TWI775020B - Money laundering prevention law and suspected case aid judgment system - Google Patents

Abstract

This case proposes a suspected case-assisted judgment system for money laundering prevention and control. The system includes a manual trial database, a big data analysis module, a speculation module, and an audit module. The manual trial database is used to store an audit result; the big data analysis module is used to analyze the transaction data and give a risk assessment; wherein the speculative module is connected with the manual trial database and the big data analysis module. To analyze the suspected case when the audit result and the risk assessment reach the risk threshold; and the audit module, which is connected with the speculative module and the manual trial database, is used to review the suspected case and generate the audit result, and transmit the audit result To the manual trial database.

Description

Auxiliary Judgment System for Suspected Cases of Money Laundering Prevention Law

The present invention relates to a money laundering prevention and control law suspected case applied to the money laundering prevention and control law by means of big data analysis and manual case review database

In recent years, in the face of an increasingly severe financial regulatory environment, more and more illegals use the management loopholes of financial institutions to launder money and conduct other criminal acts. Therefore, how to formulate financial transaction procedures and rules to prevent money laundering has also become a major issue that financial institutions must face. In order to avoid such illegal money laundering crimes, various financial institutions have also adopted the Money Laundering Prevention Act (AML) to deal with such problems. And constantly revise and improve the current rules and procedures of financial transactions to avoid the increasingly serious money laundering crimes.

However, in the process of implementing the money laundering prevention law, in order to ensure the legitimacy of each transaction, the operator must compare the data of each transaction one by one to determine whether the transaction is related to illegal elements and illegal transactions. However, this approach will increase the bank's enormous personnel costs. And such a check not only consumes a lot of time but also increases the probability of errors. And often when judging suspected blacklisted persons/organizations/transactions, because the information that can be referred to is very limited, the clerk often encounters situations where it is difficult to judge and decide, which causes the customer to be troubled by multiple checks or released for criminal acts. .

And this artificial method of judging suspected cases, in order to enable auditors to judge suspected cases more correctly and efficiently, and to ensure that auditors can judge suspected cases in accordance with the laws and regulations related to money laundering in different countries or regions. It is necessary to pass the intensive training of various financial institutions in order to make the audit Nuclear man has this ability. However, such an approach will consume considerable training costs, and increase the load and personnel costs of various financial institutions.

Therefore, in order to solve the above problems, this case proposes an auxiliary judgment system for suspected cases under the Money Laundering Prevention Law. The system includes a manual case trial database, a big data analysis module, a guessing module and an auditing module. The manual trial database is used to store an audit result; the big data analysis module is used to analyze transaction data and give risk assessment; the speculation module is connected to the manual trial database and the big data analysis module, and is used for It is used to analyze suspected cases when the audit results and risk assessment reach the risk threshold; and the audit module, which is connected with the speculation module and the manual case database, is used to audit suspected cases and generate audit results, and transmit the audit results. to the manual trial database.

According to an embodiment, the above-mentioned review result is a result after reviewing the suspected case by the reviewer.

According to an embodiment, the above-mentioned transaction data includes customer occupation information and transaction records, wherein the customer occupation information can be weighted according to whether they are classified as political figures or regional attributes; wherein the transaction records can be classified into multiple accounts according to the amount of capital flow and the same account. .

According to an embodiment, the above risk assessment is divided into high risk, medium risk and low risk by the big data analysis module by analyzing the customer's professional information and transaction records.

According to an embodiment, the above-mentioned suspected case is a transaction case with money laundering risk inferred by the speculation module, and the review result and risk assessment of past reviewers are attached to the suspected case.

According to an embodiment, the above risk threshold is set as a threshold for determining the proportion of money laundering cases in the audit result and the score of the risk assessment.

This case proposes an auxiliary judgment method for suspected cases under the Money Laundering Prevention Law. The method is as follows: firstly, the big data analysis module analyzes the customer's professional information and transaction records of the transaction data and then gives a risk assessment. The big data analysis module transmits transaction data and risk assessment to the speculation module. The manual case review database transmits the review results of the same transaction data stored to the speculation module. The speculative module analyzes the transaction data with the risk assessment and past audit results, and ends all processes if the risk threshold is not exceeded. If the transaction data exceeds the risk threshold, a suspected case will be generated, and the speculation module will transmit the suspected case, risk assessment and past audit results to the audit module. And the review module will review suspected cases by reviewers and generate review results, and finally send the review results to the manual case review database for storage. and end all processes.

According to an embodiment, the above-mentioned review result is a result entered into a manual case review database after reviewing the suspected case by reviewers.

According to an embodiment, the transaction data includes customer occupation information and transaction records, and the customer occupation information is classified according to political figures or regional attributes; and the transaction records can be classified into multiple accounts according to the flow of funds and the same account.

100: Manual trial database

110: Big Data Analysis Module

120: Speculation Module

130: Audit Module

200, 210, 220, 230, 240, 250, 260: Steps

Figure 1 is a diagram showing the relationship between the money laundering prevention law auxiliary judgment system.

Fig. 2 is a flow chart showing the auxiliary judgment method of the money laundering prevention law.

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

Please refer to Figure 1. Figure 1 is a diagram showing the relationship between the money laundering prevention law auxiliary judgment system. The system includes a manual case trial database 100 , a big data analysis module 110 , a guessing module 120 and a review module 130 . The manual case review database 100 is a database system placed on a server, used to store reviewers' review results of suspected cases, and provide a guessing module 120 to read relevant review result data. And as the amount of data in the manual case trial database 100 increases, the accuracy of the suspected case inferred by the inference module 130 also increases. Therefore, in the long run, it can be used to achieve the goal of automatically hearing cases and reduce the time for processing cases.

Please refer to Figure 1. Figure 1 is a diagram showing the relationship between the money laundering prevention law auxiliary judgment system. The big data analysis module 110 may be a computer of the bank. It is classified according to customer professional information and transaction records, and analyzed according to the classification results of big data. And according to the customer's occupational information and transaction records, it is divided into high risk, medium risk, and low risk, and a risk score is given to facilitate the speculation module 110 to speculate. For example: the client’s occupational information is a political figure or a trader with North Korea and Iran, and the transaction records are records of frequent large-scale capital flows, multiple accounts to the same account, the same account to multiple accounts, or frequent transactions with North Korea and Iran. at high risk.

Please refer to Figure 1. Figure 1 is a diagram showing the relationship between the money laundering prevention law auxiliary judgment system. The speculating module 120 may be a computer of the bank. It needs to be matched with the manual trial database 100 and the big data analysis module 110 . When judging whether a client is a suspected case or not, the inference module 120 infers the data in the manual case trial database 100 and the risk level and risk score transmitted from the big data analysis module 110 . In addition, a client must have at least 30 pieces of similar data in the manual trial database 100 before the data in the manual trial database 100 can be brought into the inference module 110 . Example: If the same customer exceeds a certain percentage (eg 50%) in the manual trial database 100, it will be judged as a money laundering case, or the risk level and risk score from the big data analysis module 110 exceed a certain threshold. 120 is judged as a suspected case.

Please refer to Figure 1. Figure 1 is a diagram showing the relationship between the money laundering prevention law auxiliary judgment system. The audit module 130 may be the host of each financial institution. It is a transaction result file and a report file generated by inferring that the speculation module 120 is a suspicious customer and transaction case, and then sent back to various financial institutions for review by means of an encrypted web server or web page. In addition to the suspected case, the returned results will also provide the previous review results of other personnel and the risk assessment after big data analysis on the suspected case, and give an appropriate risk score for the reviewer's reference.

Please refer to Figure 2. Figure 2 is a flow chart illustrating the method of assisting the judgment of the money laundering prevention law. The step 200 is to place various lists (providing a custom list upload function), customer information and transaction records in the server array of the United Center. The step 210 is to analyze the customer professional information and transaction records of the transaction data through the big data analysis module 110 to give risk assessment, and transmit the transaction data and risk assessment to the prediction module 120 . The big data analysis module 110 can be a computer of the bank. Risk assessment is a kind of classification based on customer occupational information and transaction records, and analyzes the results of big data classification, divides customer occupational information and transaction records into high risk, medium risk, and low risk, and assigns a risk score to facilitate the inference module 110 speculate. For example: the client’s occupational information is a political figure or a trader with North Korea and Iran, and the transaction records are records of frequent large-scale capital flows, multiple accounts to the same account, the same account to multiple accounts, or frequent transactions with North Korea and Iran. at high risk.

Please refer to Figure 2. Figure 2 is a flow chart illustrating the method of assisting the judgment of the money laundering prevention law. The step 220 is to analyze the transaction data through the money laundering risk assessment and the review result of the manual case review database 100 . First, the manual case review database 100 transmits the review result of the same stored transaction data to the inference module 120 to determine whether the case is a suspected case. When judging whether the client is a suspected case, the inference module 120 infers the data in the manual case trial database 100 and the risk level and risk score transmitted from the big data analysis module 110 . Example: If the judgment result is that the risk threshold is not exceeded, the risk The assessment and past audit results are passed to the audit module 130 . If the judgment result is that the transaction data exceeds the risk threshold, a suspected case is generated, and the suspected case, risk assessment and past audit results are sent to the audit module 130 . The risk threshold is set as the threshold for the judgment of the proportion of money laundering cases in the audit results and the score of the risk assessment. Example: If the same customer exceeds a certain percentage (for example: 50%) in the manual trial database 100, it will be judged as a suspected money laundering case, or the risk level and risk score from the big data analysis module 110 exceed a certain threshold. The group 120 judges the case as a suspected case. The speculation module 100 analyzes the transaction data with the risk assessment and past audit results. The speculating module 120 may be a computer of the bank. It needs to be matched with the manual trial database 100 and the big data analysis module 110 . In addition, a client must have at least 30 similar data in the manual trial database 100 before the data in the manual trial database 100 can be brought to the inference module 110 .

Please refer to Figure 2. Figure 2 is a flow chart illustrating the method of assisting the judgment of the money laundering prevention law. The step 230 is to manually determine whether it is a suspected case. Manual judgment takes time to investigate the legitimacy of customer and transaction data sources, and every time a case is confirmed, the reasons for passing and failing the scan must be written. And after the judgment is completed, the review result is sent to the manual case review database 100 . The manual case review database 100 is a database system placed on a server for storing the review results of the reviewers on suspected cases, and providing the presumption module 100 to read the relevant review result data.

Please refer to Figure 2. Figure 2 is a flow chart illustrating the method of assisting the judgment of the money laundering prevention law. The step 240 is to quickly review the suspected case by the reviewers through the review module 130, and after the review is completed, the review result is generated, and the review result is sent to the manual case review database 100 for storage (step 250), and all processes are ended (step 250). 260). The manual case review database 100 is a database system placed on a server for storing the review results of the reviewers on suspected cases, and providing the presumption module 100 to read the relevant review result data. And with the increase in the amount of data in the manual trial database 100, it is speculated that the module The accuracy of the suspected cases inferred by 130 also improved. Therefore, in the long run, it can be used to achieve the goal of automatically hearing cases and reduce the time for processing cases. Finally, all processes are ended (step 260).

Based on the above, this case proposes an automatic judgment of suspected cases by means of big data analysis and manual trial database. The reviewers will store the review results of suspected cases in the manual case review database. As more and more transaction data are stored in the manual trial database, the accuracy of judgment is also getting higher and higher, and finally a more accurate and efficient automatic trial method can be achieved. Therefore, in order to ensure the legitimacy of each transaction, it can solve the problem that the operators must compare the transaction data one by one and spend a lot of time and cost. And reduce the probability of judgment error and the training cost of training auditors. And because the available reference materials are very limited and it is difficult to judge and decide, which causes customers to be troubled by multiple inspections or released for criminal acts. Therefore, the technology in this case can solve the above problems and make the judgment of suspected cases more accurate and efficient.

However, the illustrations and descriptions disclosed above are only preferred embodiments of the present invention, and are not intended to limit the implementation of the present invention. Those who are familiar with the art may make changes based on the spirit of the present invention. or modification, should be covered within the scope of the following patent application in this case.

100: Manual trial database

110: Big Data Analysis Module

120: Speculation Module

130: Audit Module

Claims (5)

An auxiliary judgment system for suspected cases of money laundering prevention law, which includes: a manual case database for storing an audit result; a big data analysis module for analyzing a transaction record and a customer's professional information, A risk assessment is given to obtain a risk score, wherein the risk assessment is classified into a risk level of high risk, medium risk or low risk according to the client's professional information and the transaction record, and the risk score is given; a speculation module , which is connected to the manual trial database and the big data analysis module, when the proportion of the money laundering cases of a client exceeds 50% in the manual trial database, or the risk level is high risk, or the When the risk score exceeds a predetermined value, the inference module is judged as a suspected case; and a review module is connected with the inference module and the manual case review database, and is used for providing the suspected case to a reviewer personnel refer to it, generate the audit result, and transmit the audit result to update the manual case review database; wherein, when a client needs at least a predetermined number of the transaction records in the manual case review database, the The transaction record of the customer will be brought into the speculation module. The auxiliary judgment system for suspected cases of money laundering prevention law as described in claim 1, wherein the transaction data includes the client's professional information and the transaction record, and the client's professional information is classified according to political figures or regional attributes; and the transaction record can be Categorize multiple accounts by cash flow and the same account. The auxiliary judgment system for suspected cases of money laundering prevention law as described in claim 1, wherein the suspected case is a transaction case with money laundering risk inferred by the speculation module, and the audit result of the auditor in the past is attached to the suspected case and this risk assessment. An auxiliary judgment method for suspected cases of money laundering prevention law, comprising: A big data analysis module analyzes a transaction record and a client's professional information, and then gives a risk assessment to obtain a risk score, wherein the risk assessment is classified into high risk, medium risk or high risk according to the client's professional information and the transaction record. A risk level of low risk, and assign the risk score; a speculation module receives the transaction records, the proportion of money laundering cases and the risk score from the big data analysis module and a manual case database, wherein, when a The customer must have at least a predetermined number of the transaction records in the manual case review database, and then the customer's transaction record will be brought into the inference module; the inference module will match the ratio of money laundering cases or the risk score to A predetermined value is compared; when the proportion of the money laundering cases of a customer in the manual case database exceeds 50%, or the risk level is high risk, or the risk score exceeds the risk threshold, it is judged as a Suspected case, the speculation module transmits the suspected case, the money laundering case ratio, the risk score and the review result to a review module; and the review module reviews the suspected case by the reviewer and updates the review The result is sent to the manual case review database for storage. The method for assisting judgment in a suspected case of money laundering prevention law as described in claim 4, wherein the transaction data includes a client's professional information and a transaction record, and the client's professional information is classified according to political figures or regional attributes; and the transaction record can be Categorize multiple accounts by cash flow and the same account.

Images