As a guest user you are not logged in or recognized by your IP address. You have
access to the Front Matter, Abstracts, Author Index, Subject Index and the full
text of Open Access publications.
With the rapid development of digital and intelligent credit industry, credit fraud detection has become an important task to ensure financial digital intelligence. The traditional credit fraud detection model relies on artificial feature engineering and is built based on supervised learning algorithm, and its performance is greatly affected by data quality, sample distribution and other factors, and it is prone to prediction errors in the face of emerging fraud techniques. With the development of digitalization and fraud methods, these methods are often no longer applicable. There are rich information associations among users such as the users’ emergency contacts, home address, and social relationships between themselves, which make a large social network graph formed between users. In this regard, based on the Dgraph-Fin dataset, this paper uses the relationship network formed among users to better learn the weights between different edges through neighbor sampling and attention mechanism. Experimental results show that the accuracy and effectiveness are improved compared with the existing baseline.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.