Business risk control early warning interception method and system and storage medium

Abstract

The invention provides a business risk control early warning interception method and system and a storage medium, and the method comprises the steps: obtaining customer information, and determining a customer image, a customer type and historical business information based on the customer information; judging whether an abnormal service exists in the historical service information, and if yes, determining a number corresponding to the historical service information as an abnormal number; if not, calling out a number corresponding to the historical service information, receiving actual call information corresponding to the number, and judging whether the actual call information accords with reasoning call information corresponding to the customer image and the customer type; if yes, generating a maintaining signal to continue to receive actual call information corresponding to the number; and if not, generating an early warning signal and intercepting the service. According to the application, malicious disturbance of illegal and criminal behaviors such as telecommunication harassment and telecommunication fraud can be reduced by actively identifying the illegal and criminal behaviors such as telecommunication harassment and telecommunication fraud.

Description

Technical Field

[0001] This application relates to the field of early warning technology, and in particular to a business risk control early warning interception method, system and storage medium. Background Technology

[0002] With the rapid development of communication technology, telephone fraud and nuisance calls have become increasingly serious problems, causing great distress and economic losses to users. As the main providers of communication services, telecommunications operators urgently need effective technical means to monitor and manage outgoing calls, preventing users from using their numbers for illegal activities such as telecommunications harassment and fraud.

[0003] Currently, users generally decide whether to hang up a call based on their own judgment. This can easily lead users who lack awareness of illegal and criminal activities such as telecommunications harassment and fraud, or to fall into traps set by emerging scams.

[0004] Alternatively, to reduce malicious harassment from illegal and criminal activities such as telecommunications harassment and fraud, some numbers received may be blacklisted. While this may prevent harassment from those numbers, it cannot prevent malicious harassment from being carried out by new numbers. Therefore, existing methods can only passively identify illegal and criminal activities such as telecommunications harassment and fraud, and cannot effectively proactively identify such illegal activities. Summary of the Invention

[0005] In order to reduce malicious interference from illegal and criminal activities such as telecommunications harassment and telecommunications fraud by proactively identifying such activities, this application provides a business risk control early warning and interception method, system and storage medium.

[0006] Firstly, this embodiment provides a business risk control early warning and interception method, the method comprising: Obtain customer information, and determine customer image, customer type, and historical business information based on the customer information; Determine whether there are any abnormal transactions in the historical service information. If so, identify the number corresponding to the historical service information as an abnormal number. If not, call out the number corresponding to the historical service information, receive the actual call information corresponding to the number, and determine whether the actual call information matches the inferred call information corresponding to the customer image and the customer type. If the condition is met, a sustain signal is generated to continue receiving the actual call information corresponding to the number; If the conditions are not met, a warning signal is generated and the service is blocked.

[0007] In some embodiments, the customer information represents information registered to the interceptor, and the customer information includes the customer's industry, the industry of the calling group, and the customer's purpose. After obtaining the customer information, the process further includes: Determine whether the customer industry, the calling group industry, and the customer purpose correspond to each other and conform to the preset purpose and preset industry. If so, generate a correct filing signal. If not, generate a re-registration signal and send it to the customer to retrieve the customer information again.

[0008] In some embodiments, the customer information further includes a customer number, and determining the customer image, customer type, and historical business information based on the customer information includes: The system retrieves industry images corresponding to the customer's industry, dialing images corresponding to the dialing group's industry, and usage images corresponding to the customer's purpose from a preset image database. It then determines whether there are any conflicts between the industry images, dialing images, and usage images. If there are no conflicts, the system integrates the industry images, dialing images, and usage images to obtain the customer image. If present, determine an adjustment image based on the industry image, dialing image, and usage image, and integrate the adjustment image to obtain a customer image; Obtain the single-dimensional types corresponding to the customer industry, the calling group industry, and the customer purpose, and determine the type that appears most frequently among all single-dimensional types as the customer type; Use the customer number to retrieve the historical service information corresponding to the customer number from the call database.

[0009] In some embodiments, the historical service information includes several historical service sub-information entries, and determining whether the historical service information contains abnormal services includes: Send the historical business information to a preset audit model to obtain the audit result, and obtain the interception information of the intercepted historical business sub-information among all the historical business sub-information contained in the historical business information; Determine whether the interception information is consistent with the audit result. If they are consistent, obtain the actual number of times the interception was carried out. Determine whether the actual number of times exceeds the preset number of times. If it exceeds the preset number of times, the historical business information indicates abnormal business. If the number does not exceed the limit, the historical service information does not contain any abnormal services; If there is a discrepancy, obtain the difference information, and adjust the audit results and / or blocking information based on the difference information to obtain the adjustment information; Obtain the actual number of times the adjustment information was blocked, and determine whether the actual number of times exceeds the preset number. If it does, the historical service information indicates abnormal service. If the number of cases does not exceed a certain threshold, the historical service information does not contain any abnormal services.

[0010] In some embodiments, determining whether the actual call information matches the inferred call information corresponding to the customer image and the customer type includes: Based on the customer type, determine the general call analysis model corresponding to the customer from the preset analysis model; The customer image is used to adjust the weight parameters representing personality in the general call analysis model to obtain the customer call analysis model. The customer type is sent to the customer call analysis model to obtain inferred call information; Obtain the actual topic information in the actual call information and the inferred topic information in the inferred call information, determine whether the actual topic information falls into the inferred topic information, and if so, the actual call information matches the inferred call information corresponding to the customer image and the customer type. If not, the actual call information does not match the inferred call information corresponding to the customer image and the customer type.

[0011] In some embodiments, generating the warning signal and intercepting the service includes: Obtain target topic information that does not belong to the inferred topic information from the actual topic information, determine the warning level corresponding to the target topic information, generate a corresponding warning signal based on the warning level, and block the service using the corresponding blocking level.

[0012] In some embodiments, after determining the number corresponding to the historical service information as an abnormal number, the method further includes: An alarm signal is generated and sent to the customer and the corresponding supervisor, and the abnormal number is not called within a preset time period.

[0013] In some embodiments, the method further includes: If the intercepted information is inconsistent with the audit result, the preset audit model and / or the customer call analysis model are adjusted based on the difference information.

[0014] Secondly, this embodiment provides a business risk control early warning and interception system, the system comprising: a pre-call module, a call in-call module, and an early warning and interception module; wherein, The pre-call module is used to obtain customer information, determine the customer image, customer type and historical business information based on the customer information; determine whether there are abnormal business in the historical business information, and if so, determine the number corresponding to the historical business information as an abnormal number; The in-call module is used to, if the number does not exist, call out the number corresponding to the historical service information, receive the actual call information corresponding to the number, and determine whether the actual call information matches the inferred call information corresponding to the customer image and the customer type; if it matches, generate a sustain signal to continue receiving the actual call information corresponding to the number. The warning and interception module is used to generate a warning signal and intercept the service if it does not meet the requirements.

[0015] Thirdly, this embodiment provides a computer-readable storage medium storing a computer program that can run on a processor, wherein the computer program, when executed by the processor, implements a business risk control early warning and interception method as described in the first aspect.

[0016] By employing the above method, this application first obtains customer information, and then determines the customer image, customer type, and historical business information based on this information. This combination of accurate customer information to obtain the customer image and customer type, taking into account customer personality and other inherent characteristics, facilitates a more accurate determination of the customer's inferred call information. This indirectly and more accurately determines whether business risk control warnings and interceptions are necessary, reducing malicious interference from illegal and criminal activities such as telecommunications harassment and fraud, as customer characteristics can significantly impact business operations. Furthermore, it allows for the priority use of historical business information to further determine whether the business needs to be warned and intercepted, enabling proactive and timely identification of illegal and criminal activities such as telecommunications harassment and fraud.

[0017] Then, it checks if there are any abnormal transactions in the historical service information. If so, the number corresponding to the historical service information is identified as an abnormal number. Before any service is initiated, i.e., before a call is made, it proactively identifies illegal and criminal activities such as telecommunications harassment and fraud to reduce malicious disruptions caused by these activities.

[0018] If the number does not exist, the system calls the number corresponding to the historical service information, receives the actual call information corresponding to the number, and determines whether the actual call information matches the inferred call information corresponding to the customer image and customer type. If it matches, a sustain signal is generated to continue receiving the actual call information corresponding to the number; if it does not match, a warning signal is generated and the service is blocked. During service operation, i.e., during a call, by comparing the customer's actual call information with the accurate inferred call information obtained by combining the customer image and customer type, the system proactively and accurately identifies illegal and criminal activities such as telecommunications harassment and telecommunications fraud, thereby reducing malicious interference from such illegal and criminal activities. Attached Figure Description

[0019] Figure 1 This is a flowchart of a business risk control early warning and interception method provided in this application.

[0020] Figure 2 This is a flowchart illustrating the method provided in this application for determining customer images, customer types, and historical business information based on customer information.

[0021] Figure 3 This is a flowchart of a method for determining whether there are abnormal business transactions in historical business information, as provided in this application.

[0022] Figure 4 This is a flowchart of a method provided in this application for determining whether actual call information matches the inferred call information corresponding to the customer image and customer type.

[0023] Figure 5 This is a connection diagram of a business risk control early warning and interception system provided in this application. Detailed Implementation

[0024] To better understand the purpose, technical solutions, and advantages of this application, it has been described and illustrated below with reference to the accompanying drawings and embodiments. However, those skilled in the art should understand that this application can be implemented without these details. It will be apparent to those skilled in the art that various modifications can be made to the embodiments disclosed in this application, and the general principles defined in this application can be applied to other embodiments and application scenarios without departing from the principles and scope of this application. Therefore, this application is not limited to the illustrated embodiments, but is consistent with the broadest scope claimed in this application.

[0025] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0026] The specific application scenario of this application is that telecommunications operators actively intercept abnormal telephone calls. Figure 1 This is a flowchart of a business risk control early warning and interception method provided in this application. For example... Figure 1 As shown, a business risk control early warning and interception method includes the following steps: Step S100: Obtain customer information and determine customer image, customer type, and historical business information based on the customer information.

[0027] This application describes the interception mechanism used for business risk control and early warning interception. The aforementioned customer information refers to information registered with the interception mechanism. This customer information includes the customer's industry, the industry of the target group, the customer's purpose, and the customer's phone number. Specifically, when a customer receives a phone number, they must first register the number with the interception mechanism before using it to make calls. During registration, the following information needs to be recorded: the phone number (customer number); the user's purpose (customer purpose); the user's industry (customer industry); and the industry of the target group (target group industry). For subsequent calls made using this phone number, customer information can be retrieved directly from the registered information using this phone number.

[0028] Specifically, if a phone number changes in at least one of the following during its use: the customer's industry, the industries of the groups calling, or the customer's intended use, the customer is required to proactively update the information stored on the interceptor. Failure to update this information promptly will result in discrepancies between the actual call information received and the inferred call information, thus impacting the smooth execution of normal business operations. Therefore, subsequent risk control measures also require customers to update their customer information in a timely manner.

[0029] After obtaining customer information, and before determining the customer's image, type, and historical business information based on that information, the interceptor will also review the obtained customer information. This involves verifying the accuracy of the customer information and whether it conforms to normal business practices. Specifically, it determines whether the customer's industry, the industry of the calling group, and the customer's purpose correspond to each other and conform to preset purposes and industries. If so, a correct filing signal is generated; if not, a re-filing signal is generated and sent to the customer to re-obtain customer information.

[0030] Specifically, the interceptor stores preset uses and preset industries, as well as tables corresponding to customer industries, calling group industries, and customer uses. Each table contains several corresponding relationships. It can be checked whether any relationship covers the customer industry, calling group industry, and customer use in the obtained customer information. If so, it indicates that the customer industry, calling group industry, and customer use correspond to each other. If not, it indicates that the customer industry, calling group industry, and customer use do not correspond to each other. The preset uses stored in the interceptor correspond to multiple uses representing normal business, and the preset industries correspond to multiple industries representing normal business. It can be checked whether the preset industries and preset uses contain the obtained customer industry, calling group industry, and customer use. If they do, it indicates that the customer industry and calling group industry match the preset industries, and the customer use matches the preset uses. If at least one of them is not included, it indicates that it does not match the preset uses and / or preset industries.

[0031] When the customer's industry, the industry of the calling group, and the customer's purpose are mutually consistent and conform to the preset purpose and industry, it indicates that the customer's registered information meets the requirements for normal business operations. At this point, a correct registration signal is generated, allowing the customer to continue conducting business. Conversely, when the customer's industry, the industry of the calling group, and the customer's purpose are not mutually consistent and / or do not conform to the preset purpose and / or industry, it indicates that the customer's registered information does not meet the requirements for normal business operations. In this case, a maintenance signal is generated and sent to the customer, prompting the customer to re-register with the interceptor, thus repeatedly acquiring customer information. This self-verification of customer information reduces the occurrence of further non-compliant business operations, thereby indirectly and proactively identifying illegal and criminal activities such as telecommunications harassment and fraud. This reduces malicious interference from such activities, as customers intending to engage in telecommunications harassment or fraud would likely have their registered customer information not falling within the preset purpose or industry, or exhibiting a mismatch between the two. Therefore, proactive identification of some illegal and criminal activities such as telecommunications harassment and fraud is possible.

[0032] Figure 2 This is a flowchart illustrating the method provided in this application for determining customer images, customer types, and historical business information based on customer information. For example... Figure 2 As shown, determining customer images, customer types, and historical business information based on customer information includes the following steps: Step S101: Obtain the industry image corresponding to the customer's industry, the dialing image corresponding to the dialing group's industry, and the usage image corresponding to the customer's purpose from the preset image database. Determine whether there is a conflict between the industry image, dialing image, and usage image. If not, integrate the industry image, dialing image, and usage image to obtain the customer image.

[0033] Step S102: If it exists, determine the adjustment image based on the industry image, dialing image, and usage image, and integrate the adjustment image to obtain the customer image.

[0034] Step S103: Obtain the single-dimensional types corresponding to the customer industry, the calling group industry, and the customer purpose, and determine the type that appears most frequently among all single-dimensional types as the customer type.

[0035] Step S104: Use the customer number to retrieve the historical service information corresponding to the customer number from the call database.

[0036] The aforementioned preset image database refers to a database composed of common feature images from all historical actual images corresponding to each customer's industry, common feature images from all historical actual images corresponding to each caller group's industry, and common feature images from all historical actual images corresponding to each customer's purpose industry. Specifically, after obtaining and verifying customer information, the interceptor can substitute the customer's industry from the customer information into the preset database to obtain the industry image corresponding to that customer's industry; substitute the caller group's industry from the customer information into the preset database to obtain the call image corresponding to that caller group's industry; and substitute the customer's purpose from the customer information into the preset database to obtain the purpose image corresponding to that customer's purpose. Here, the image specifically refers to a user's facial image.

[0037] Next, the industry image, dialing image, and application image are compared to see if they all correspond to the same feature for the same body part. If so, then there is no conflict between the industry image, dialing image, and application image. If not, then there is a conflict. For example, for the nose, the industry image, dialing image, and application image all correspond to the feature of a prominent nose, regardless of the degree of prominence. This is considered as the industry image, dialing image, and application image all corresponding to the same feature for the same body part. If at least one of the industry image, dialing image, and application image corresponds to a flat nose instead of a prominent nose, then it is considered that the industry image, dialing image, and application image do not all correspond to the same feature for the same body part.

[0038] If there are no conflicts between the industry image, dialing image, and application image, then the obtained industry image, dialing image, and application image can be directly integrated to obtain an image for each part, i.e., a customer image.

[0039] When conflicts arise between the industry image, call image, and purpose image, the first step is to identify which parts conflict through the aforementioned comparison. Then, based on the principle of majority rule, the actual image corresponding to that part is determined. If the majority rule cannot determine the actual image corresponding to that part, the characteristics corresponding to that part are sent to the customer to receive feedback. This customer feedback is then used as the actual image corresponding to that part, and the industry image, call image, and purpose image are adjusted accordingly to obtain the adjusted image. These three adjusted images must not conflict with each other. Finally, these three adjusted images are integrated to obtain the customer image. This method of obtaining a customer image by combining accurate customer information, taking into account the customer's personality and other characteristics, facilitates a more accurate determination of the customer's inferred call information. This indirectly and more accurately determines whether business risk control and early warning interception are needed, reducing malicious interference from illegal and criminal activities such as telecommunications harassment and fraud, as the customer's own characteristics will have a certain impact on business operations. Using images allows for a more intuitive understanding of the customer's personality and other characteristics.

[0040] The customer industry, call group industry, and customer purpose obtained above each correspond to at least one single-dimensional type. The interceptor stores the single-dimensional types corresponding to each historical customer industry, historical call group industry, and historical customer purpose. These single-dimensional types can be obtained by viewing the information stored on the interceptor. Then, information belonging to the same type among all obtained single-dimensional types is grouped, and this type is designated as the group name. Next, the number of information items in each group is read, and the type corresponding to the group with the largest number is determined as the customer type. If there are at least two groups with the largest number, the group names are sent to the customer to receive feedback and confirm the customer type. This method of combining accurate customer information to obtain the customer type, and taking the customer type into account, facilitates a more accurate determination of the customer's inferred call information, thereby indirectly and more accurately determining whether business risk control and early warning interception are necessary. This reduces malicious interference from illegal and criminal activities such as telecommunications harassment and fraud, as the customer type has a certain impact on business operations.

[0041] The interceptor stores call information for each number, allowing it to retrieve historical service information for that customer number from a database. This historical service information specifically refers to the historical call information of the corresponding customer number, comprising several sub-information entries, each corresponding to one call. This facilitates prioritizing the use of historical service information to further determine whether a service needs to be blocked, enabling proactive and timely identification of illegal activities such as telecommunications harassment and fraud.

[0042] Step S200: Determine whether there are any abnormal services in the historical service information. If so, identify the number corresponding to the historical service information as an abnormal number.

[0043] Figure 3 This is a flowchart illustrating the method provided in this application for determining whether historical business information contains abnormal business activity. For example... Figure 3 As shown, determining whether there are any abnormal transactions in historical business information includes the following steps: Step S201: Send historical business information to the preset audit model to obtain audit results, and obtain the interception information of the intercepted historical business sub-information among all historical business sub-information contained in the historical business information.

[0044] Step S202: Determine whether the intercepted information is consistent with the audit result. If they are consistent, obtain the actual number of times the interception was carried out and determine whether the actual number of times exceeds the preset number of times. If it exceeds the preset number of times, there is abnormal business in the historical business information.

[0045] Step S203: If the number of cases does not exceed the limit, there are no abnormal business transactions in the historical business information.

[0046] Step S204: If there is a discrepancy, obtain the difference information, and adjust the audit results and / or blocking information based on the difference information to obtain the adjustment information.

[0047] Step S205: Obtain the actual number of times the adjustment information was blocked, and determine whether the actual number of times exceeds the preset number. If it does, there is abnormal business in the historical business information.

[0048] Step S206: If the number of cases does not exceed the limit, there are no abnormal business transactions in the historical business information.

[0049] The aforementioned pre-set review model was trained offline using historical business information and corresponding correct review results from a large number of revoked phone numbers. Once the historical business information is obtained, it is sent to the pre-set review model for processing to obtain the corresponding review results. These results mark which historical business sub-information segments need to be blocked. Simultaneously, the blocking endpoint stores the results of each blocking action for that number. By viewing the information stored on the blocking endpoint, the blocking information for all blocked historical business sub-information segments within the historical business information can be obtained. This blocking information is determined autonomously by the blocking endpoint.

[0050] To accurately verify whether there are any abnormal transactions in the historical service information of this number, the review results obtained by the model and the interception information obtained by the interception terminal need to be compared. If they match, it indicates that both the review results and the interception information are correct. Then, the review results or interception information can be used to further examine the actual number of times the number has been blocked, and this actual number is compared with a preset number. If the actual number exceeds the preset number, it indicates that the number needs to be marked, as it has been involved in illegal or criminal activities such as telecommunications harassment or telecommunications fraud, meaning there are abnormal transactions in the historical service information. If the actual number does not exceed the preset number, it indicates that the number does not need to be marked, as it is uncertain whether there are any illegal or criminal activities such as telecommunications harassment or telecommunications fraud, meaning there are no abnormal transactions in the historical service information. The preset number refers to the number of times the system determines whether the number needs to be blacklisted.

[0051] If the audit results and interception information are inconsistent, it indicates that at least one of them is incorrect. In this case, the audit results and interception information can be sent to the auditing end, allowing them to manually verify the discrepancies and determine the necessary adjustments to the audit results and / or interception information. This discrepancy information is then used to adjust the audit results and / or interception information to obtain the correct adjustments. Finally, the actual number of interceptions corresponding to this adjusted information is obtained and compared to a preset number. If the actual number is greater than the preset number, the historical business information is considered to contain abnormal business activity. If the actual number is not greater than the preset number, the historical business information is considered to not contain abnormal business activity. This means that the audit results obtained by the model and the interception information obtained by the intercepting end may both be incorrect, one may be correct, or both may be correct. Determining the existence of abnormal business activity solely based on one of the audit results or interception information would reduce the accuracy of judging whether historical business activity contains abnormal activity, thus indirectly reducing the accuracy of subsequent interception decisions. Combining the audit results and interception information improves the accuracy of judging whether historical business activity contains abnormal activity, thereby indirectly improving the accuracy of subsequent interception decisions.

[0052] When abnormal business activity is identified in historical service information, the corresponding phone number is directly designated as an abnormal number, effectively restricting its subsequent business operations. This reduces malicious disruptions from telecommunications harassment, fraud, and other illegal activities. In this way, telecommunications harassment and fraud are proactively identified before any business activity is initiated, i.e., before a call is made, thus minimizing malicious disruptions from such activities.

[0053] Step S300: If the number does not exist, call out the number corresponding to the historical service information, receive the actual call information corresponding to the number, and determine whether the actual call information matches the inferred call information corresponding to the customer image and customer type.

[0054] Step S400: If the condition is met, a sustain signal is generated to continue receiving the actual call information corresponding to the number.

[0055] If the condition is not met in step S500, a warning signal is generated and the service is blocked.

[0056] If it is determined that there are no abnormal transactions in the historical service information, it means that it is impossible to identify whether the call is for illegal or criminal activities such as telecommunications harassment or telecommunications fraud. At this point, the caller is allowed to continue the service, that is, to call the number corresponding to the historical service information and then listen to the actual call information of the number after the call is connected. Based on the actual call information received, it is determined in real time whether the service needs to be blocked. Figure 4 This is a flowchart illustrating the method provided in this application for determining whether actual call information matches the inferred call information corresponding to the customer image and customer type. For example... Figure 4 As shown, determining whether the actual call information matches the inferred call information corresponding to the customer image and customer type includes the following steps: Step S301: Determine the general call analysis model corresponding to the customer from the preset analysis model based on the customer type.

[0057] Step S302: Adjust the weight parameters representing personality in the general call analysis model using customer images to obtain the customer call analysis model.

[0058] Step S303: Send the customer type to the customer call analysis model to obtain inference call information.

[0059] Step S304: Obtain the actual topic information in the actual call information and the inferred topic information in the inferred call information; determine whether the actual topic information falls into the inferred topic information; if so, the actual call information matches the inferred call information corresponding to the customer image and customer type.

[0060] Step S305: If not included, the actual call information does not match the inferred call information corresponding to the customer image and customer type.

[0061] The aforementioned preset analysis model refers to the model used to analyze call information when customers conduct business. Each customer type corresponds to a preset analysis model, stored at the interception end. A general call analysis model corresponding to the customer type obtained above can be used. This general call analysis model is universal and applicable to this type of customer analysis. However, to more accurately analyze this business, the customer image can be used to adjust the weight parameters representing personality in the general call analysis model. After all, customers of the same type may have different personalities, and customers with different personalities will use different methods when conducting business. At this point, the customer's personality can be determined by viewing the customer image, and then the weight parameters representing personality in the general call analysis model can be adjusted accordingly, i.e., increasing the value of the weight parameter corresponding to that personality in the personality weight parameters. Originally, the weight parameters representing personality in the general call analysis model are all consistent. This yields the customer call analysis model corresponding to this customer, allowing the customer call analysis model to more accurately obtain the call information theoretically expected for this business. Then, the customer type is sent to the customer call analysis model, allowing the model to process the customer type to obtain the theoretical call information for this business. Among them, theoretical call information represents information that customers of this type should be able to make calls when conducting business, provided they do not engage in illegal or criminal activities such as telecommunications harassment or telecommunications fraud.

[0062] Next, information extraction technology is used to summarize the actual topics involved in the actual call information from the actual call information, and to summarize the inference topics involved in the inference call information from the inference call information. If the actual topic information falls into the inference topic information, it indicates that the customer is conducting business normally and there is no illegal or criminal activity such as telecommunications harassment or telecommunications fraud. At this point, a sustain signal is generated to continue receiving the actual call information corresponding to that number.

[0063] If the actual topic information does not fall within the inferred topic information, it indicates that the customer is not conducting normal business and is engaging in illegal or criminal activities such as telecommunications harassment or fraud. This generates a warning signal and blocks the service, thereby proactively identifying and reducing malicious disruptions caused by telecommunications harassment and fraud. By combining customer type and customer profile, more accurate inferred call information can be obtained, thus improving the accuracy of proactively identifying and reducing malicious disruptions caused by telecommunications harassment and fraud.

[0064] The process of generating early warning signals and intercepting services includes: obtaining target topic information that is not part of the inferred topic information, determining the early warning level corresponding to the target topic information, generating a corresponding early warning signal based on the early warning level, and intercepting the service using the corresponding interception level.

[0065] By comparing the actual topic information with the inferred topic information, we can determine which topics in the actual topic information do not belong to the inferred topic information, and thus identify these topics as target topics. The interceptor stores a warning level table corresponding to different topics. By consulting this warning level table, we can determine the warning level corresponding to the target topic information, and generate a warning signal according to the highest warning level obtained, thereby blocking the service and minimizing the malicious interference from illegal and criminal activities such as telecommunications harassment and fraud.

[0066] Preferably, after identifying the number corresponding to the historical business information as an abnormal number, the method further includes: generating an alarm signal, sending the alarm signal to the customer and the corresponding regulatory personnel, and refraining from calling the abnormal number within a preset time period. This serves to punish the customer, indirectly reducing the likelihood of them engaging in illegal and criminal activities such as telecommunications harassment and fraud, and minimizing malicious disruption.

[0067] Preferably, if the intercepted information is inconsistent with the audit results, the preset audit model and / or customer call analysis model are adjusted based on the abnormal information.

[0068] Specifically, if the intercepted information does not match the pre-screening results, it indicates that the accuracy of the preset review model is poor, and / or the accuracy of the results obtained by the interceptor through the customer call analysis model is poor. In this case, check whether the discrepancy corresponds to a review result or intercepted information. If only a review result is present, it indicates that the accuracy of the preset review model is poor, and the preset review model should be adjusted based on the anomaly. If only intercepted information is present, it indicates that the accuracy of the results obtained by the preset interceptor through the customer call analysis model is poor, and the general call analysis model corresponding to the customer should be adjusted based on the anomaly. If both review results and intercepted information are present, it indicates that the accuracy of both the preset review model and the results obtained by the preset interceptor through the customer call analysis model is poor, and the preset review model and the general call analysis model corresponding to the customer should be adjusted based on the anomaly. This facilitates the accuracy of subsequent business risk control and early warning interception, thereby ensuring the feasibility of proactively identifying illegal and criminal activities such as telecommunications harassment and fraud, and reducing malicious interference from such activities.

[0069] Figure 5 This is a connection diagram of a business risk control early warning and interception system provided in this application. For example... Figure 5As shown, a business risk control early warning and interception system includes: a pre-call module, a call-in-process module, and an early warning and interception module.

[0070] The system comprises several modules: a pre-call module, which acquires customer information and determines the customer's image, type, and historical service information; it also checks for any abnormal services in the historical service information and, if found, identifies the corresponding number as an abnormal number. The in-call module, if the abnormal number is not found, calls the number corresponding to the historical service information, receives the actual call information, and determines whether the actual call information matches the inferred call information based on the customer's image and type; if it matches, it generates a sustain signal to continue receiving the actual call information. The warning and interception module, if the call does not match, generates a warning signal and intercepts the service.

[0071] The other functions performed by the pre-call module, the call-in-call module, and the early warning and interception module, as well as the technical details of each function, are the same as or similar to the corresponding features in the business risk control early warning and interception method described above, so they will not be repeated here.

[0072] This application also provides a computer storage medium storing a computer program that, when run on a computer, enables the computer to execute the steps in the business risk control early warning and interception method described above.

[0073] It should be understood that although the steps in the flowcharts in the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order requirement for the execution of these steps, and they can be performed in other orders.

[0074] The above description is only a partial embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A business risk control early warning and interception method, characterized in that, The method includes: Obtain customer information, and determine customer image, customer type, and historical business information based on the customer information; Determine whether there are any abnormal transactions in the historical service information. If so, identify the number corresponding to the historical service information as an abnormal number. If not, call out the number corresponding to the historical service information, receive the actual call information corresponding to the number, and determine whether the actual call information matches the inferred call information corresponding to the customer image and the customer type. If the condition is met, a sustain signal is generated to continue receiving the actual call information corresponding to the number; If the conditions are not met, a warning signal is generated and the service is blocked.

2. The method according to claim 1, characterized in that, The customer information represents the information recorded at the interception terminal. The customer information includes the customer's industry, the industries of the calling group, and the customer's purpose. After obtaining the customer information, the process further includes: Determine whether the customer industry, the calling group industry, and the customer purpose correspond to each other and conform to the preset purpose and preset industry. If so, generate a correct filing signal. If not, generate a re-registration signal and send it to the customer to retrieve the customer information again.

3. The method according to claim 2, characterized in that, The customer information also includes a customer number, and determining the customer image, customer type, and historical business information based on the customer information includes: The system retrieves industry images corresponding to the customer's industry, dialing images corresponding to the dialing group's industry, and usage images corresponding to the customer's purpose from a preset image database. It then determines whether there are any conflicts between the industry images, dialing images, and usage images. If there are no conflicts, the system integrates the industry images, dialing images, and usage images to obtain the customer image. If present, determine an adjustment image based on the industry image, dialing image, and usage image, and integrate the adjustment image to obtain a customer image; Obtain the single-dimensional types corresponding to the customer industry, the calling group industry, and the customer purpose, and determine the type that appears most frequently among all single-dimensional types as the customer type; Use the customer number to retrieve the historical service information corresponding to the customer number from the call database.

4. The method according to claim 1, characterized in that, The historical service information includes several historical service sub-information entries, and determining whether the historical service information contains abnormal services includes: Send the historical business information to a preset audit model to obtain the audit result, and obtain the interception information of the intercepted historical business sub-information among all the historical business sub-information contained in the historical business information; Determine whether the interception information is consistent with the audit result. If they are consistent, obtain the actual number of times the interception was carried out. Determine whether the actual number of times exceeds the preset number of times. If it exceeds the preset number of times, the historical business information indicates abnormal business. If the number does not exceed the limit, the historical service information does not contain any abnormal services; If there is a discrepancy, obtain the difference information, and adjust the audit results and / or blocking information based on the difference information to obtain the adjustment information; Obtain the actual number of times the adjustment information was blocked, and determine whether the actual number of times exceeds the preset number. If it does, the historical service information indicates abnormal service. If the number of cases does not exceed a certain threshold, the historical service information does not contain any abnormal services.

5. The method according to claim 4, characterized in that, The step of determining whether the actual call information matches the inferred call information corresponding to the customer image and the customer type includes: Based on the customer type, determine the general call analysis model corresponding to the customer from the preset analysis model; The customer image is used to adjust the weight parameters representing personality in the general call analysis model to obtain the customer call analysis model. The customer type is sent to the customer call analysis model to obtain inferred call information; Obtain the actual topic information in the actual call information and the inferred topic information in the inferred call information, determine whether the actual topic information falls into the inferred topic information, and if so, the actual call information matches the inferred call information corresponding to the customer image and the customer type. If not, the actual call information does not match the inferred call information corresponding to the customer image and the customer type.

6. The method according to claim 5, characterized in that, The generation of the early warning signal and the interception of the service include: Obtain target topic information that does not belong to the inferred topic information from the actual topic information, determine the warning level corresponding to the target topic information, generate a corresponding warning signal based on the warning level, and block the service using the corresponding blocking level.

7. The method according to claim 1, characterized in that, After determining the number corresponding to the historical business information as an abnormal number, the process also includes: An alarm signal is generated and sent to the customer and the corresponding supervisor, and the abnormal number is not called within a preset time period.

8. The method according to claim 5, characterized in that, The method further includes: If the intercepted information is inconsistent with the audit result, the preset audit model and / or the customer call analysis model are adjusted based on the difference information.

9. A business risk control early warning and interception system, characterized in that, The system includes: a pre-call module, a call-in-call module, and an early warning and interception module; wherein... The pre-call module is used to obtain customer information, determine the customer image, customer type and historical business information based on the customer information; determine whether there are abnormal business in the historical business information, and if so, determine the number corresponding to the historical business information as an abnormal number; The in-call module is used to, if the number does not exist, call out the number corresponding to the historical service information, receive the actual call information corresponding to the number, and determine whether the actual call information matches the inferred call information corresponding to the customer image and the customer type; if it matches, generate a sustain signal to continue receiving the actual call information corresponding to the number. The warning and interception module is used to generate a warning signal and intercept the service if it does not meet the requirements.

10. A computer-readable storage medium having a computer program stored thereon that can run on a processor, characterized in that, When the computer program is executed by the processor, it implements a business risk control early warning and interception method as described in any one of claims 1 to 8.

Images