The Application of Neural Networks to Fraud Detection

Abstract

Nowadays, with the rapid development of the Internet, social reviews conducted through the Internet have become the main source for people to obtain product information. These reviews help individuals, companies and institutions make decisions. Although social commentary can help people provide more objective and comprehensive information, some individuals or organizations use this method to spread false and untrue information to the outside world, thereby affecting the outside world's judgment on the authenticity of the information, resulting in economic losses. Here is a study of user behavior and comment language to address the difficulties of money fraud. Social fraud detection uses a framework of three key components for review: the review itself, the user performing the review, and the item being reviewed are three key components used by social fraud detection. Under this framework, we do this through appropriate sequence modeling methods, Hidden Markov Models (HMM) and Artificial Neural Networks (ANN) are two examples. By summarizing and expanding the contributions of key persons in the subject of financial fraud, we assist new scholars in the field in providing some theoretical support.