Publication

Abstract : This study suggests using graph algorithms to detect healthcare fraud, a serious issue that can lead to significant financial losses and compromise patient care. Health care fraud is a representation of deceitful acts carried out by health providers, patients, or other individuals for the purpose of gaining undue financial profit from the health insurance program. It includes false billing, misrepresentation of diagnosis and procedure, and kickbacks at the expense of substantial financial loss and deterioration of patient care. Our approach models the relationships between patients, healthcare providers, physicians, and procedures as a graph to identify unusual patterns that may indicate fraud. We utilize publicly available tabular datasets, including beneficiary data and claims from inpatient stays and outpatient visits, transforming them into a graph structure to create a unified dataset. The system employs algorithmic analysis and graph construction techniques to identify significant nodes, abnormal clusters, and outliers within the healthcare network. This strategy offers a viable method for reducing healthcare fraud, enhancing the integrity of insurance systems, and ensuring that funds are appropriately allocated to legitimate medical needs. Ultimately, this study aims to decrease the rate of fraud in society.