devising innovative methods and tactics including pressure tactics, favoritism and nepotism form a part of fraud which is a hazard growing by leaps and bounds since the last decade.To establish that a fraud has been committed requires furnishing of relevant proof. An indepth analysis of the health insurance policyholders intention may also be taken into consideration. It is estimated that the number of false health insurance claims in the industry is approximately 15 per cent of total claims. The report suggests that the healthcare industry in India is losing approximately Rs. 600 to Rs. 800 crores incurred on fraudulent claims annually. Health insurance is a bleeding sector with very high claims ratio. Hence, in order to make health insurance a viable sector, it is essential to concentrate on elimination or minimization of fake claims.
[1] Dr. BiswenduBardhan. Fraud in Health Insurance?, http://healthcare.financialexpress.com/200711/market13.shtml.
[2] MelihKirlidoga, CuneytAsuk(2012) A fraud detection approach with data mining in health insurance. Procedia - Social and Behavioral Sciences 62 ( 2012 ) 989 ? 994.
[3] Dan Ventura. Class Lecture, Topic: ”SVM Example”. BYU University of Physics and Mathematical Sciences, Mar. 12, 2009.
[4] Shunzhi Zhu, Yan Wang, Yun Wu, ”Health Care Fraud Detection Using Nonnegative Matrix Factorization”, The 6th International Conference on Computer Science Education (ICCSE 2011) August 3-5, 2011. SuperStar Virgo, Singapore.
[5] Zhongyuan Zhang, Tao Li, Chris Ding, Xiangsun Zhang, ”Binary Matrix Factorization with Applications”, Proceeding ICDM ’07 Proceedings of the 2007 Seventh IEEE International Conference on Data Mining Pages 391-400.
[6] Mohammad SajjadGhaemi. Class Lecture, Topic: ”Clustering and Nonnegative Matrix Factorization”. DAMAS LAB, Computer Science and Software Engineering Department, Laval University. Apr.12, 2013.
[7] Haesun Park. Class Lecture, Topic: ”Nonnegative Matrix Factorization for Clustering”. School of Computational Science and Engineering Georgia Institute of Technology Atlanta, GA, USA, July 2012.
[8] Fashoto Stephen G., OwolabiOlumide, Sadiku J., Gbadeyan Jacob A, ”Application of Data Mining Technique for Fraud Detection in Health
Insurance Scheme Using Knee-Point K-Means Algorithm”, Australian Journal of Basic and Applied Sciences, 7(8): 140-144, 2013 ISSN 19918178.
[9] Leonard WafulaWakoli. ”APPLICATION OF THE K-MEANS CLUS-
TERING ALGORITHM IN MEDICAL CLAIMS FRAUD/ABUSE DETECTION,” MSc Thesis, Jomo Kenyatta University Of Agriculture And Technology, 2012.
[10] Guido Cornelis van Capelleveen, ”Outlier based Predictors for Health Insurance Fraud Detection within U.S. Medicaid”, University of Twente University of California, San Diego December 2013.
[11] Qun Song, Nikola Kasabov, ”ECM A Novel On-line, Evolving Clustering Method and Its Applications”, Department of Information Science, University of Otago. H. Kopka and P. W. Daly, A Guide to LATEX, 3rd ed.
Harlow, England: Addison-Wesley, 1999.
