A TYPICAL STUDY OF IMPROVING ACCURACY IN DETECTING INSURANCE FRAUD ON UNSTRUCTURED DATA SETS

Fraud in insurance health care brings significant financial and personal loss on individuals, business, government and society as a whole. The size of health care sector and the enormous volume of money involved make it an important fraud target. The big data trend, (the growth in unstructured data) always leaves lots of rooms for a fraud going undetected if data is not analyzed properly. Performing big data analysis can identify repetitive errors that are hidden and prevent the occurrence of them in future. The primary objective of this paper is to define existing challenges of fraud detection for the different types of large data sets and ways to extract the features tha t cause fraud. It also deals with the methods for improving accuracy by considering both true positives and true negatives, thereby performing data analytics.