A method and system for analyzing bank retail customer fund behavior

Abstract

The present invention relates to the field of fund management technology, and in particular to a method and system for analyzing fund behavior of retail bank customers, comprising a fund supervision platform, a transaction supervision unit, a behavior assessment unit, a transaction risk unit, a fund tracking unit and a management early warning unit; the present invention analyzes historical transaction flow data of customers to preliminarily analyze whether the fund transaction behavior of target customers is abnormal, and conducts behavior trend security supervision analysis through information feedback to determine whether the target customer's transaction activity is abnormal, so as to improve the security of the target customer's fund transaction. Under the premise that the customer's fund transaction is abnormally active, the method analyzes from two perspectives, namely, the customer's transaction performance and the transaction object, to understand the customer's fund transaction risk situation, so as to reduce the target customer's fund transaction risk, which is helpful to protect the target customer's funds and prevent economic transaction losses.

Description

Technical Field

[0001] The present invention relates to the technical field of fund management, and in particular to a method and system for analyzing fund behavior of retail bank customers. Background Art

[0002] With the rapid development of Internet technologies, various online banking transaction methods have emerged, and personal money transactions have become faster and easier. The resulting transaction data has become more and more as time goes by.

[0003] Bank customer flow refers to the detailed record of fund flow in customer accounts recorded by the bank. However, in the existing analysis methods and technologies, the analysis methods are usually too simplified and limited, and cannot accurately monitor customer transaction flow, which leads to increased risk of abnormal customer transactions, and cannot conduct in-depth security transaction supervision based on customer transaction behavior and transaction objects, and thus cannot timely protect target customer funds;

[0004] In view of the above technical defects, a solution is now proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for analyzing the fund behavior of retail bank customers to solve the technical defects mentioned above. The present invention analyzes the historical transaction flow data of customers to preliminarily analyze whether the fund transaction behavior of target customers is abnormal, that is, to perform transaction behavior supervision analysis on the historical transaction flow data to understand the customer's transaction flow activity, and to perform behavior trend security supervision analysis based on information feedback to determine whether the target customer's transaction activity is abnormal, so as to further perform security supervision on the target customer's fund transactions to improve the security of the target customer's fund transactions. Under the premise that the customer's fund transactions are abnormally active, the present invention analyzes from two perspectives: the customer's transaction performance and the transaction object to understand the customer's fund transaction risk situation, so as to reduce the target customer's fund transaction risk, which helps to protect the target customer's funds and prevent economic transaction losses.

[0006] The object of the present invention can be achieved by the following technical solutions: A bank retail customer fund behavior analysis system, including a fund supervision platform, a transaction supervision unit, a behavior assessment unit, a transaction risk unit, a fund tracking unit and a management early warning unit;

[0007] The transaction monitoring unit is used to collect historical transaction flow data of the target customer, perform transaction behavior monitoring analysis on the historical transaction flow data, and send the obtained active signal to the behavior evaluation unit;

[0008] After receiving the active signal, the behavior assessment unit immediately conducts a behavior trend security supervision analysis and sends the obtained abnormal active signal to the transaction risk unit and the fund tracking unit;

[0009] After receiving the abnormal activity signal, the transaction risk unit immediately collects the target customer's transaction performance data, which represents the transaction risk value and capital flow risk value, and conducts transaction risk safety supervision feedback analysis on the transaction performance data, and sends the obtained warning signals and risk signals to the management warning unit;

[0010] After receiving the abnormally active signal, the fund tracking unit immediately collects the target customer's transaction object information, conducts a mesh transaction security supervision analysis on the transaction object information, and sends the obtained suspicious signal to the management warning unit.

[0011] Preferably, the transaction behavior supervision and analysis process of the transaction supervision unit is as follows:

[0012] A monitoring cycle is set as a time threshold, and historical transaction flow data of target customers within the time threshold is obtained. The historical transaction flow data includes daily consumption, deposits, withdrawals, transfers, online banking transactions, and credit card transactions. The number of fund transfers out and the number of fund transfers in of the target customers within the time threshold are obtained from the historical transaction flow data. The ratio between the number of fund transfers out and the number of fund transfers in of the target customers within the time threshold is set as the transaction propensity value. The average interval duration of the number of fund transfers out and the average interval duration of the number of fund transfers in of the target customers within the time threshold are obtained from the historical transaction flow data. The product of the average interval duration of the number of fund transfers out and the average interval duration of the number of fund transfers in of the target customers within the time threshold is set as the flow transaction value;

[0013] Compare and analyze the transaction tendency value and flow transaction value with the preset transaction tendency value threshold and preset flow transaction value threshold that are stored internally:

[0014] If the trading propensity value is less than the preset trading propensity value threshold, and the flow trading value is less than the preset flow trading value threshold, no signal is generated;

[0015] If the transaction propensity value is greater than or equal to a preset transaction propensity value threshold, or the flow transaction value is greater than or equal to a preset flow transaction value threshold, an active signal is generated.

[0016] Preferably, the trend safety supervision analysis process of the behavior assessment unit is as follows:

[0017] The transaction propensity value and flow transaction value of the target customer within the time threshold are obtained, and then the transaction propensity value and flow transaction value of the target customer within the historical j time thresholds are obtained, j is a natural number greater than zero, and a rectangular coordinate system is established with the number of time thresholds as the X-axis and the transaction propensity value and flow transaction value as the Y-axis respectively. The transaction propensity value curve and the flow transaction value curve are drawn by plotting points respectively, and the transaction propensity value and flow transaction value corresponding to the transaction propensity value curve and the flow transaction value curve are collected once every t1 time, and the transaction propensity value and flow transaction value corresponding to the transaction propensity value curve and the flow transaction value curve are collected once every t2 time, and the transaction propensity value and flow transaction value corresponding to the transaction propensity value curve and the flow transaction value curve are collected once every tg time, g is a natural number greater than zero, t1, t2...tg are not equal, and the sets A and B of the transaction propensity value and the flow transaction value are respectively constructed, and then the discrete coefficients of the sets A and B are obtained, and they are set as the transaction fluctuation value and the flow floating value respectively;

[0018] Compare and analyze the transaction fluctuation value and liquidity floating value with the preset transaction fluctuation value threshold and preset liquidity floating value threshold that are stored internally:

[0019] If the trading volatility value is less than the preset trading volatility value threshold, and the liquidity float value is less than the preset liquidity float value threshold, no signal is generated;

[0020] If the transaction fluctuation value is greater than or equal to the preset transaction fluctuation value threshold, or the liquidity floating value is greater than or equal to the preset liquidity floating value threshold, an abnormal activity signal is generated.

[0021] Preferably, the transaction risk safety supervision feedback analysis process of the transaction risk unit is as follows:

[0022] S1: Obtain the transaction risk value of the target customer within the time threshold from the historical transaction flow data. The transaction risk value represents the product of the number of consecutive transactions of the same amount and the number of occurrences of the corresponding transaction objects of the same amount after data normalization, and the ratio of the product of the number of consecutive transactions of the same amount and the average interval time after data normalization;

[0023] S2: At the same time, the target customer's capital flow risk value within the time threshold is obtained from the historical transaction flow data. The capital flow risk value indicates the proportion of the number of times without filling in the transaction remarks in the total number of times the transaction amount exceeds the preset transaction amount threshold. The transaction risk value and the capital flow risk value are marked as JF and ZL respectively;

[0024] S3: Obtain the transaction behavior risk coefficient F according to the formula, obtain the transaction behavior risk coefficient of the target customer within the historical j time thresholds, obtain the maximum value of the transaction behavior risk coefficient of the target customer within the historical j time thresholds, and set it as the transaction behavior risk critical value, and compare and analyze the transaction behavior risk coefficient F with the transaction behavior risk critical value:

[0025] If the ratio between the trading behavior risk coefficient F and the trading behavior risk threshold is less than 1, a feedback instruction is generated;

[0026] If the ratio between the trading behavior risk coefficient F and the trading behavior risk critical value is greater than or equal to 1, a risk signal is generated.

[0027] Preferably, when the transaction risk unit generates a feedback instruction:

[0028] The target customer's transaction period information is obtained from the historical transaction flow data. Based on the target customer's transaction period information, the target customer's regular behavior characteristics are obtained. The difference between the regular behavior characteristics and the set regular behavior characteristics is obtained and set as the behavior difference degree. The behavior difference degree is compared and analyzed with the preset behavior difference degree threshold value stored in the internal data entry:

[0029] If the behavior difference is less than the preset behavior difference threshold, no signal is generated;

[0030] If the behavior difference is greater than or equal to the preset behavior difference threshold, an early warning signal is generated.

[0031] Preferably, the mesh transaction security supervision and analysis process of the funds tracking unit is as follows:

[0032] The transaction object information of the target customer within the time threshold is obtained from the historical transaction flow data. The transaction object information includes the number of unfamiliar transaction objects and the number of known transaction objects. The unfamiliar transaction objects indicate that they do not appear in the historical transaction list, and the known transaction objects indicate that they appear in the historical transaction list. The associated transaction value between the unfamiliar transaction objects and the known transaction objects is obtained. The associated transaction value indicates the number of unfamiliar transaction objects corresponding to the transaction behaviors between the unfamiliar transaction objects and the known transaction objects, and the associated transaction value is set as the associated value. The value obtained by subtracting the associated value from the number of unfamiliar transaction objects is set as the unfamiliar risk value. The sum of the associated value and the number of known transaction objects is set as the network flow value. The ratio between the unfamiliar risk value and the network flow value is set as the capital transaction value.

[0033] Compare and analyze the funds transaction value with the stored preset funds transaction value threshold:

[0034] If the fund turnover value is less than the preset fund turnover value threshold, no signal is generated;

[0035] If the fund transaction value is greater than or equal to the preset fund transaction value threshold, a suspicious signal is generated.

[0036] The beneficial effects of the present invention are as follows:

[0037] The present invention analyzes the historical transaction flow data of the customer to preliminarily analyze whether the target customer's fund transaction behavior is abnormal, that is, the historical transaction flow data is subjected to transaction behavior supervision analysis to understand the customer's transaction flow activity, and the behavior trend security supervision analysis is performed based on the information feedback to determine whether the target customer's transaction activity is abnormal, so as to further conduct security supervision on the target customer's fund transaction to improve the security of the target customer's fund transaction;

[0038] Under the premise that customer fund transactions are abnormally active, the present invention analyzes from two perspectives, namely, the customer's transaction performance and transaction object, to understand the customer fund transaction risk situation. That is, from the perspective of transaction performance, the analysis is carried out, and transaction risk security supervision feedback analysis is performed on the transaction performance data to determine whether the transaction risk of the target customer is too high, so as to carry out early warning feedback management in a timely manner to reduce the target customer's fund transaction risk and help protect the target customer's funds. From the perspective of transaction objects, the analysis is carried out, and a meshed transaction security supervision analysis is carried out on the transaction object information to determine whether the target customer's transaction object has suspicious transaction behavior, so as to carry out timely early warning supervision and improve the transaction security of the target customer. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The present invention will be further described below in conjunction with the accompanying drawings;

[0040] Figure 1 is a reference diagram of the method of the present invention;

[0041] Figure 2 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] Embodiment 1:

[0044] See also Figure 1 to Figure 2 As shown, the present invention is a method for analyzing the fund behavior of retail bank customers, comprising the following steps:

[0045] Step 1: Collect the historical transaction flow data of the target customer and conduct transaction behavior supervision and analysis on the historical transaction flow data to preliminarily analyze whether the target customer's fund transaction behavior is abnormal. If an active signal is obtained, proceed to step 2;

[0046] Step 2: Conduct a behavior trend security supervision analysis in a progressive manner to determine whether the target customer's transaction activity is abnormal. If an abnormal activity signal is obtained, proceed to Step 3 and Step 5;

[0047] Step 3: Conduct transaction risk safety supervision feedback analysis on the target customer’s transaction performance data based on information feedback to determine whether the target customer’s transaction risk is too high. If feedback instructions are received, proceed to step 4. If risk signals are received, early warning feedback is given;

[0048] Step 4: Perform a difference analysis on the regular behavior characteristics of the benchmark customers in an in-depth manner to obtain the degree of behavioral difference, and then perform a discriminant analysis on the degree of behavioral difference. If a warning signal is obtained, a warning feedback is given;

[0049] Step 5: Conduct a mesh transaction security supervision analysis on the target customer’s transaction object information based on information feedback to determine whether the target customer’s transaction object has any suspicious transaction behavior. If a suspicious signal is obtained, an early warning feedback will be issued.

[0050] Embodiment 2:

[0051] A bank retail customer fund behavior analysis system includes a fund supervision platform, a transaction supervision unit, a behavior assessment unit, a transaction risk unit, a fund tracking unit and a management warning unit, wherein the fund supervision platform is connected to the transaction supervision unit in a one-way communication, the transaction supervision unit is connected to the behavior assessment unit in a one-way communication, the behavior assessment unit is connected to the transaction risk unit and the fund tracking unit in a one-way communication, and the transaction risk unit and the fund tracking unit are connected to the management warning unit in a one-way communication;

[0052] When the fund supervision platform generates a supervision instruction, it sends the supervision instruction to the transaction supervision unit. After receiving the supervision instruction, the transaction supervision unit immediately collects the historical transaction flow data of the target customer and conducts transaction behavior supervision analysis on the historical transaction flow data, so as to preliminarily analyze whether the target customer's fund transaction behavior is abnormal, so as to provide timely feedback management. The specific transaction behavior supervision analysis process is as follows:

[0053] Set a monitoring cycle and set it as a time threshold, obtain the historical transaction flow data of the target customer within the time threshold, the historical transaction flow data includes daily consumption, deposits, withdrawals, transfers, online banking transactions, credit card transactions, etc., obtain the number of fund transfers out and the number of fund transfers in of the target customer within the time threshold from the historical transaction flow data, set the ratio between the number of fund transfers out and the number of fund transfers in of the target customer as the transaction propensity value, obtain the average interval duration of the number of fund transfers out and the average interval duration of the number of fund transfers in of the target customer within the time threshold from the historical transaction flow data, set the product of the average interval duration of the number of fund transfers out and the average interval duration of the number of fund transfers in of the target customer as the flow transaction value. It should be noted that the transaction propensity value and the flow transaction value are two influencing parameters that reflect the transaction characteristics of the target customer. The larger the values ​​of the transaction propensity value and the flow transaction value, the higher the transaction supervision demand of the target customer;

[0054] Compare and analyze the transaction tendency value and flow transaction value with the preset transaction tendency value threshold and preset flow transaction value threshold that are stored internally:

[0055] If the trading propensity value is less than the preset trading propensity value threshold, and the flow trading value is less than the preset flow trading value threshold, no signal is generated;

[0056] If the transaction tendency value is greater than or equal to a preset transaction tendency value threshold, or the flow transaction value is greater than or equal to a preset flow transaction value threshold, an active signal is generated and sent to the behavior evaluation unit;

[0057] After receiving the active signal, the behavior assessment unit immediately conducts a behavior trend security supervision analysis to determine whether the target customer's transaction activity is abnormal, so as to further conduct security supervision on the target customer's fund transactions. The specific trend security supervision analysis process is as follows:

[0058] The transaction tendency value and flow transaction value of the target customer within the time threshold are obtained, and then the transaction tendency value and flow transaction value of the target customer within the historical j time thresholds are obtained, where j is a natural number greater than zero. A rectangular coordinate system is established with the number of time thresholds as the X-axis and the transaction tendency value and flow transaction value as the Y-axis respectively. The transaction tendency value curve and flow transaction value curve are drawn by plotting points respectively. The transaction tendency value and flow transaction value corresponding to the transaction tendency value curve and flow transaction value curve are collected once every t1 time length, and the transaction tendency value and flow transaction value corresponding to the transaction tendency value curve and flow transaction value curve are collected once every t2 time length, and so on. Every tg The transaction tendency value and flow transaction value corresponding to the transaction tendency value curve and flow transaction value curve are collected once in a long time. g is a natural number greater than zero. t1, t2…tg gradually increase, and t1, t2…tg are not equal. Sets A and B of transaction tendency values ​​and flow transaction values ​​are constructed respectively, and then the discrete coefficients of sets A and B are obtained, and they are set as transaction fluctuation value and flow floating value respectively. It should be noted that transaction fluctuation value and flow floating value are two influencing parameters that reflect the abnormality of historical transaction flow data of target customers. The larger the values ​​of transaction fluctuation value and flow floating value, the greater the risk of abnormal activity of historical transaction flow data of target customers.

[0059] Compare and analyze the transaction fluctuation value and liquidity floating value with the preset transaction fluctuation value threshold and preset liquidity floating value threshold that are stored internally:

[0060] If the trading volatility value is less than the preset trading volatility value threshold, and the liquidity float value is less than the preset liquidity float value threshold, no signal is generated;

[0061] If the transaction fluctuation value is greater than or equal to the preset transaction fluctuation value threshold, or the liquidity floating value is greater than or equal to the preset liquidity floating value threshold, an abnormal activity signal is generated and sent to the transaction risk unit and the funds tracking unit.

[0062] Embodiment 3:

[0063] After receiving the abnormal activity signal, the transaction risk unit immediately collects the target customer's transaction performance data, which represents the transaction risk value and the capital flow risk value, and conducts transaction risk safety supervision feedback analysis on the transaction performance data to determine whether the target customer's transaction risk is too high, so as to conduct early warning feedback management in a timely manner to reduce the target customer's capital transaction risk and help protect the target customer's funds. The specific transaction risk safety supervision feedback analysis process is as follows:

[0064] The transaction risk value of the target customer within the time threshold is obtained from the historical transaction flow data. The transaction risk value represents the product of the number of consecutive transactions of the same amount and the number of occurrences of the transaction object corresponding to the same amount of consecutive transactions after data normalization, and the ratio obtained by normalizing the average interval time. It should be noted that the larger the value of the transaction risk value, the higher the transaction risk;

[0065] At the same time, the target customer's capital flow risk value within the time threshold is obtained from the historical transaction flow data. The capital flow risk value indicates the proportion of the number of times that the transaction amount exceeds the preset transaction amount threshold and the number of times the transaction remarks are not filled in. It should be noted that the larger the value of the capital flow risk value, the higher the transaction risk. The transaction risk value and the capital flow risk value are marked as JF and ZL respectively; according to the formula The trading behavior risk coefficient is obtained, where a1 and a2 are the preset proportional factor coefficients of the trading risk value and the capital flow risk value respectively. The proportional factor coefficient is used to correct the deviation of various parameters in the formula calculation process, so as to make the calculation result more accurate. A3 is the preset correction factor coefficient. A1, a2 and a3 are all greater than zero. F is the trading behavior risk coefficient;

[0066] Obtain the target customer's transaction behavior risk coefficient within the historical j time thresholds, obtain the maximum value of the target customer's transaction behavior risk coefficient within the historical j time thresholds, and set it as the transaction behavior risk critical value, and compare and analyze the transaction behavior risk coefficient F with the transaction behavior risk critical value:

[0067] If the ratio between the transaction behavior risk coefficient F and the transaction behavior risk critical value is less than 1, a feedback instruction is generated. When the feedback instruction is generated, the target customer's transaction period information is obtained from the historical transaction flow data, and the target customer's regular behavior characteristics are obtained according to the target customer's transaction period information. The difference between the regular behavior characteristics and the set regular behavior characteristics is obtained and set as the behavior difference degree. The behavior difference degree is compared and analyzed with the preset behavior difference degree threshold value stored in the internal data entry:

[0068] If the behavior difference is less than the preset behavior difference threshold, no signal is generated;

[0069] If the behavior difference is greater than or equal to the preset behavior difference threshold, a warning signal is generated and sent to the management warning unit;

[0070] If the ratio between the transaction behavior risk coefficient F and the transaction behavior risk threshold is greater than or equal to 1, a risk signal is generated and sent to the management warning unit. After receiving the risk signal and the warning signal, the management warning unit immediately performs the preset warning operations corresponding to the risk signal and the warning signal, which helps to improve the target customer's fund security and prevent economic transaction losses, thereby achieving the purpose of reducing the target customer's transaction risk;

[0071] After receiving the abnormally active signal, the fund tracking unit immediately collects the target customer's transaction object information and conducts a mesh transaction security supervision analysis on the transaction object information to determine whether the target customer's transaction object has any suspicious transaction behavior, so as to provide timely early warning supervision to improve the target customer's transaction security. The specific mesh transaction security supervision analysis process is as follows:

[0072] The transaction object information of the target customer within the time threshold is obtained from the historical transaction flow data. The transaction object information includes the number of unfamiliar transaction objects and the number of known transaction objects. The unfamiliar transaction objects indicate that they do not appear in the historical transaction list, and the known transaction objects indicate that they appear in the historical transaction list. The associated transaction value between the unfamiliar transaction objects and the known transaction objects is obtained. The associated transaction value indicates the number of unfamiliar transaction objects corresponding to the transaction behaviors between the unfamiliar transaction objects and the known transaction objects, and the associated transaction value is set as the associated value. The value obtained by subtracting the associated value from the number of unfamiliar transaction objects is set as the unfamiliar risk value. The sum of the associated value and the number of known transaction objects is set as the network flow value. The ratio between the unfamiliar risk value and the network flow value is set as the capital transaction value. It should be noted that the larger the value of the capital transaction value is, the greater the transaction risk of the target customer is.

[0073] Compare and analyze the funds transaction value with the stored preset funds transaction value threshold:

[0074] If the fund turnover value is less than the preset fund turnover value threshold, no signal is generated;

[0075] If the fund transaction value is greater than or equal to the preset fund transaction value threshold, a suspicious signal is generated and sent to the management warning unit. After receiving the suspicious signal, the management warning unit immediately performs the preset warning operation corresponding to the suspicious signal, which helps to improve the fund security of the target customer;

[0076] To summarize, the present invention analyzes the historical transaction flow data of the customer to preliminarily analyze whether the target customer's fund transaction behavior is abnormal, that is, conducts transaction behavior supervision analysis on the historical transaction flow data to understand the customer's transaction flow activity, and conducts behavior trend security supervision analysis through information feedback to determine whether the target customer's transaction activity is abnormal, so as to further conduct security supervision on the target customer's fund transaction. On the premise that the customer's fund transaction is abnormally active, the present invention analyzes from the two perspectives of the customer's transaction performance and transaction object to understand the customer's fund transaction risk, that is, analyzes from the perspective of transaction performance, conducts transaction risk security supervision feedback analysis on the transaction performance data to determine whether the target customer's transaction risk is too high, so as to conduct early warning feedback management in a timely manner to reduce the target customer's fund transaction risk, which is helpful to protect the target customer's funds. Analyze from the perspective of transaction objects, conduct meshed transaction security supervision analysis on transaction object information to determine whether the target customer's transaction object has suspicious transaction behavior, so as to conduct timely early warning supervision to improve the target customer's transaction security.

[0077] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by technicians in this field for each group of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0078] The above formulas are obtained by collecting a large amount of data for software simulation and selecting a formula that is close to the actual value. The coefficients in the formula are set by technical personnel in this field according to actual conditions. The above is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited to this. Any technical personnel familiar with the technical field within the technical scope disclosed by the present invention, according to the technical solution and the inventive concept of the present invention, make equivalent replacement or change, which should be covered within the protection scope of the present invention.

Claims

1. A method for analyzing the fund behavior of retail bank customers, characterized in that: The following steps are involved: Step 1: Collect the historical transaction flow data of the target customer and conduct transaction behavior supervision and analysis on the historical transaction flow data to preliminarily analyze whether the target customer's fund transaction behavior is abnormal. If an active signal is obtained, proceed to step 2; Step 2: Conduct a behavior trend security supervision analysis in a progressive manner to determine whether the target customer's transaction activity is abnormal. If an abnormal activity signal is obtained, proceed to Step 3 and Step 5; Step 3: Conduct transaction risk safety supervision feedback analysis on the target customer’s transaction performance data based on information feedback to determine whether the target customer’s transaction risk is too high. If feedback instructions are received, proceed to step 4. If risk signals are received, early warning feedback is given; Step 4: Perform a difference analysis on the regular behavior characteristics of the benchmark customers in an in-depth manner to obtain the degree of behavioral difference, and then perform a discriminant analysis on the degree of behavioral difference. If a warning signal is obtained, a warning feedback is given; Step 5: Conduct a network transaction security supervision analysis on the target customer's transaction partner information based on the information feedback to determine whether the target customer's transaction partner has any suspicious transaction behavior. If a suspicious signal is obtained, an early warning feedback will be issued; The transaction behavior supervision and analysis process is as follows: Set a monitoring cycle and set it as a time threshold, obtain historical transaction flow data of target customers within the time threshold, the historical transaction flow data includes daily consumption, deposits, withdrawals, and transfers, obtain the number of fund transfers out and the number of fund transfers in of the target customer within the time threshold from the historical transaction flow data, set the ratio between the number of fund transfers out and the number of fund transfers in of the target customer as the transaction propensity value, obtain the average interval duration of the number of fund transfers out and the average interval duration of the number of fund transfers in of the target customer within the time threshold from the historical transaction flow data, and set the product of the average interval duration of the number of fund transfers out and the average interval duration of the number of fund transfers in of the target customer as the flow transaction value; Compare and analyze the transaction tendency value and flow transaction value with the preset transaction tendency value threshold and preset flow transaction value threshold that are stored internally: If the trading propensity value is less than the preset trading propensity value threshold, and the flow trading value is less than the preset flow trading value threshold, no signal is generated; If the transaction propensity value is greater than or equal to the preset transaction propensity value threshold, or the flow transaction value is greater than or equal to the preset flow transaction value threshold, an active signal is generated; The behavior trend safety supervision analysis process is as follows: The transaction propensity value and flow transaction value of the target customer within the time threshold are obtained, and then the transaction propensity value and flow transaction value of the target customer within the historical j time thresholds are obtained, j is a natural number greater than zero, and a rectangular coordinate system is established with the number of time thresholds as the X-axis and the transaction propensity value and flow transaction value as the Y-axis respectively. The transaction propensity value curve and the flow transaction value curve are drawn by plotting points respectively, and the transaction propensity value and flow transaction value corresponding to the transaction propensity value curve and the flow transaction value curve are collected once every t1 time, and the transaction propensity value and flow transaction value corresponding to the transaction propensity value curve and the flow transaction value curve are collected once every t2 time, and the transaction propensity value and flow transaction value corresponding to the transaction propensity value curve and the flow transaction value curve are collected once every tg time, g is a natural number greater than zero, t1, t2...tg are not equal, and the sets A and B of the transaction propensity value and the flow transaction value are respectively constructed, and then the discrete coefficients of the sets A and B are obtained, and they are set as the transaction fluctuation value and the flow floating value respectively; Compare and analyze the transaction fluctuation value and liquidity floating value with the preset transaction fluctuation value threshold and preset liquidity floating value threshold that are stored internally: If the trading volatility value is less than the preset trading volatility value threshold, and the liquidity float value is less than the preset liquidity float value threshold, no signal is generated; If the transaction fluctuation value is greater than or equal to the preset transaction fluctuation value threshold, or the liquidity floating value is greater than or equal to the preset liquidity floating value threshold, an abnormal activity signal is generated.

2. A bank retail customer fund behavior analysis system, which is applied to a bank retail customer fund behavior analysis method according to claim 1, characterized in that: It includes fund supervision platform, transaction supervision unit, behavior assessment unit, transaction risk unit, fund tracking unit and management early warning unit; The transaction monitoring unit is used to collect historical transaction flow data of the target customer, perform transaction behavior monitoring analysis on the historical transaction flow data, and send the obtained active signal to the behavior evaluation unit; After receiving the active signal, the behavior assessment unit immediately conducts a behavior trend security supervision analysis and sends the obtained abnormal active signal to the transaction risk unit and the fund tracking unit; After receiving the abnormal activity signal, the transaction risk unit immediately collects the target customer's transaction performance data, which represents the transaction risk value and capital flow risk value, and conducts transaction risk safety supervision feedback analysis on the transaction performance data, and sends the obtained warning signals and risk signals to the management warning unit; After receiving the abnormally active signal, the fund tracking unit immediately collects the target customer's transaction object information, conducts a mesh transaction security supervision analysis on the transaction object information, and sends the obtained suspicious signal to the management warning unit.

3. A retail banking customer fund behavior analysis system according to claim 2, characterized in that: The transaction risk safety supervision feedback analysis process of the transaction risk unit is as follows: S1: Obtain the transaction risk value of the target customer within the time threshold from the historical transaction flow data. The transaction risk value represents the product of the number of consecutive transactions of the same amount and the number of occurrences of the corresponding transaction objects of the same amount after data normalization, and the ratio of the product of the number of consecutive transactions of the same amount and the average interval time after data normalization; S2: At the same time, the target customer's capital flow risk value within the time threshold is obtained from the historical transaction flow data. The capital flow risk value indicates the proportion of the number of times without filling in the transaction remarks in the total number of times the transaction amount exceeds the preset transaction amount threshold. The transaction risk value and the capital flow risk value are marked as JF and ZL respectively; S3: Obtain the transaction behavior risk coefficient F according to the formula, obtain the transaction behavior risk coefficient of the target customer within the historical j time thresholds, obtain the maximum value of the transaction behavior risk coefficient of the target customer within the historical j time thresholds, and set it as the transaction behavior risk critical value, and compare and analyze the transaction behavior risk coefficient F with the transaction behavior risk critical value: If the ratio between the trading behavior risk coefficient F and the trading behavior risk threshold is less than 1, a feedback instruction is generated; If the ratio between the trading behavior risk coefficient F and the trading behavior risk critical value is greater than or equal to 1, a risk signal is generated.

4. A retail banking customer fund behavior analysis system according to claim 3, characterized in that: When the transaction risk unit generates a feedback instruction: The target customer's transaction period information is obtained from the historical transaction flow data. Based on the target customer's transaction period information, the target customer's regular behavior characteristics are obtained. The difference between the regular behavior characteristics and the set regular behavior characteristics is obtained and set as the behavior difference degree. The behavior difference degree is compared and analyzed with the preset behavior difference degree threshold value stored in the internal data entry: If the behavior difference is less than the preset behavior difference threshold, no signal is generated; If the behavior difference is greater than or equal to the preset behavior difference threshold, an early warning signal is generated.

5. A retail banking customer fund behavior analysis system according to claim 2, characterized in that: The mesh transaction security supervision and analysis process of the fund tracking unit is as follows: The transaction object information of the target customer within the time threshold is obtained from the historical transaction flow data. The transaction object information includes the number of unfamiliar transaction objects and the number of known transaction objects. The unfamiliar transaction objects indicate that they do not appear in the historical transaction list, and the known transaction objects indicate that they appear in the historical transaction list. The associated transaction value between the unfamiliar transaction objects and the known transaction objects is obtained. The associated transaction value indicates the number of unfamiliar transaction objects corresponding to the transaction behaviors between the unfamiliar transaction objects and the known transaction objects, and the associated transaction value is set as the associated value. The value obtained by subtracting the associated value from the number of unfamiliar transaction objects is set as the unfamiliar risk value. The sum of the associated value and the number of known transaction objects is set as the network flow value. The ratio between the unfamiliar risk value and the network flow value is set as the capital transaction value. Compare and analyze the funds transaction value with the stored preset funds transaction value threshold: If the fund turnover value is less than the preset fund turnover value threshold, no signal is generated; If the fund transaction value is greater than or equal to the preset fund transaction value threshold, a suspicious signal is generated.

Images