CARD TRANSACTION FRAUD DETECTION SYSTEM

Abstract

This invention relates to a system (1) that enables the identification of potential fraudulent transactions by analyzing the customer’s past spending, financial status and card usage habits in merchants where numerous card payment transactions are carried out without additional security verification in the banking and electronic payment field.

Description

1 TARIFF CARD TRANSACTION FRAUD DETECTION SYSTEM Technical Area This invention has implications for banking and electronic payments, involving numerous card payments. Additional security verification at merchant locations where the transaction is carried out. transactions made without using the service, customer's past expenses, financial status and by analyzing card usage habits, potential fraudulent transactions can be identified. It is related to a system that enables its determination. 10 Previous Technique Today, card payment transactions involve analyzing transaction data and customer history. This allows for the identification of transactions that carry a risk of fraud. 15 Current systems involve transaction rules, customer confirmation, analyst review, and risk management. Scores are used. However, high-volume and secure verification is required. In large merchant locations that accept transactions without implementing nonsecure, the actual It is becoming difficult to distinguish between legitimate transactions and fraudulent attempts. Therefore... The current rules are also considered risky as a result of the assessment of actual operations, resulting in 20 incorrect decisions. Deficiencies include alarms, unnecessary customer confirmations, and decreased rule efficiency. It is emerging. Therefore, considering the studies and shortcomings in the current technique... When this is taken into account, customers' past spending habits, payment 25 by analyzing their behavior, banking product usage, and transaction history a payment transaction being carried out is related to the customer's previous financial behavior a system is needed to determine whether it is compliant or not It is understood. 2 United States Regulation US2025190992A1, which is included in the known state of the art. The patent document describes payments by estimating the risk of fraud. This refers to a system that manages transactions according to risk level. The subject of the invention is the use of historical accountant data and credit account data. It includes a trained machine learning model. The machine learning model is 5 The system consists of gradient-augmented decision trees. The system is based on credit account transactions. It retrieves data related to an authorized payment transaction. Then it sends it to the account holder. and historical data related to the relevant credit account are requested from the database. Historical data and current payment transaction data are combined for machine learning. This information is transferred to the model. The model uses this information to generate a 10 for the payment process. It generates a fraud prediction score. The generated score determines when the payment will be returned in the future. This indicates the possibility of withdrawal or invalidation. The system obtained fraud prediction score with a predetermined threshold value It compares them. If the risk score is low, a short payment is made. A waiting period is determined. If the risk score is high, it may be extended to 15 days. A long payment waiting period applies. During this period, the payment is expected to be processed. The use of the accumulated credit balance may be restricted. Brief Description of the Invention The aim of this invention is to provide additional security in the banking and electronic payments sectors. In card payment transactions made without using verification, the customer past transaction habits, financial status, use of cards and banking products. By analyzing security behaviors based on this information, fraud can be detected for each transaction. (fraud) risk is indicated by a score, and that score is used to calculate the current fraud risk. 25 Using prevention rules as additional decision input to compare real transactions with potential identifying fraudulent attempts, rejecting unnecessary transactions, and customer a system that reduces contact and lowers financial losses to accomplish. 3 Detailed Description of the Invention The "Card Transaction Fraud" investigation was carried out to achieve the purpose of this invention. The "Detection System" is shown in the attached diagram; Figure 1 shows a schematic view of the system that is the subject of the invention. The parts shown in the figure are individually numbered, and the corresponding numbers correspond to these numbers. given below: 1. System 2. Electronic device 3. Database 4. Server In the banking and electronic payments sector, numerous card payment transactions... carried out at member businesses without using additional security verification transactions, customer's past spending, financial status and card usage Identifying potential fraudulent activities by analyzing their habits The system that provides the subject of the invention (1); 20 - the result of the security scan of the customer's card payment transactions the receipt of the transaction amount, merchant information, and the generated fraud risk score. This process displays the evaluation information to the user. at least one electronic device configured to provide at least one interface that provides 25 (2), - customer's age, relationship with the bank and credit card usage duration, transaction volume, account balance, credit card limit, statement and payment information, demand deposits transactions, banking products, preferred channel usage rate, card security features, statement and income / expense viewing numbers, password and 30 Security settings usage related to merchant, merchant group and merchant 4 To store past transaction count, amount, and rejection information in the category code. at least one database structured (3) and - using any remote communication protocol with the electronic device (2) to communicate with customer and past transaction data obtained from the database (3) Analyzing data related to card payment transactions, determining the transaction amount is the customer's 5 Determining the ratios to past financial and transaction values, fraud. Effective variables in decomposing processes using supervised machine learning to detect with the algorithm, with different combinations of these variables By comparing the classification models created, those with high discrimination power can be identified. Choosing a model indicates the likelihood of fraud with the chosen model (10). to generate a score and use that score as additional decision input in fraud prevention rules It includes at least one server (4) configured for use as such. The electronic device (2) in the system (1) which is the subject of the invention, any remote communication to communicate with the server (4) using the protocol and this established communication 15 smartphones and tablets configured to exchange data via A computer is a device in the form of a desktop computer or a laptop computer. Electronic device (2), card verification performed without additional security verification. the transaction amount for the payment transaction, the merchant where the transaction was made, and the customer. The security scan result used in assessing the security situation is 20. It is structured to ensure that the electronic device (2), transaction amount, merchant location where the transaction was carried out, transaction time, and security scan result. transmitting the transaction and security information in this form to the server (4), by the server (4) Customer's past expenses, transaction frequency, balance, debt recorded in the database (3), credit card limit, payment habits, past transactions at the same merchant, and 25 Fraud risk is determined as a result of analyzing security behaviors. to obtain the score and the transaction amount, merchant information and fraud risk to provide at least one interface that allows the score to be displayed It is being structured. 5 The database (3) in the system (1) which is the subject of the invention, the customer's age, relationship with the bank duration, time elapsed since the first credit card was issued, and professional information. Customer profile data such as account balance, credit card limit, and outstanding balance, outstanding balance, payment and statement information, demand deposit transactions, assets Financial data in the form of values ​​and banking products owned; preference 5 the usage rate of the chosen banking channel, blocking the card for international transactions, the number of times the statement and income-expense menus are viewed, and password changes. to store behavioral data regarding the use of security settings is being structured. The database (3) contains the customer's last 30, 90 and 180 days Information on the number and amount of transactions carried out during the periods, by the relevant merchant, 10 member merchant group and member merchant category code in the last one, three and six months the number and amounts of transactions carried out, rejected at large merchant locations to keep records of transaction numbers and domestic spending amounts is structured. The database (3) contains the transaction amount of the customer's past transaction amount. averages, asset values, balance, debt amount, credit card limit and related member 15 historical values ​​to be used in the comparison with past transaction amounts in the workplace is configured to store in a way that is accessible by the server (4). In the system that is the subject of the invention, the server (4) located in (1) communicates with the electronic device (2) to establish and exchange data with electronic devices (2) through this established communication 20 It is configured to perform. Server (4), electronic device (2) transaction amount, merchant information, transaction time, and security scan result. Transaction and security data in the form of (3) customer profiles registered in the database, Adding security by gathering financial status, past transaction and banking usage data. Card payment transactions made without verification are considered fraud. 25 It is structured to enable evaluation in terms of the server (4), customer's age, relationship with the bank and credit card usage duration, transaction numbers, balance, Debt, credit card limit, payment and statement information, demand deposit transactions, banking products used, merchant information regarding the use of security settings, Member business group and MCC (Merchant Category Code) 30 to analyze the number and amount of past transactions within the scope; current transaction amount 6 the customer's average transactions over the last 30, 90 and 180 days, domestic spending, assets values, balance, debt, credit card limit, and past transactions at the same merchant. by proportionally comparing the current transaction to the customer's past transaction profile. to create the variables to be used in the evaluation is configured. Server (4), variables supervised machine learning 5 (Supervised Machine Learning) based Analyzing victims of fraud using a classification algorithm Which variables are more important in differentiating customers from other customers? to determine that it has weight and separating power and the determined Variables differentiate customers in different transaction types where fraud is observed. 10 It is structured to evaluate the success of doing. The server (4) is configured to evaluate the success of doing. Multiple classifications can be achieved by using different combinations of variables. creating models, models of customers' security behaviors and transactions Comparing their separation performance based on their profiles reveals which separation power is stronger. Selecting the model with the highest rating and examining the nonsecure 15 based on the selected model. An indicator of the possibility of fraud for an (insecure) card payment transaction. It is configured to generate a risk score. The server (4) is created Fraud risk score is available in real-time transaction review processes. Using real customers as additional decision input to fraud prevention rules 20 to provide and monitor changing customer behaviors within the scope of model tracking studies. Adding variables used in line with new banking products, to enable model updating by removing or modifying it It is being structured. Industrial Application of the Invention Thanks to the system (1) that is the subject of the invention, in the banking and electronic payment sector, Members who conduct numerous card transactions, such as those on Google and Facebook. 3D Secure (Three-Domain Secure) in workplaces, additional 30 Payments made without using verification, customers' past spending, 7 Account and card information, payment habits, and security behaviors are collected by assessment in terms of fraud risk and the risk scores created by identifying suspicious transactions and increasing payment security Its use is ensured. Based on these fundamental concepts, the invention focuses on "Card Transaction Fraud Detection". It is possible to develop a wide variety of applications related to the System (1)”, and the invention This cannot be limited to the examples described here, but is primarily stated in the claims. It is like that.

Claims

8 REQUESTS 1. In the banking and electronic payment sector, a large number of card payment transactions without using additional security verification at member businesses where it is carried out transactions made, customer's past spending, financial status and card 5 by analyzing usage habits, potential fraudulent activities can be identified. enabling its determination; - the result of the security scan of the customer's card payment transactions the receipt of the transaction amount, merchant information, and the generated fraud risk score. The evaluation information is displayed to the user through the process described above. 10 at least one electronic device configured to provide at least one interface that enables (2), - customer's age, relationship with the bank and credit card usage duration, transaction volume, account balance, credit card limit, statement and payment information, demand deposits transactions, banking products, preferred channel usage rate, card 15 security features, statement and income / expense viewing numbers, password and Security settings usage related to merchant, merchant group and merchant To store past transaction count, amount, and rejection information in the category code. containing at least one structured database (3) and − using any remote communication protocol with electronic device (2) 20 to communicate with customer and past transaction data obtained from the database (3) Analyzing data related to card payment transactions, determining the transaction amount for the customer Determining the ratios to past financial and transaction values, fraud. Effective variables in decomposing processes using supervised machine learning to detect with the algorithm, 25 with different combinations of these variables By comparing the classification models created, those with high discrimination power can be identified. Choosing a model indicates the likelihood of fraud associated with the chosen model. to generate a score and use that score as additional decision input in fraud prevention rules characterized by at least one server (4) configured to be used as a system (1). 30 9 2. Communicate with the server (4) using any remote communication protocol. to establish and exchange data through this established communication configured smartphone, tablet computer, desktop computer or electronic device (2) characterized by a portable computer-like device A system like the one in Request 1 (1). 5 3. Card payment transactions made without using additional security verification. the transaction amount, the merchant's information where the transaction was carried out, and the customer's security information. the result of the security scan used in assessing the situation 10 characterized by an electronic device (2) configured to enable its acquisition. A system like the one in Request 1 or 2 (1).

4. Transaction amount, merchant location, transaction time, and security. transmit the process and security information in the form of the scan result to the server (4), The server (4) records the customer's past expenses (3) in the database, 15 transaction frequency, balance, debt, credit card limit, payment habits, same member analyzing past workplace transactions and safety behaviors to obtain the resulting fraud risk score and the transaction in question the amount, merchant information and fraud risk score are displayed. electronic device (2) configured to provide at least one interface that provides 20 a system like any of the above characterized demands (1).

5. Customer's age, length of relationship with the bank, since the issuance of the first credit card. Customer profile data such as elapsed time and professional information; account balance, credit card limit, due date, carryover balance, payment and statement information, 25 demand deposit movements, asset values, and banking holdings. Financial data in the form of products; use of the preferred banking channel rate, blocking international transactions on the card, statement and income-expense menus view counts, password changes, and security settings usage. database structured to store related behavioral data (3) and 30 a system like any of the above characterized demands (1).

6. Transactions made by the customer in the last 30, 90 and 180 days. Quantity and amount information, relevant merchant, merchant group and merchant category. the number of transactions performed in the code over the last one, three and six months their amounts, the number of rejected transactions at major merchant locations, and domestically database structured to keep records of expenditure amounts (3) and 5 a system like any of the above characterized demands (1).

7. Transaction amount based on the customer's past transaction averages, asset values, balance, debt amount, credit card limit, and past transactions at the relevant merchant. The historical values ​​to be used in the ratio of the amounts are provided by the server (4) 10 characterized by the database (3) structured to store accessible data. a system like any of the above-mentioned requests (1).

8. To communicate with the electronic device (2) and through this communication 15 configured to exchange data with electronic devices (2) as in any of the above requests characterized by the server (4) a system (1).

9. Transaction amount received from the electronic device (2), merchant information, transaction time and transaction and security data in the form of security scan results and data 20 (3) registered customer profile, financial status, past transactions and banking By collecting usage data, without using additional security verification. card payment transactions in terms of fraud characterized by the server (4) configured to enable its evaluation. a system like any of the above requests (1). 25 10. Customer's age, relationship with the bank and credit card usage duration, transaction volume, Balance, debt, credit card limit, payment and statement information, demand deposits transactions, banking products used, and the use of security settings The relevant merchant, merchant group, and past transaction count within the scope of MCC and 30 to analyze the amounts; the current transaction amount of the customer in the last 30, 90 and 180 11 daily transaction averages, domestic spending, asset values, balance, debt, by proportionally comparing it to your credit card limit and past transaction amounts at the same merchant. the current transaction is evaluated based on the customer's past transaction profile the server (4) configured to create the variables to be used A system like any of the above-mentioned requirements characterized (1). 5 11. Classification algorithm based on supervised machine learning for variables. by using this analysis, customers who have been subjected to fraud and others Which variables have higher weight in customer segmentation? to determine that it has discriminatory power and that the determined variables are 10 Identifying customers in different transaction types where fraud occurs characterized by the server (4) configured to evaluate its success a system like any of the above requests (1).

12. By using the specified variables in different combinations, you can create more than one 15 Creating a classification model, analyzing customer security behaviors using these models. and to compare parsing performance based on process profiles, parsing Selecting the model with higher power and examining the selected model. A risk indicating the possibility of fraud for non-secure card payment transactions. The above 20 is characterized by the server (4) configured to produce the score. a system like any of the requests (1).

13. Real-time transaction review of the generated fraud risk score. as additional decision input to existing fraud prevention rules in their processes using real customer transactions to effectively detect potential fraud attempts 25 to ensure its separation in this way and within the scope of model monitoring studies in line with changing customer behaviors and new banking products by adding, removing or modifying the variables used with the server (4) configured to enable model updating A system like any of the above-mentioned demands characterized (1). 30

Images