Bank card issuing protection method and system and storable medium

Abstract

The invention relates to the field of government affair services, in particular to a bank card opening protection method and system and a storable medium, and the method comprises the steps: carrying out the feature extraction of first basic data information of a fraud victim, and obtaining and confirming an effective element and an effective feature data set; training a victim identification model according to the effective elements and the effective feature data set, and obtaining an identification reference value; substituting the second basic information of the jurisdiction users into the victim identification model to obtain a simulation identification value, and when the simulation identification value is greater than or equal to an identification reference value, filling the user basic information in the second reference information and the simulation identification value into a root grey code database; and data in the root grey code database is encrypted and then synchronously transmitted to the sub grey code database in the slave server, so that the slave server compares the received card opening application basic information to obtain a comparison result, and an anti-fraud alarm signal is triggered when the comparison result is consistent. According to the invention, victims and potential victims are automatically identified and information leakage is avoided.

Description

Technical Field

[0001] The present application relates to the field of government services, and in particular to a bank card opening protection method, system and storable medium. Background Art

[0002] It is urgent to effectively identify victims, carefully and accurately judge bank card freezing strategies, avoid causing trouble to non-victims due to misjudgment, and thus serve the people more efficiently and protect their property safety. Summary of the invention

[0003] In order to improve the accuracy of identifying victims and potential victims, the present application provides a bank card opening protection method, system and storable medium.

[0004] The present application provides a bank card opening protection method, system and storage medium, which adopt the following technical solutions: A bank card opening protection method is used for a master server associated with a server network, wherein the server network includes the master server and at least one slave server, and the at least one slave server is respectively connected to the corresponding master server in communication, and the method includes: Obtain the first basic data information of the fraud victim; Perform feature extraction on the first basic data information to obtain a feature data set corresponding to each preset element, and confirm valid elements and valid feature data sets; Training a victim identification model based on the valid elements and the valid feature data set, and obtaining an identification benchmark value; Obtain the second basic information of users in the jurisdiction; Substituting the second basic information into the victim identification model to obtain a simulated identification value, and when the simulated identification value is greater than or equal to the identification reference value, filling the user basic information in the second reference information and the simulated identification value into the root gray code database; The data in the root gray code database is encrypted and then synchronously transmitted to the gray code database in the slave server for comparison with the received basic information of the card application by the slave server to obtain a comparison result, and an anti-fraud alarm signal is triggered when the comparison result is consistent.

[0005] By adopting the above technical solution, firstly, the first basic information of all victims who have been defrauded is obtained, and the first basic information includes at least one of the victim's identity information, the victim's movement trajectory, the victim's capital flow, the victim's incoming call records, and the victim's web browsing records; Then, the first basic information of each victim is feature extracted according to the preset elements, so as to obtain the feature data set of each preset element, and realize the feature extraction and classification of the first basic information. For example, if the preset element is the call record answered by the victim, then the corresponding feature data set is the call record of multiple victims. After confirmation, the valid elements and valid feature data are obtained, which supports manual screening of the preset elements and feature data sets.

[0006] Next, the victim identification model is trained based on the effective elements and effective feature data sets, thereby obtaining the victim's fraud pattern and the critical fraud identification benchmark value. This is conducive to the subsequent substitution of the second basic information to determine the key fraud protection group among the users in the jurisdiction, so that the staff can carry out key publicity and refuse to execute the card opening instruction.

[0007] Optionally, the preset factors include a network risk behavior index, a telephone risk behavior index, a capital chain risk index, an enterprise personnel risk index, and an offline risk behavior index; the feature extraction of the first basic data information to obtain a feature data set corresponding to each preset factor includes: Filter the first basic data information corresponding to the fraud type according to the preset elements, and extract the corresponding element entries in the first basic data information; The element feature value corresponding to the element entry is calculated according to the element rule corresponding to the preset element, and the feature data set corresponding to the preset element is composed of multiple groups of element feature values ​​of the first basic information.

[0008] By adopting the above technical solution, the preset fraud type corresponds to one or more preset elements. Therefore, the first basic data information corresponding to the fraud type is screened out according to the preset elements to achieve accurate division of the first basic data.

[0009] Optionally, confirm valid features and valid feature datasets, including: According to the selection confirmation signal received from the user, the preset element corresponding to the selection confirmation signal is confirmed as a valid element from the corresponding preset elements and the feature data set, and the feature data set corresponding to the preset element is used as a valid feature data set.

[0010] By adopting the above technical solution, according to the selection confirmation signal from the user, the corresponding valid element is found from multiple preset elements, and the corresponding valid feature data set is set. The user's confirmation signal can confirm a preset element or set multiple preset elements, so that the user can set and adjust according to their own needs. The user's selection confirmation signal determines the number of preset elements, and then determines the accuracy of the trained victim recognition model, supports the reduction of preset elements, reduces the accuracy of the victim recognition model to improve the overall operation speed.

[0011] Optionally, the training of a victim identification model based on the valid elements and the valid feature data set, and obtaining an identification benchmark value, includes: According to the effective elements and the effective feature data set, a neural network model is trained to obtain a victim identification model, and the first basic data information is substituted into the victim identification model to obtain a corresponding identification value, and the lowest identification value is taken as the identification reference value.

[0012] By adopting the above technical solution, the effective elements and effective feature data sets are used as historical data to train the neural network model, thereby obtaining a victim identification model. The victim identification model summarizes the comprehensive rules of the effective elements and effective feature data sets of the victim. Subsequently, the first basic data is substituted into the victim identification model to obtain the identification value, thereby taking the lowest identification value as the identification benchmark value, and then using the identification benchmark value as the judgment standard, that is, whether the user substituted into the victim identification model is easily defrauded and whether it needs to be paid special attention.

[0013] Optionally, the training of a victim identification model based on the valid elements and the valid feature data set, and obtaining an identification benchmark value, includes: According to the effective elements and the effective feature data set, a neural network model is trained to obtain a victim identification model, and the first basic data information is substituted into the victim identification model to obtain a corresponding identification value, and the lowest identification value is taken as a first reference value; Counting the historical risk coefficient in the first basic information, wherein the historical risk coefficient is positively correlated with the percentage coefficient of the total amount of fraud to the total amount of the user and the frequency of fraud; The first reference value is used as a basis, and the sum of the products of the historical risk coefficient and the preset reference step length is subtracted therefrom to obtain the identification reference value.

[0014] By adopting the above technical solution, the setting of the identification benchmark value is more personalized, and various personalized factors of the user are added on the basis of the first benchmark value, such as the percentage coefficient of the total amount of fraud to the total amount of the user, and the frequency of fraud. This can achieve a high level of alertness to fraud for users with a history of risk.

[0015] Optionally, the historical risk coefficient is also positively correlated with the frequency of the user's recent travel to high-risk areas, the frequency of the user's financial transactions with high-risk accounts, and the frequency of phone calls between the user and high-risk accounts.

[0016] By adopting the above technical solution, the size of the identification benchmark value is positively correlated with the historical risk coefficient, and the historical risk coefficient is positively correlated with the frequency of the user's recent travel to high-risk areas, the frequency of the user's economic transactions with high-risk accounts, and the frequency of the user's calls with high-risk accounts. In this case, the size of the identification benchmark value is affected by the user's recent travel to high-risk areas, the frequency of the user's economic transactions with high-risk accounts, and the frequency of the user's calls with high-risk accounts, thereby raising the identification benchmark for users who have transportation, economic, and phone transactions with high-risk areas / personnel, thereby achieving key inspections and strictly restricting card opening permissions.

[0017] Optionally, encrypting the data in the root gray code database and then synchronously transferring the data into the sub-gray code database in the slave server includes: By adopting the above technical solution, Optionally, the method further includes: Obtaining the anti-fraud confirmation signal, card application basic information and the anti-fraud alarm signal generated according to the comparison result from the slave server; The second basic information corresponding to the card application basic information is updated to the new first basic information.

[0018] By adopting the above technical solution, when the anti-fraud alarm signal is triggered from the server when the comparison result is consistent, the staff will understand the actual situation and give advice to the user who submitted the basic information for the card opening application. When it is found that it is indeed a fraud case, an instruction is input to the server to generate an anti-fraud confirmation signal. Subsequently, the anti-fraud confirmation signal, the basic information for the card opening application and the anti-fraud alarm signal are transmitted from the server to the root gray code database, and the second basic information corresponding to the basic information for the card opening (if it is a resident within the jurisdiction, the second basic information of the resident) is updated to the new first basic information, that is, the applicant who submitted the basic information for the card opening application is registered as a fraud-related person.

[0019] The present invention also provides a bank card opening protection method, which is used for a slave server associated with a server network, wherein the server network includes the master server and at least one slave server, and the at least one slave server is respectively connected to the corresponding master server in communication, and the method includes: Get the basic card application information entered by the user; Searching whether there is corresponding user basic information in the gray code database of the slave server according to the card opening application information, and if so, the comparison result is consistent, triggering an anti-fraud alarm signal; if neither of them exists, the comparison result is inconsistent, triggering a card opening approval signal; Among them, the user basic information in the gray code database is transmitted by the main server after encrypting the root gray code database. The root gray code database is composed of a simulated identification value that exceeds the benchmark identification value and a second basic information that is substituted into the victim identification model to obtain the simulated identification value. The victim identification model and the benchmark identification value are both obtained by training after feature extraction from the first basic data information of the fraud victim.

[0020] By adopting the above technical solution, the first basic information of all victims who have been defrauded is first obtained, and then the first basic information of each victim is extracted according to the preset elements, so as to obtain the feature data set of each preset element, and realize the feature extraction and classification of the first basic information. For example, if the preset element is the call record answered by the victim, then the corresponding feature data set is the call record of multiple victims. After confirmation, valid elements and valid feature data are obtained, which supports manual screening of preset elements and feature data sets. Then, a victim identification model is trained based on the valid elements and valid feature data sets to obtain the victim's deception pattern and the critical identification benchmark value of being defrauded. It is conducive to the subsequent substitution of the second basic information to determine the key fraud protection group among each user in the jurisdiction, so that the staff can carry out key publicity and refuse to execute the card opening instruction.

[0021] Optionally, when the comparison result is consistent, the method further includes: Obtain an anti-fraud confirmation signal, and transmit the anti-fraud confirmation signal, the basic information of the card application, and the anti-fraud alarm signal generated according to the comparison result to the main server, so as to update the second basic information corresponding to the basic information of the card application to the first basic information; and after a preset time period, the slave server deletes the comparison result.

[0022] By adopting the above technical solution, when the anti-fraud alarm signal is triggered from the server when the comparison result is consistent, the staff will understand the actual situation and give advice to the user who submitted the basic information for the card opening application. When it is found that it is indeed a fraud case, an instruction is input to the server to generate an anti-fraud confirmation signal. Subsequently, the anti-fraud confirmation signal, the basic information for the card opening application and the anti-fraud alarm signal are transmitted from the server to the root gray code database, and the second basic information corresponding to the basic information for the card opening (if it is a resident within the jurisdiction, the second basic information of the resident) is updated to the new first basic information, that is, the applicant who submitted the basic information for the card opening application is registered as a fraud-related person.

[0023] The present invention also provides a bank card activation protection system, comprising a main server and at least one slave server constituting a server network, wherein the at least one slave server is respectively connected to a corresponding main server for communication, and the main server comprises a main data storage module, a data access module, a feature extraction module, a modeling module, a calculation module, an encryption module, and a data transmission module; The main data storage module is used to store the first basic data information of the fraud victim and store the root gray code database; The data access module is used to obtain the second basic information; The feature extraction module is connected to the main data storage module and is used to extract features from the first basic data information to obtain a feature data set corresponding to each preset element, and to confirm valid elements and valid feature data sets; The modeling module is connected to the feature extraction module and is used to train a victim recognition model based on the effective elements and the effective feature data set, and obtain a recognition reference value; The operation module is connected to the data access module, the modeling module and the main data storage module, and is used to substitute the second basic information into the victim identification model to obtain a simulated identification value, and when the simulated identification value is greater than or equal to the identification reference value, fill the user basic information in the second reference information and the simulated identification value into the root gray code database; The encryption module is connected to the main data storage module and is used to encrypt the data in the root gray code database; The data transmission module is connected to the encryption module and is used to transmit the encrypted data in the root gray code database to the slave server; The slave server includes a data synchronization module, a slave storage module, a collection module, a verification module, an alarm module and an output module; The data synchronization module is connected to the slave storage module and is used to receive the encrypted data in the root code database transmitted by the master server and store the data in the slave storage module as the gray code database after decryption; The acquisition module is used to obtain the basic information of the card application input by the user; The verification module is connected to the acquisition module and the slave storage module, and is used to find out whether there is corresponding user basic information in the gray code database of the slave server according to the card application information to obtain a comparison result; The alarm module is connected to the verification module, and when the comparison result is consistent, an anti-fraud alarm signal is activated; The output module is connected to the verification module, and outputs a card opening pass signal when the comparison result is inconsistent.

[0024] By adopting the above technical solution, the first basic information of all victims who have been defrauded is first obtained, and then the first basic information of each victim is extracted according to the preset elements, so as to obtain the feature data set of each preset element, and realize the feature extraction and classification of the first basic information. For example, if the preset element is the call record answered by the victim, then the corresponding feature data set is the call record of multiple victims. After confirmation, valid elements and valid feature data are obtained, which supports manual screening of preset elements and feature data sets. Then, a victim identification model is trained based on the valid elements and valid feature data sets to obtain the victim's deception pattern and the critical identification benchmark value of being defrauded. It is conducive to the subsequent substitution of the second basic information to determine the key fraud protection group among each user in the jurisdiction, so that the staff can carry out key publicity and refuse to execute the card opening instruction.

[0025] The present invention provides a storage medium, characterized in that the storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the bank card activation protection method as described in any one of the above items.

[0026] In summary, the present application includes at least one of the following beneficial technical effects: 1. In this case, the first basic information memory features of the fraud victims are extracted and model training is carried out based on this, thereby obtaining a victim identification model, which shows the unified fraud characteristics of all the imported fraud victims and the identification benchmark value that judges the law of fraud victims. The information of the fraud victims is summarized and concluded, and the victim identification model is conducive to the subsequent identification of the possibility of people in the existing jurisdiction or people in other external jurisdictions being victims of fraud.

[0027] 2. In this case, all users within the jurisdiction are first screened, and the second basic information is substituted into the victim identification model to obtain a simulated identification value. The comparison result between the simulated identification value and the identification reference value is used to determine whether the user corresponding to the second basic information meets the pattern of fraud victims, thereby determining that the user corresponding to the current second basic information is at risk of being defrauded.

[0028] 3. In this case, all the second basic information judged to be at risk of victimization will be stored in the root gray code database and transmitted to multiple slave servers after encryption, so that the slave servers can restrict card opening and issue anti-fraud alarms for users corresponding to the second basic information at multiple locations, thereby ensuring the stability of the root gray code database and the independence of the sub-gray code database.

[0029] 4. When using the gray code database to determine whether the user corresponding to the basic information of the card application is at risk of being defrauded, after manual confirmation, the first basic information is automatically updated according to the second basic information corresponding to the basic information of the application to ensure that the first basic data information of all victims of fraud can be updated in a timely manner.

[0030] 5. In the face of the first basic data information of different users, after setting up a unified victim identification model, a variety of identification benchmark values ​​will be set accordingly, so as to achieve flexible standards for victim identification for different situations and different degrees of severity, so as to prevent fraud in advance. For example, A has been defrauded 4 times, and B has been defrauded 0 times. When the victim identification model is used for identification, the identification benchmark value for A is higher than that for B, and the victim identification for A is more strict to avoid missed judgments as much as possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a structural diagram of server networking in the first embodiment of the present invention; Figure 2 It is a flowchart of the bank card opening protection method in the first embodiment of the present invention; Figure 3 It is a flow chart of the bank card opening protection method in the second embodiment of the present invention; Figure 4 is a schematic diagram of a module of a main server in a third embodiment of the present invention; Figure 5 It is a module diagram of a slave server in the fourth embodiment of the present invention. DETAILED DESCRIPTION

[0032] The present application is further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0033] In the following description, for the purpose of explanation, many specific details are set forth in order to provide a thorough understanding of the inventive concept. As part of this specification, some of the drawings in the drawings of the present disclosure represent structures and devices in the form of block diagrams to avoid making the disclosed principles complicated and difficult to understand. For the sake of clarity, not all features of the actual implementation are necessarily described. References to "one specific implementation" or "specific implementations" in this disclosure mean that the specific features, structures, or characteristics described in conjunction with the specific implementation are included in at least one specific implementation, and multiple references to "one specific implementation" or "specific implementations" should not be understood as necessarily all referring to the same specific implementation.

[0034] Unless expressly limited, the terms "a", "an", and "the" are not intended to refer to a singular entity, but rather to include general categories of which specific examples may be used for illustration. Thus, the use of the terms "a" or "an" may mean any number of at least one, including "one", "one or more", "at least one", and "one or more than one". The term "or" means any of the alternatives and any combination of the alternatives, including all alternatives, unless the alternatives are expressly indicated to be mutually exclusive. The phrase "at least one of" when combined with a list of items refers to a single item in the list or any combination of the items in the list. The phrase does not require all of the listed items unless expressly limited to that.

[0035] First embodiment: The present application embodiment discloses a bank card opening protection method. Figure 1 The bank card opening protection method is used for a main server associated with a server network, wherein the server network includes the main server and at least one slave server, and the at least one slave server is respectively communicated with the corresponding main server.

[0036] refer to Figure 2 The bank card opening protection method includes: obtaining first basic data information of a fraud victim; performing feature extraction on the first basic data information to obtain a feature data set corresponding to each preset element, and confirming a valid element and a valid feature data set; training a victim recognition model based on the valid element and the valid feature data set, and obtaining a recognition benchmark value; Obtain the second basic information of users in the jurisdiction; substitute the second basic information into the victim identification model to obtain a simulated identification value, and when the simulated identification value is greater than or equal to the identification reference value, fill the user basic information and the simulated identification value in the second reference information into the root gray code database; The data in the root gray code database is encrypted and then synchronously transmitted to the gray code database in the slave server for comparison with the basic information of the card application received by the slave server to obtain the comparison result. When the comparison result is consistent, the anti-fraud alarm signal is triggered.

[0037] In the specific implementation, the scheme first obtains the first basic information of all victims who have been defrauded, and then extracts the features of the first basic information of each victim according to the preset elements, thereby obtaining a feature data set for each preset element, and realizing feature extraction and classification of the first basic information. For example, if the preset element is the call record answered by the victim, then the corresponding feature data set is the call record of multiple victims. After confirmation, valid elements and valid feature data are obtained, which supports manual screening of preset elements and feature data sets.

[0038] Next, the victim identification model is trained based on the effective elements and effective feature data sets, thereby obtaining the victim's fraud pattern and the critical fraud identification benchmark value. This is conducive to the subsequent substitution of the second basic information to determine the key fraud protection group among the users in the jurisdiction, so that the staff can carry out key publicity and refuse to execute the card opening instruction.

[0039] The slave server is usually a server in a bank or other business place. The slave server will only obtain the target data (usually the gray list) in the root gray code database after the level 1 encryption from the master server to prevent the irrelevant data in the master server from being leaked to the slave server, which would bring data security risks.

[0040] The following is a detailed description of the implementation details of the bank card opening protection method of this embodiment. The following content is only for the convenience of understanding the implementation details provided, and is not necessary for the implementation of this solution. The specific process of this embodiment is as follows Figure 2 As shown, the following steps are included: Step 101, obtaining the first basic data information of the fraud victim.

[0041] Specifically, fraud victims usually include: victims who have registered with the public security organs, and persons involved in fraud cases cracked by the public security organs. The first basic data information includes at least one of the victim's identity information, the victim's movement trajectory, the victim's capital flow, the victim's incoming call records, and the victim's web browsing records.

[0042] Step 102: extract features from the first basic data information to obtain a feature data set corresponding to each preset element, and confirm valid elements and valid feature data sets.

[0043] Specifically, the preset elements are used to extract features from the first basic information of each victim, thereby obtaining a feature data set for each preset element, and realizing feature extraction and classification of the first basic information. For example, if the preset element is the call record answered by the victim, then the corresponding feature data set is the call record of multiple victims. After confirmation, valid elements and valid feature data are obtained, and manual screening of preset elements and feature data sets is supported. The confirmation step is established to facilitate quality inspection personnel to further screen, merge, add or delete preset elements and feature data sets.

[0044] In some examples, feature extraction is performed on the first basic data information to obtain a feature data set corresponding to each preset element, including: S2-1, filtering first basic data information corresponding to the fraud type according to preset factors; S2-2, extracting corresponding element terms in the first basic data information after screening; S2-3, calculating the element feature value corresponding to the element entry according to the element rule corresponding to the preset element, and forming a feature data set corresponding to the preset element by multiple groups of element feature values ​​of the first basic information.

[0045] Specifically, the preset factors include network risk behavior index, telephone risk behavior index, capital chain risk index, enterprise personnel risk index, and offline risk behavior index. The preset fraud type corresponds to one or more preset factors. Therefore, the first basic data information corresponding to the fraud type is screened out according to the preset factors to achieve accurate division of the first basic data.

[0046] Taking the preset factor of "Internet risk behavior index" as an example, the preset factor is a preset factor related to the type of "Internet fraud". When S2-1 is executed, it will only filter out the first basic data information of the type of "Internet fraud" from all the first basic data information.

[0047] S2-2 then extracts data from the first basic data information screened out by S2-1 to obtain element entries corresponding to the preset element "network risk behavior index", wherein the extraction of element entries can be performed based on the keywords corresponding to the "network risk behavior index" to obtain element entries, and the keywords corresponding to the "network risk behavior index" are also pre-set.

[0048] After S2-2 is run, S2-3 is executed. The element rule corresponding to the preset element is actually the rule for calculating the element characteristic value based on the element entry. For example: the first element entry contains the detection value a, and the second element entry contains the time value t. Then the element rule is the formula for calculating the element characteristic value by combining different element entries. This formula is preset. Subsequently, the element characteristic value is calculated based on the element entry and the element rule, and then summarized to obtain the characteristic data set of the preset element. Each first basic data screened out by S2-1 corresponds to an element characteristic value of a preset element. In this case, multiple groups of element characteristic values ​​of the first basic information constitute a characteristic data set corresponding to the preset element. In this way, the characteristic data set is the set of all element characteristic values ​​corresponding to the "Internet Risk Behavior Index" under the "Internet Fraud" type.

[0049] In some examples, valid elements and valid feature data sets are identified, including: According to the selection confirmation signal received from the user, the preset element corresponding to the selection confirmation signal is confirmed as a valid element from the corresponding preset elements and feature data sets, and the feature data set corresponding to the preset element is used as a valid feature data set.

[0050] In this example, if no user confirmation signal is received within the preset time period, the current preset elements and feature data sets will be used as valid elements and valid feature data sets by default. This state is the default state. The preset time period here is a waiting time, which is usually set by the staff.

[0051] A user selection confirmation signal is received within a preset time period. According to the selection confirmation signal from the user, the corresponding valid element is found from multiple preset elements, and the valid feature data set of the corresponding valid element is set. The user's confirmation signal can confirm a preset element or set multiple preset elements, so that the user can set and adjust according to their own needs. The user's selection confirmation signal determines the number of preset elements, and then determines the accuracy of the trained victim recognition model. It supports the reduction of preset elements, reduces the accuracy of the victim recognition model, and improves the overall operation speed.

[0052] Step 103: Train a victim recognition model based on the valid elements and valid feature data sets, and obtain a recognition benchmark value.

[0053] In some examples, the victim identification model is trained based on the valid elements and the valid feature set. Step 103 includes: S3-1, based on the valid elements and valid feature data sets, train the neural network model to obtain the victim identification model; S3-2, substitute the first basic data information into the victim identification model to obtain the corresponding identification value, and take the lowest identification value as the identification reference value.

[0054] The effective elements and effective feature data sets are used as historical data to train the neural network model, thereby obtaining a victim identification model. The victim identification model summarizes the comprehensive rules of the effective elements and effective feature data sets of the victim. Subsequently, the first basic data is substituted into the victim identification model to obtain the identification value, thereby taking the lowest identification value as the identification benchmark value, and then using the identification benchmark value as the judgment standard, that is, whether the user substituted into the victim identification model is very likely to be defrauded and whether it needs to be paid special attention.

[0055] In some examples, the victim identification model is pre-set and stored, so when applied, the corresponding victim identification model and the corresponding identification reference value are directly found according to the valid elements. Usually, the victim identification model is announced by the government and can be pre-entered and set by the staff.

[0056] In some examples, implementation of step 103 includes: S3-1, based on the valid elements and valid feature data sets, train the neural network model to obtain the victim identification model; S3-2', substituting the first basic data information into the victim identification model to obtain a corresponding identification value, and taking the lowest identification value as the first reference value; S3-3, calculating the historical risk coefficient in the first basic information, where the historical risk coefficient is positively correlated with the percentage coefficient of the total amount of fraud to the total amount of the user and the frequency of fraud; S3-4, taking the first benchmark value as a basis, subtract the sum of the products of the historical risk coefficient and the preset benchmark step length from it to obtain the identification benchmark value.

[0057] In this example, after obtaining the victim identification model in S3-1, the lowest identification value is taken as the first benchmark value, and then subtracted from the first base value based on the historical risk coefficient to obtain the identification benchmark value. In this example, the setting of the identification benchmark value is more personalized, and various personalized factors of the user are added on the basis of the first benchmark value. The percentage coefficient of the total amount defrauded to the total amount of the user and the frequency of being defrauded are used to increase the awareness of being defrauded for users with historical risks and users with high amounts of defrauded.

[0058] For example, user A has been a victim of telephone fraud four times, and user B has no experience of being a victim of telephone fraud. In this case, user B's basic identification value is the lowest. User A's basic identification value is lower than that of B. A lower base value means that A is more likely to be included in the gray code database, and when handling banking and other financial services, staff confirmation and persuasion are more needed.

[0059] Furthermore, the historical risk coefficient is positively correlated with the frequency of users' recent travel to high-risk areas, the frequency of users' economic transactions with high-risk accounts, and the frequency of users' phone calls with high-risk accounts.

[0060] In this example, the size of the identification benchmark value is positively correlated with the historical risk coefficient, which is positively correlated with the frequency of the user's recent travel to high-risk areas, the frequency of the user's economic transactions with high-risk accounts, and the frequency of the user's calls with high-risk accounts. The size of the identification benchmark value is affected by the frequency of the user's recent travel to high-risk areas, the frequency of the user's economic transactions with high-risk accounts, and the frequency of the user's calls with high-risk accounts. This increases the identification benchmark for users who have transportation, economic, and phone transactions with high-risk areas / people, thereby achieving key inspections and strictly restricting card opening permissions.

[0061] Step 104, obtaining the second basic information of users in the jurisdiction.

[0062] Specifically, the second basic information is actively input or imported by the staff. The setting of the jurisdiction is set by the staff themselves, and the second basic information usually includes at least one of the following: the identity information of the users in the jurisdiction, the movement trajectory of the users in the jurisdiction, the capital flow of the users in the jurisdiction, the call records received by the users in the jurisdiction, and the web browsing records of the users in the jurisdiction.

[0063] Step 105, substitute the second basic information into the victim identification model to obtain a simulated identification value, and when the simulated identification value is greater than or equal to the identification reference value, fill the user basic information and the simulated identification value in the second reference information into the root gray code database.

[0064] Specifically, the second basic information is substituted into the victim identification model to calculate the simulated identification value. When the simulated identification value reaches the identification reference value, it is determined that the user corresponding to the second basic information meets the victim user rule; then the user basic information and simulated identification value in the second reference information are filled into the root gray code database. It is used to synchronize with each gray code database. At the same time, the root gray code database strictly controls the content of the stored data to prevent data leakage.

[0065] Step 106, encrypt the data in the root gray code database, and then synchronously transfer it to the sub-gray code database in the slave server for comparison with the received card application basic information to obtain a comparison result, and trigger an anti-fraud alarm signal when the comparison result is consistent.

[0066] Specifically, in step 106, the data in the root gray code database is encrypted and then synchronously transferred to the sub-gray code database in the slave server, including: The data in the root gray code database is encrypted using the national encryption algorithm SM4+SM2. The encrypted data in the root gray code database is transferred to the sub-gray code database, realizing data encryption and synchronization between the root gray code database of the master server and the sub-gray code databases of multiple slave servers. This prevents the user's confidential information (travel itinerary, call records, etc.) from being leaked after the data of the master server and the slave server are intercepted.

[0067] After comparing the basic information of the card application received from the server with the data in the gray code database, the comparison result is obtained. If there is a match, an anti-fraud alarm signal is triggered, thereby prompting the staff at the location to determine whether it is a fraud and provide timely guidance.

[0068] In some examples, after step 106 is completed, the method of the present invention further includes: Step 107, obtaining the anti-fraud confirmation signal, the basic information of the card application, and the anti-fraud alarm signal generated according to the comparison result from the server; and updating the second basic information corresponding to the basic information of the card application to the new first basic information.

[0069] Specifically, when the anti-fraud alarm signal is triggered from the server when the comparison result is consistent, the staff will understand the actual situation and give advice to the user who submitted the basic information for the card opening application. When it is found that it is indeed a fraud case, an instruction will be input to the server to generate an anti-fraud confirmation signal. The slave server will then transmit the anti-fraud confirmation signal, the basic information for the card opening application, and the anti-fraud alarm signal to the root gray code database, and update the second basic information corresponding to the basic information for the card opening (if it is a resident within the jurisdiction, the second basic information of the resident) to the new first basic information, that is, the applicant who submitted the basic information for the card opening application will be registered as a fraud-related person.

[0070] The step division of the above-mentioned methods is only for the purpose of clear description. When implemented, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this patent. Adding insignificant modifications or introducing insignificant designs to the algorithm or process without changing the core design of the algorithm and process are all within the scope of protection of this patent.

[0071] Second implementation method: The second embodiment of the present invention provides a bank card opening protection method, such as Figure 3 As shown, a slave server associated with a server network, wherein the server network includes a master server and at least one slave server, and at least one slave server is respectively connected to a corresponding master server in communication, and the method includes: Step 201, obtaining basic card application information input by the user; Step 202: Check whether the corresponding basic user information exists in the gray code database of the server according to the card application information. If it exists, the comparison result is consistent, and an anti-fraud alarm signal is triggered. If it does not exist, the comparison result is inconsistent, and a card opening approval signal is triggered; Among them, the user basic information in the gray code database is transmitted to the root gray code database after data encryption by the main server. The root gray code database is composed of a simulated identification value that exceeds the benchmark identification value and a second basic information that is substituted into the victim identification model to obtain a simulated identification value. The victim identification model and the benchmark identification value are both obtained by training after feature extraction from the first basic data information of the fraud victim.

[0072] In this embodiment, the main server obtains a victim identification model that represents the pattern of victims in fraud based on the first basic information of the victim through feature extraction training, and relies on the first basic information and the victim identification model to calculate and determine whether it is a benchmark identification value for determining whether it is a victim. The main server substitutes the second basic information of each user in the jurisdiction into the victim identification model, thereby obtaining the second basic information of the suspected victim after screening the benchmark identification value, and stores the second basic information of the suspected victim in the root gray code database of the main server, and transmits it to the gray code database of the slave server in the corresponding jurisdiction after encryption.

[0073] In step 201, there are many ways to obtain the card application information input by the user from the server, for example: first, the user inputs the card application information to the slave server through an input device such as a keyboard; second, the user can scan a QR code representing the current slave server to enter a web page interface, which displays a preset blank template. The user fills in the corresponding information in the blank part of the web page interface, and the web page interface directly generates the corresponding card application information after the user confirms; third, the user imports the data of a third party and uses the third party's data as the card application information in order of precedence.

[0074] In step 202, if the card application information is consistent with the basic user information in the gray code database from the server, an anti-fraud alarm signal will be triggered to prompt the staff to go to the user who applied for the card to understand whether the card opening or financial services such as bank transfers are fraudulent; if the card application information is inconsistent with the basic user information in the gray code database from the server, a card opening approval signal will be triggered to notify the staff to remember the card opening or bank transfer and other financial services for the customer.

[0075] In some examples, the method further includes: 203, the card application basic information and the anti-fraud alarm signal generated according to the comparison result are transmitted to the main server, so as to update the second basic information corresponding to the card application basic information to the first basic information; 204: After a preset period of time, the comparison result is deleted from the server.

[0076] In this example, after being manually determined to be a fraudulent act, the card application information is directly synchronized to the first basic information as a successful fraud case, and the victim database where the first basic information is located is updated to ensure the accurate operation of the main server, and the comparison result is deleted from the slave server the next day to avoid a large amount of redundant data in the slave server and improve the data processing efficiency of the slave server.

[0077] The bank card opening protection method disclosed in this application is used for a slave server associated with a server network. Other implementation details and working methods are the same or similar to the bank card opening protection method described above, which is used for a slave server associated with a server network and will not be repeated here.

[0078] Third embodiment: The present invention also provides a main server of the bank card opening protection system. The main server establishes a communication connection with at least one slave server, and a server network is formed between the main server and the slave servers.

[0079] refer to Figure 4 , the main server 30 includes a main data storage module 301, a data access module 302, a feature extraction module 303, a modeling module 304, a calculation module 305, an encryption module 306, and a data transmission module 307; The main data storage module 301 is used to store the first basic data information of the fraud victim and store the root gray code database; The data access module 302 is used to obtain the second basic information; The feature extraction module 303 is connected to the main data storage module 301 and is used to extract features from the first basic data information, obtain a feature data set corresponding to each preset element, and confirm valid elements and valid feature data sets; The modeling module 304 is connected to the feature extraction module 303 and is used to train a victim recognition model based on the valid elements and the valid feature data set, and obtain a recognition benchmark value; The operation module 305 is connected to the data access module 302, the modeling module 304 and the main data storage module 301, and is used to substitute the second basic information into the victim identification model to obtain a simulated identification value, and when the simulated identification value is greater than or equal to the identification reference value, fill the user basic information and the simulated identification value in the second reference information into the root gray code database; An encryption module 306, connected to the main data storage module 301, is used to encrypt the data in the root gray code database; The data transmission module 307 is connected to the encryption module 306 and is used to transmit the encrypted data in the root gray code database to the slave server 40 for comparison with the received basic information of the card application from the slave server 40 to obtain a comparison result, and trigger an anti-fraud alarm signal when the comparison result is consistent.

[0080] In the implementation method, the main server first obtains the first basic information of all victims who have been defrauded, and then extracts features from the first basic information of each victim according to the preset elements, thereby obtaining a feature data set for each preset element, and realizing feature extraction and classification of the first basic information. For example, if the preset element is the call record answered by the victim, then the corresponding feature data set is the call record of multiple victims. After confirmation, valid elements and valid feature data are obtained, which supports manual screening of preset elements and feature data sets. Then, a victim identification model is trained based on the valid elements and valid feature data sets, thereby obtaining the victim's deception pattern and the critical identification benchmark value of being defrauded. It is beneficial to subsequently substitute the second basic information to judge the key fraud protection group among each user in the jurisdiction, so that the staff can carry out key publicity and refuse to execute the card opening instruction.

[0081] Fourth embodiment: A fourth embodiment of the present invention provides a slave server of a bank card opening protection system, wherein at least one slave server establishes a communication connection with a master server, and a server network is formed between the master server and the slave servers.

[0082] refer to Figure 5 , the slave server 40 includes a data synchronization module 401, a slave storage module 402, a collection module 403, a verification module 404, an alarm module 405 and an output module 406; The data synchronization module 401 is connected to the slave storage module 402 and is used to receive the encrypted data in the root code database transmitted by the main server 30 and store it in the slave storage module as the gray code database after decryption; The acquisition module 403 is used to obtain the basic information of the card application input by the user; The verification module 404 is connected to the acquisition module 403 and the storage module 402, and is used to find out whether there is corresponding user basic information in the gray code database of the slave server according to the card application information to obtain a comparison result; The alarm module 405 is connected to the verification module 404, and when the comparison result is consistent, an anti-fraud alarm signal is activated; The output module 406 is connected to the verification module 404, and outputs a card opening pass signal when the comparison result is inconsistent; Among them, the user basic information in the gray code database is transmitted to the main server 30 after encrypting the root gray code database. The root gray code database is composed of a simulated identification value that exceeds the benchmark identification value and a second basic information that is substituted into the victim identification model to obtain a simulated identification value. The victim identification model and the benchmark identification value are both obtained by training after feature extraction from the first basic data information of the fraud victim.

[0083] In this embodiment, the main server obtains a victim identification model representing the law of victims in fraud through feature extraction training based on the first basic information of the victim who was defrauded, and relies on the first basic information and the victim identification model to calculate and obtain a benchmark identification value for judging whether it is a victim. The main server substitutes the second basic information of each user in the jurisdiction into the victim identification model, so that after the screening of the benchmark identification value, the second basic information of the suspected victim is obtained, and the second basic information determined to be the victim is stored in the root gray code database of the main server, and is encrypted and transmitted to the gray code database of the slave server in the corresponding jurisdiction. According to the consistency between the card application information and the basic information of the user in the gray code database of the slave server, an anti-fraud alarm signal will be triggered to prompt the staff to go to the user who applied for the card to understand whether the card opening or bank transfer and other financial services are fraudulent; if the card application information is inconsistent with the basic information of the user in the gray code database of the slave server, a card opening pass signal will be triggered, thereby notifying the staff to remember the card opening or bank transfer and other financial services for the customer.

[0084] Other implementation details and working methods of the slave server of the bank card opening protection system disclosed in this application are the same or similar to the bank card opening protection method described in the second embodiment, and will not be repeated here.

[0085] Fifth embodiment: The fifth embodiment of the present invention provides a bank card opening protection system, referring to Figure 4 and Figure 5 , including a main server 30 and at least one slave server 40 constituting a server network, at least one slave server 40 is respectively connected to the corresponding main server 30 in communication, the main server 30 includes a main data storage module 301, a data access module 302, a feature extraction module 303, a modeling module 304, a calculation module 305, an encryption module 306, and a data transmission module 307; the slave server 40 includes a data synchronization module 401, a slave storage module 402, a collection module 403, a verification module 404, an alarm module 405, and an output module 406. The structure of the main server 30 refers to the third embodiment, and the structure of the slave server 40 refers to the fourth embodiment.

[0086] The present application also provides a storage medium, which stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the bank card activation protection method as described above.

[0087] Other implementation details and working methods of a bank card opening protection system and storage medium disclosed in this application are the same or similar to the bank card opening protection method described above and will not be repeated here.

[0088] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A bank card opening protection method, used for a main server associated with a server network, wherein: The server network includes the master server and at least one slave server, and the at least one slave server is respectively connected to the corresponding master server for communication, and the method includes: Obtain the first basic data information of the fraud victim; Perform feature extraction on the first basic data information to obtain a feature data set corresponding to each preset element, and confirm valid elements and valid feature data sets; Training a victim identification model based on the valid elements and the valid feature data set, and obtaining an identification benchmark value; Obtain the second basic information of users in the jurisdiction; Substituting the second basic information into the victim identification model to obtain a simulated identification value, and when the simulated identification value is greater than or equal to the identification reference value, filling the user basic information in the second reference information and the simulated identification value into the root gray code database; The data in the root gray code database is encrypted and then synchronously transmitted to the gray code database in the slave server for comparison with the received basic information of the card application by the slave server to obtain a comparison result, and an anti-fraud alarm signal is triggered when the comparison result is consistent.

2. A bank card opening protection method according to claim 1, characterized in that: The preset factors include a network risk behavior index, a telephone risk behavior index, a capital chain risk index, an enterprise personnel risk index, and an offline risk behavior index; the feature extraction of the first basic data information to obtain a feature data set corresponding to each preset factor includes: Filter the first basic data information corresponding to the fraud type according to the preset elements, and extract the corresponding element entries in the first basic data information; The element feature value corresponding to the element entry is calculated according to the element rule corresponding to the preset element, and the feature data set corresponding to the preset element is composed of multiple groups of element feature values ​​of the first basic information.

3. A bank card opening protection method according to claim 1, characterized in that: Confirm valid features and valid feature datasets, including: According to the selection confirmation signal received from the user, the preset element corresponding to the selection confirmation signal is confirmed as a valid element from the corresponding preset elements and the feature data set, and the feature data set corresponding to the preset element is used as a valid feature data set.

4. A bank card opening protection method according to claim 1, characterized in that: The step of training a victim identification model based on the valid elements and the valid feature data set and obtaining an identification reference value includes: According to the effective elements and the effective feature data set, a neural network model is trained to obtain a victim identification model, and the first basic data information is substituted into the victim identification model to obtain a corresponding identification value, and the lowest identification value is taken as the identification reference value.

5. A bank card opening protection method according to claim 1, characterized in that: The step of training a victim identification model based on the valid elements and the valid feature data set and obtaining an identification reference value includes: According to the effective elements and the effective feature data set, a neural network model is trained to obtain a victim identification model, and the first basic data information is substituted into the victim identification model to obtain a corresponding identification value, and the lowest identification value is taken as a first reference value; Counting the historical risk coefficient in the first basic information, wherein the historical risk coefficient is positively correlated with the percentage coefficient of the total amount of fraud to the total amount of the user and the frequency of fraud; The first reference value is used as a basis, and the sum of the products of the historical risk coefficient and the preset reference step length is subtracted therefrom to obtain the identification reference value.

6. A bank card opening protection method according to claim 5, characterized in that: The historical risk coefficient is also positively correlated with the frequency of the user's recent travel to high-risk areas, the frequency of the user's financial transactions with high-risk accounts, and the frequency of phone calls between the user and high-risk accounts.

7. A bank card opening protection method according to claim 1, characterized in that: The data in the root gray code database is encrypted and then synchronously transferred to the sub-gray code database in the slave server, including: By adopting the above technical solution, Optionally, the method further includes: Obtaining the anti-fraud confirmation signal, card application basic information and the anti-fraud alarm signal generated according to the comparison result from the slave server; The second basic information corresponding to the card application basic information is updated to the new first basic information.

8. A bank card opening protection method, used for a slave server associated with a server network, wherein: The server network includes the master server and at least one slave server, and the at least one slave server is respectively connected to the corresponding master server for communication, and the method includes: Get the basic card application information entered by the user; Searching whether there is corresponding user basic information in the gray code database of the slave server according to the card opening application information, and if so, the comparison result is consistent, triggering an anti-fraud alarm signal; if neither of them exists, the comparison result is inconsistent, triggering a card opening approval signal; Among them, the user basic information in the gray code database is transmitted by the main server after encrypting the root gray code database. The root gray code database is composed of a simulated identification value that exceeds the benchmark identification value and a second basic information that is substituted into the victim identification model to obtain the simulated identification value. The victim identification model and the benchmark identification value are both obtained by training after feature extraction from the first basic data information of the fraud victim.

9. A bank card opening protection method according to claim 8, characterized in that: When the comparison result is consistent, the method comprises: Obtain an anti-fraud confirmation signal, and transmit the anti-fraud confirmation signal, the basic information of the card application, and the anti-fraud alarm signal generated according to the comparison result to the main server, so as to update the second basic information corresponding to the basic information of the card application to the first basic information; and after a preset time period, the slave server deletes the comparison result.

10. A bank card activation protection system, comprising a master server and at least one slave server constituting a server network, wherein the at least one slave server is respectively connected to a corresponding master server for communication, characterized in that: The main server includes a main data storage module, a data access module, a feature extraction module, a modeling module, a calculation module, an encryption module, and a data transmission module; The main data storage module is used to store the first basic data information of the fraud victim and store the root gray code database; The data access module is used to obtain the second basic information; The feature extraction module is connected to the main data storage module and is used to extract features from the first basic data information to obtain a feature data set corresponding to each preset element, and to confirm valid elements and valid feature data sets; The modeling module is connected to the feature extraction module and is used to train a victim recognition model based on the effective elements and the effective feature data set, and obtain a recognition reference value; The operation module is connected to the data access module, the modeling module and the main data storage module, and is used to substitute the second basic information into the victim identification model to obtain a simulated identification value, and when the simulated identification value is greater than or equal to the identification reference value, fill the user basic information in the second reference information and the simulated identification value into the root gray code database; The encryption module is connected to the main data storage module and is used to encrypt the data in the root gray code database; The data transmission module is connected to the encryption module and is used to transmit the encrypted data in the root gray code database to the slave server; The slave server includes a data synchronization module, a slave storage module, a collection module, a verification module, an alarm module and an output module; The data synchronization module is connected to the slave storage module and is used to receive the encrypted data in the root code database transmitted by the master server and store the data in the slave storage module as the gray code database after decryption; The acquisition module is used to obtain the basic information of the card application input by the user; The verification module is connected to the acquisition module and the slave storage module, and is used to find out whether there is corresponding user basic information in the gray code database of the slave server according to the card application information to obtain a comparison result; The alarm module is connected to the verification module, and when the comparison result is consistent, an anti-fraud alarm signal is activated; The output module is connected to the verification module, and outputs a card opening pass signal when the comparison result is inconsistent.

11. A storage medium, characterized in that: The storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the bank card activation protection method as described in any one of claims 1-8.

Images