Browse Prior Art Database

System and Method to Detect and Report Financial Suspicious Transactions

IP.com Disclosure Number: IPCOM000198084D
Publication Date: 2010-Jul-26
Document File: 3 page(s) / 136K

Publishing Venue

The IP.com Prior Art Database

Abstract

We propose the method to detect and report financial suspicious transactions with high efficiency. Commercial banks are facing a larger number of transactions and how to detect suspicious transactions with a high efficiency especially when the detection rules become more complex. In this paper, we assign a belief to each suspicious transaction detection rule, where this belief is also assigned to the suspicious transactions identified by this rule. We also propose the method to update the belief of each suspicious transaction detection rule according the manual affirms’ feedback, during the running of the suspicious transaction detection system. Therefore, we decide to report the suspicious transactions with the help of beliefs. For example, we report the suspicious transactions whose beliefs are larger than a given threshold, or report the given number of suspicious transactions with the largest beliefs.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 3

System and Method to Detect and Report Financial Suspicious Transactions

  According to the People's Bank of China, all commercial banks in China should report the suspicious transactions since 2007. China Anti-Money Laundering Monitoring & Analysis Center will process and analyze the reported suspicious transactions. During the detection of suspicious transactions by commercial banks, one of the most important problems is that the existing system and method detects too many suspicious transactions. Therefore, banks should arrange the manual confirms for the increasing number of transactions. Currently, the manual confirms do not feedback to a closed system. At the same time, to deal with the changing, adaptive models are very important.

  For the detection of suspicious transactions by a rule based system, if the number of transactions which satisfy the given rules (template, model, patterns) is very large, should all transactions be reported,

just as the existing application

system does? If not, how to select the suspicious transactions from thousands of hundreds? How to adaptively leverage the information from manual confirms to deal with the changing regulations and/or the changing criminal behaviors?

  In this paper, we propose the method to assign a belief to each suspicious transaction detection rule, where this belief is also assigned to the suspicious transactions identified by this rule.

  We also propose the method to update the belief of each suspicious transaction detection rule according the manual affirms' feedback, during the running of the suspicious transaction detection...