Processing method and device of insurance policy correction application, electronic equipment and storage medium

Abstract

The embodiment of the invention provides an insurance policy correction application processing method and device, electronic equipment and a storage medium, belongs to the technical field of data processing, and is suitable for the field of finance. The method comprises the following steps: acquiring insurance policy correction application information, insurance applicant information and historical claim settlement records of a target user; performing insurance policy modification operation risk assessment according to the insurance policy correction application information to obtain a risk operation assessment vector; performing user portrait deviation degree evaluation according to the insurance applicant information and the insurance policy correction application information to obtain a risk portrait evaluation vector; performing claim settlement anti-fraud evaluation according to the insurance policy correction application information and the historical claim settlement record to obtain a risk claim settlement evaluation vector; screening preset application processing schemes based on the risk operation evaluation vector, the risk portrait evaluation vector and the risk claim settlement evaluation vector to obtain a target processing scheme; and processing the insurance policy correction application information according to the target processing scheme. According to the embodiment of the invention, the accuracy of policy correction application processing can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and is suitable for the financial field, in particular to a processing method and device for policy amendment application, an electronic device and a storage medium. BACKGROUND

[0002] The policy amendment application refers to the application for modifying the content of the policy within the effective period of the policy. For example, adjusting the guarantee responsibility, modifying the premium, converting the payment method, etc. In the prior art, the processing of the policy amendment application mostly depends on manual review, and the risk assessment and processing decision are made by the underwriting personnel. Due to the variety of amendment types, there are often inconsistencies in the judgment criteria in the manual processing process, which easily leads to inaccurate judgment of the risk degree of the amendment application, and further gives unreasonable processing schemes. Therefore, how to improve the accuracy of the processing of the policy amendment application has become a technical problem to be solved. SUMMARY

[0003] The main purpose of the embodiments of the present application is to provide a processing method and device for policy amendment application, an electronic device and a storage medium, which aims to improve the accuracy of the processing of the policy amendment application.

[0004] To achieve the above-mentioned purpose, the first aspect of the embodiments of the present application provides a processing method for policy amendment application, which comprises:

[0005] Obtaining the policy amendment application information of a target user, obtaining the policyholder information and the historical claim record of the target user;

[0006] Performing risk assessment of policy modification operation according to the policy amendment application information, to obtain a risk operation assessment vector;

[0007] Performing user portrait deviation assessment according to the policyholder information and the policy amendment application information, to obtain a risk portrait assessment vector;

[0008] Performing anti-fraud assessment of claim according to the policy amendment application information and the historical claim record, to obtain a risk claim assessment vector;

[0009] Filtering a preset application processing scheme based on the risk operation assessment vector, the risk portrait assessment vector and the risk claim assessment vector, to obtain a target processing scheme;

[0010] Performing application processing on the policy amendment application information according to the target processing scheme.

[0011] In some embodiments, the policy amendment application information includes premium adjustment data, guarantee liability change data, amendment application frequency data, and application operation attribute data; the risk operation assessment vector is obtained by performing risk operation assessment on the policy amendment application information, including:

[0012] The premium adjustment risk index is obtained by matching the premium adjustment data with a preset premium adjustment index;

[0013] The liability adjustment risk index is obtained by matching the guarantee liability change data with a preset liability adjustment index;

[0014] The application batch risk index is obtained by matching the amendment application frequency data with a preset adjustment threshold index;

[0015] The application operation risk index is obtained by matching the application operation attribute data with a preset operation risk index;

[0016] The risk operation assessment vector is obtained by vector construction based on the premium adjustment risk index, the liability adjustment risk index, the application batch risk index, and the application operation risk index.

[0017] In some embodiments, the application operation attribute data includes step advancement time data, sensitive field switching frequency, application initiation address data, application initiation time data, and application initiation device data; the application operation risk index is obtained by matching the application operation attribute data with a preset operation risk index, including:

[0018] The advancement time index is obtained by matching the step advancement time data with the preset operation risk index;

[0019] The field switching index is obtained by matching the sensitive field switching frequency with the preset operation risk index;

[0020] The initiation address index is obtained by matching the application initiation address data with the preset operation risk index;

[0021] The initiation time index is obtained by matching the application initiation time data with the preset operation risk index;

[0022] The initiation device index is obtained by matching the application initiation device data with the preset operation risk index;

[0023] The application operation risk index is obtained by vector construction based on the advancement time index, the field switching index, the initiation address index, the initiation time index, and the initiation device index.

[0024] In some embodiments, performing a user profile deviation assessment based on the policyholder information and the policy modification application information to obtain a risk profile assessment vector includes:

[0025] Acquire reference user information, and perform user clustering based on the policyholder information and the reference user information to obtain a target clustering area; wherein the policyholder information is located in the target clustering area;

[0026] Obtaining user information of a target cluster area, obtaining target user information, and obtaining historical revision application information of the target user information;

[0027] Comparing the policy revision application information with the historical revision application information and a preset application abnormality threshold indicator to obtain abnormal revision application information;

[0028] The abnormal correction application information is vectorized to obtain the risk profile assessment vector.

[0029] In some embodiments, the policy modification application information includes at least one policy field and a policy field modification record of the policy field; performing a claim anti-fraud assessment based on the policy modification application information and the historical claim records to obtain a risk claim assessment vector includes:

[0030] Obtaining sample revision application information and sample claim records; wherein the sample revision application information includes at least one sample field modification record, and the sample field modification record corresponds to the policy field modification record;

[0031] Performing self-attention calculation on the sample field modification record and the sample claim record through a preset attention model, determining a sample claim association value of each sample field modification record, and using the sample claim association value of the sample field modification record as the field claim association value of the policy field modification record corresponding to the sample field modification record;

[0032] An aggregate calculation is performed based on the field claim association value, the policy field, and the historical claim records to obtain the risk claim assessment vector.

[0033] In some embodiments, the historical claim records include claim record text, and performing aggregation calculation based on the field claim association value, the policy field, and the historical claim records to obtain the risk claim assessment vector includes:

[0034] Performing field extraction on the claim record text according to a preset field extraction model to obtain matching fields;

[0035] assigning a first numerical value to a field in the policy field that matches the matching field and a second numerical value to a field in the policy field that does not match the matching field;

[0036] performing a weighted calculation according to the field claim correlation value and the numerical value in the policy field to determine a risk claim evaluation value of each of the policy fields;

[0037] performing vectorization processing according to each of the risk claim evaluation values to obtain the risk claim evaluation vector.

[0038] In some embodiments, the target processing scheme is obtained by screening a preset application processing scheme based on the risk operation evaluation vector, the risk portrait evaluation vector and the risk claim evaluation vector, including:

[0039] performing vector splicing on the risk operation evaluation vector, the risk portrait evaluation vector and the risk claim evaluation vector to obtain a spliced evaluation vector;

[0040] performing risk rating on the spliced evaluation vector through a preset random forest model to obtain a risk grade label;

[0041] screening the application processing scheme according to the risk grade label to obtain the target processing scheme.

[0042] To achieve the above-mentioned purpose, a second aspect of the embodiment of the present application proposes a policy amendment application processing device, the device comprising:

[0043] an acquisition data module configured to acquire policy amendment application information of a target user, and acquire policyholder information and historical claim records of the target user;

[0044] an operation evaluation module configured to perform policy modification operation risk evaluation according to the policy amendment application information to obtain a risk operation evaluation vector;

[0045] a deviation evaluation module configured to perform user portrait deviation evaluation according to the policyholder information and the policy amendment application information to obtain a risk portrait evaluation vector;

[0046] a fraud evaluation module configured to perform claim fraud evaluation according to the policy amendment application information and the historical claim records to obtain a risk claim evaluation vector;

[0047] a scheme screening module configured to screen a preset application processing scheme based on the risk operation evaluation vector, the risk portrait evaluation vector and the risk claim evaluation vector to obtain a target processing scheme;

[0048] An application processing module is configured to perform application processing on the policy amendment application information according to the target processing scheme.

[0049] To achieve the above object, a third aspect of the embodiments of the present application provides an electronic device, which comprises a memory and a processor, the memory stores a computer program, and the processor implements the method of the first aspect when executing the computer program.

[0050] To achieve the above object, a fourth aspect of the embodiments of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the method of the first aspect.

[0051] The method and device for processing a policy amendment application, the electronic device and the storage medium provided by the present application obtain policy amendment application information, applicant information and historical claim records; perform risk operation assessment on policy modification operation according to the policy amendment application information to obtain a risk operation assessment vector; perform user portrait deviation assessment according to the applicant information and the policy amendment application information to obtain a risk portrait assessment vector; perform anti-fraud assessment on claim according to the policy amendment application information and the historical claim records to obtain a risk claim assessment vector; filter a preset application processing scheme based on the risk operation assessment vector, the risk portrait assessment vector and the risk claim assessment vector to obtain a target processing scheme; and perform application processing on the policy amendment application information according to the target processing scheme. In this way, the embodiments of the present application comprehensively process the policy amendment application information, the applicant information and the historical claim records by introducing a multi-dimensional assessment mechanism, jointly assess the risk of policy modification operation, the user portrait deviation and the anti-fraud risk of claim, and then filter a processing scheme matching the risk level based on multiple risk assessment results, so as to improve the pertinence and rationality of the processing strategy, and finally process the amendment application according to the target processing scheme. The embodiments of the present application can accurately judge the rationality of the amendment application based on the cooperative assessment in different risk dimensions, significantly reduce the misjudgment rate caused by subjective judgment of artificial, and improve the standardization and accuracy of the processing of the policy amendment. BRIEF DESCRIPTION OF DRAWINGS

[0052] Figure 1 is a flowchart of the method for processing a policy amendment application provided by the embodiments of the present application;

[0053] Figure 2 is a flowchart of step S102 in Figure 1 ;

[0054] Figure 3 is a flowchart of step S204 in Figure 2 ;

[0055] Figure 4is a flowchart of step S103 in Figure 1 is a flowchart of step S104 in

[0056] Figure 5 is a flowchart of step S104 in Figure 1 is a flowchart of step S105 in

[0057] Figure 6 is a flowchart of step S503 in Figure 5 is a flowchart of step S503 in

[0058] Figure 7 is a flowchart of step S105 in Figure 1 is a flowchart of step S105 in

[0059] Figure 8 is a structural schematic diagram of a policy amendment application processing device provided by an embodiment of the present application;

[0060] Figure 9 is a hardware structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0061] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. 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.

[0062] It should be noted that although the functional modules are divided in the device schematic diagram, and the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a manner different from the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification and claims and the above-described drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence.

[0063] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application, and are not intended to limit the present application.

[0064] The policy amendment application refers to the application for modifying the content of the policy within the effective period of the policy. For example, adjusting the guarantee liability, modifying the premium, converting the payment method, etc. In the prior art, the processing of the policy amendment application mostly relies on manual review, and the risk assessment and processing decision are made by the underwriting personnel. Due to the variety of amendment types, there are often inconsistencies in the judgment criteria in the manual processing process, which easily leads to inaccurate judgment of the risk degree of the amendment application, and further gives unreasonable processing schemes. Therefore, how to improve the accuracy of the processing of the policy amendment application has become a technical problem to be solved.

[0065] Based on this, the embodiment of the application provides a processing method and device for a policy amendment application, an electronic device and a storage medium, aiming to improve the accuracy of policy amendment application processing.

[0066] The processing method and device for a policy amendment application, the electronic device and the storage medium provided by the embodiment of the application are specifically described through the following embodiment. First, the processing method for a policy amendment application in the embodiment of the application is described.

[0067] The embodiment of the application can acquire and process related data based on artificial intelligence technology. The artificial intelligence (AI) is a theory, method, technology and application system for simulating, extending and expanding human intelligence by using a digital computer or a machine controlled by a digital computer, perceiving an environment, acquiring knowledge and using the knowledge to obtain the best results.

[0068] The artificial intelligence basic technology generally includes technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics, etc. The artificial intelligence software technology mainly includes computer vision technology, robot technology, biometric technology, speech processing technology, natural language processing technology and machine learning / deep learning, etc.

[0069] The processing method for a policy amendment application provided by the embodiment of the application relates to the technical field of data processing and is suitable for the financial field. The processing method for a policy amendment application provided by the embodiment of the application can be applied to a terminal, can be applied to a server end, and can also be software running in a terminal or a server end. In some embodiments, the terminal can be a smart phone, a tablet computer, a notebook computer, a desktop computer, etc.; the server end can be configured as an independent physical server, can be configured as a server cluster or a distributed system composed of multiple physical servers, can also be configured as a cloud server providing basic cloud computing services such as cloud service, cloud database, cloud computing, cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN and big data and artificial intelligence platform, etc.; and the software can be an application for implementing the processing method for a policy amendment application, etc., but is not limited to the above forms.

[0070] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments in which tasks are performed by remote processing devices connected via a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media, including storage devices.

[0071] It should be noted that in each specific embodiment of the present application, when it comes to the need to perform relevant processing based on data related to the user's identity or characteristics, such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of such data will comply with relevant laws, regulations, and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0072] Figure 1 This is an optional flow chart of the method for processing an insurance policy amendment application provided in an embodiment of the present application. Figure 1 The method may include but is not limited to steps S101 to S106.

[0073] Step S101: Obtain the target user's insurance policy modification application information, and obtain the target user's policyholder information and historical claims records;

[0074] Step S102: Perform a risk assessment of the policy modification operation based on the policy modification application information to obtain a risk operation assessment vector;

[0075] Step S103: Perform user profile deviation assessment based on the policyholder information and the policy revision application information to obtain a risk profile assessment vector;

[0076] Step S104: performing a claim anti-fraud assessment based on the policy revision application information and historical claim records to obtain a risk claim assessment vector;

[0077] Step S105: Screening preset application processing solutions based on the risk operation assessment vector, the risk profile assessment vector, and the risk claim assessment vector to obtain a target processing solution;

[0078] Step S106: Process the policy revision application information according to the target processing solution.

[0079] In steps S101 to S106 shown in the embodiment of the present application, policy modification application information, policyholder information and historical claims records are obtained; a risk assessment of the policy modification operation is performed based on the policy modification application information to obtain a risk operation assessment vector; a user profile deviation assessment is performed based on the policyholder information and the policy modification application information to obtain a risk profile assessment vector; a claims anti-fraud assessment is performed based on the policy modification application information and historical claims records to obtain a risk claims assessment vector; a preset application processing plan is screened based on the risk operation assessment vector, the risk profile assessment vector and the risk claims assessment vector to obtain a target processing plan; and application processing is performed on the policy modification application information according to the target processing plan. In this way, the embodiment of the present application introduces a multi-dimensional evaluation mechanism to comprehensively process the policy modification application information with the insured information, historical claims records, etc., to achieve a joint evaluation of the policy modification operation risk, user profile deviation and claims anti-fraud risk, and then screen out a processing solution that matches the risk level based on multiple risk assessment results, thereby improving the pertinence and rationality of the processing strategy, and finally processing the modification application according to the target processing solution. The embodiment of the present application can accurately judge the rationality of the modification application based on the collaborative evaluation of different risk dimensions, significantly reduce the misjudgment rate caused by human subjective judgment, and improve the standardization and accuracy of the policy modification processing.

[0080] In step S101 of some embodiments, the target user refers to an applicant who initiates a policy modification application in the insurance business system.

[0081] Policy modification requests refer to the specific content of a user's request to modify an existing insurance contract. These requests typically include the type of modification, reason for the modification, and request date. For example, a policy modification request might indicate that the user wishes to change the vehicle number on their existing auto insurance policy from "12345" to "12321," with the reason being "the original vehicle has been sold."

[0082] Policyholder information refers to the basic information related to the target user, which is submitted and stored in the system during the policy signing process. It usually includes identity information, occupational information, contact information, risk preference level, historical insurance behavior, etc.

[0083] The historical claim record refers to the detailed data of the claim request initiated and processed by the target user or his / her related insurance policy within the effective period of the insurance contract, including information such as claim occurrence time, claim type, claim amount, claim conclusion, and whether it involves a dispute. For example, a historical claim record for target user Zhang San is that he filed a claim for a car insurance policy in May 2023 due to a traffic accident, with an application amount of 15,000 yuan, and the claim amount was 12,000 yuan after investigation and appraisal.

[0084] Please refer to Figure 2 In some embodiments, the policy amendment application information includes premium adjustment data, guarantee liability change data, amendment application frequency data, and application operation attribute data, and step S102 can include but is not limited to steps S201 to S205:

[0085] Step S201, matching the premium adjustment data with the preset premium adjustment index to obtain a premium adjustment risk index;

[0086] Step S202, matching the guarantee liability change data with the preset liability adjustment index to obtain a liability adjustment risk index;

[0087] Step S203, matching the amendment application frequency data with the preset adjustment threshold index to obtain an application batch risk index;

[0088] Step S204, matching the application operation attribute data with the preset operation risk index to obtain an application operation risk index;

[0089] Step S205, constructing a vector based on the premium adjustment risk index, the liability adjustment risk index, the application batch risk index, and the application operation risk index to obtain a risk operation evaluation vector.

[0090] The steps S201 to S205 shown in the embodiments of the present application match the premium adjustment data with the preset premium adjustment index to obtain a premium adjustment risk index; match the guarantee liability change data with the preset liability adjustment index to obtain a liability adjustment risk index; match the amendment application frequency data with the preset adjustment threshold index to obtain an application batch risk index; match the application operation attribute data with the preset operation risk index to obtain an application operation risk index; and construct a vector based on the premium adjustment risk index, the liability adjustment risk index, the application batch risk index, and the application operation risk index to obtain a risk operation evaluation vector. In this way, the embodiments of the present application introduce indexes corresponding to the premium adjustment data, the guarantee liability change data, the amendment application frequency data, and the application operation attribute data, respectively, so as to comprehensively and quantitatively consider the risk of the policy amendment application information.

[0091] In step S201 of some embodiments, the preset insured amount adjustment indicator refers to a reference value for determining potential risk behaviors in the insured amount adjustment data, based on historical statistical data, business rules, and regulatory standards. For example, the preset insured amount adjustment indicator may include a rule that "the adjustment ratio is greater than 50% and occurs within three days of the effective date."

[0092] The insurance premium adjustment risk indicator is a binary indicator that indicates whether the insurance premium adjustment behavior meets the risk characteristics. If any of the pre-set risk rules in the insurance premium adjustment indicator are met, the insurance premium adjustment risk indicator is 1, indicating that risk exists; otherwise, it is 0, indicating that no risk has been identified. For example, if a target user adjusts their auto insurance premium from 100,000 yuan to 200,000 yuan two days before the policy takes effect, this adjustment behavior meets the risk rule of "large-scale increase near the effective date", and the insurance premium adjustment risk indicator value is 1.

[0093] In step S202 of some embodiments, the preset liability adjustment indicator refers to a rule set to identify behaviors in the coverage change data that may constitute abnormal adjustments. The purpose is to identify risky operations caused by changes in coverage. Examples include situations where the liability insurance type is expanded from basic liability to additional high-value liability, the scope of liability is significantly broadened, and the adjustment time is unusually concentrated. For example, the preset liability adjustment indicator includes the rule of "adjusting the additional theft and robbery insurance from uninsured to insured, and submitting the application within 1 day before the policy takes effect."

[0094] The liability adjustment risk indicator is a binary result that indicates whether a change in coverage liability triggers a risk signature. If any of the preset liability adjustment indicators are met, the indicator is 1; otherwise, it is 0. For example, if a user adds theft and robbery insurance coverage the day before the policy takes effect, this action meets the risk rule, and the liability adjustment risk indicator value is 1.

[0095] In step S203 of some embodiments, the preset adjustment threshold indicator refers to a set of rules used to determine whether the frequency of revision requests is abnormal, and is constructed based on the frequency distribution of revision behaviors within the normal business scope. The preset adjustment threshold indicator is implemented by calculating the average number of revision requests by different users within a unit time period and setting a warning threshold for high-frequency behaviors based on business experience. For example, "more than three revision requests for the same policy within seven days" is considered an abnormal frequency. For example, the preset adjustment threshold indicator is "more than five revision requests within 30 days."

[0096] The application batch risk indicator is a binary result that indicates whether the revision frequency exceeds a reasonable threshold. If the target user's revision frequency exceeds the set threshold within a specified period, the application batch risk indicator is 1; otherwise, it is 0. For example, if a user submits 6 revision applications for the same policy within 30 days, the application batch risk indicator value is 1.

[0097] See also Figure 3In some embodiments, the application operation attribute data includes step advancement time data, sensitive field switching frequency, application initiation address data, application initiation time data, and application initiation device data. Step S204 may include, but is not limited to, steps S301 to S306:

[0098] Step S301, matching the step advancement time data with the preset operational risk indicator to obtain the advancement time indicator;

[0099] Step S302: Match the sensitive field switching frequency with a preset operational risk indicator to obtain a field switching indicator;

[0100] Step S303: Match the application initiation address data with the preset operational risk indicator to obtain an initiation address indicator;

[0101] Step S304: Match the application initiation time data with the preset operational risk indicator to obtain an initiation time indicator;

[0102] Step S305: Match the application initiating device data with the preset operational risk index to obtain the initiating device index;

[0103] Step S306 , constructing a vector based on the advancement time indicator, the field switching indicator, the initiation address indicator, the initiation time indicator, and the initiation device indicator to obtain the application operation risk indicator.

[0104] In the steps S301 to S306 shown in the embodiment of the present application, the step advancement time data, the sensitive field switching frequency, the application initiation address data, the application initiation time data, and the application initiation device data are matched with the preset operation risk indicators respectively, and then the matched indicators are vector-constructed to obtain the application operation risk indicators. In this way, the embodiment of the present application extracts sub-item indicators from the time dimension, field change dimension, network address dimension, behavior time dimension and device information dimension involved in the correction operation process, and performs risk feature matching, and finally constructs the application operation risk indicator vector. The abnormal advancement interval between the operation steps is characterized by the advancement time indicator, the frequent operation behavior of sensitive fields is measured by the field switching indicator, the abnormal regional access behavior is identified by the initiation address indicator, the night or non-business period behavior is identified by the initiation time indicator, and the suspicious device or unknown terminal behavior is screened by the initiation device indicator. The attribute information of the application initiation is evaluated in all aspects, thereby improving the audit and control level of the correction link.

[0105] In step S301 of some embodiments, the step promotion time data refers to the time interval information between key interaction steps in the policy amendment application process, including the relative promotion time of multiple operation nodes such as starting application, field filling, confirmation submission, etc. The step promotion time data reflects the operation rhythm and stability of the application behavior. The preset operation risk indicator includes a risk rule for identifying "step promotion anomaly", specifically: any continuous step promotion time interval is less than a preset threshold (such as 1 second), indicating that there may be non-manual natural operation or batch submission behavior driven by automatic script. The promotion time indicator is the result of judging whether the step promotion time data meets the promotion anomaly rule. If there is a time interval anomaly, the value of the promotion time indicator is 1, otherwise it is 0. For example, when the time interval between filling and submission operations in the amendment application is 0.5 seconds, the value of the promotion time indicator is 1.

[0106] In step S302 of some embodiments, the sensitive field switching frequency refers to the frequency statistics result of switching or repeatedly modifying behaviors of key fields such as beneficiary information, insurance subject, insurance amount, guarantee period, etc. in the policy amendment application process. The sensitive field switching frequency reflects the repeated operation of key elements in the application behavior. The preset operation risk indicator includes a risk rule for identifying "frequent field switching", specifically: the number of sensitive field switching times in a single application exceeds a preset threshold (such as 3 times). The field switching indicator is used to judge whether the sensitive field switching frequency meets the abnormal operation condition. If it exceeds the threshold, the value of the field switching indicator is 1, otherwise it is 0. For example, when the same user repeatedly modifies the insurance amount field 4 times before submitting the application, the value of the field switching indicator is 1.

[0107] In step S303 of some embodiments, the application initiation address data refers to the network address information recorded in the policy amendment application, including IP address, home address information, etc., which is used to identify the location source of operation initiation. This type of data is used to judge whether there is a regional abnormal access behavior. The preset operation risk indicator includes a risk rule for identifying "abnormal initiation address", specifically: the current application operation IP address is inconsistent with the registered address home address, or the IP address is in the high-risk list. The initiation address indicator is used to judge whether the initiation address data matches the address anomaly rule. If it matches, the value of the initiation address indicator is 1, otherwise it is 0. For example, when a user registers a policy in the inland, and the IP address of the application initiation is an overseas address, the value of the initiation address indicator is 1.

[0108] In step S304 of some embodiments, the application initiation time data refers to the specific time point information of the submission of the policy amendment application, including hours, weekdays, etc., for identifying the reasonableness of the operation time. The preset operation risk indicator includes a risk rule for identifying “abnormal operation time”, specifically: the operation time is in the time period from 0:00 to 5:00 in the morning, or the operation is concentrated in non-business time (such as holidays, weekends at night) and the frequency is abnormally high. The initiation time indicator is used to determine whether the application initiation time data matches the abnormal time period. If it matches, the value of the initiation time indicator is 1, otherwise it is 0. For example, when the user initiates the amendment application at 2:15 in the morning, the value of the initiation time indicator is 1.

[0109] In step S305 of some embodiments, the application initiation device data refers to the device information of the amendment application terminal, including device type, device number, operating system, browser fingerprint, etc., for identifying whether there is abnormal device access behavior. The preset operation risk indicator includes a risk rule for identifying “suspicious device”, specifically: the operation device is associated with multiple different accounts in a short period of time, or it is a first-time login unknown device and is inconsistent with the device registered in the insurance stage. The initiation device indicator is used to determine whether the application initiation device data meets the above-mentioned suspicious device determination condition. If it meets, the value of the initiation device indicator is 1, otherwise it is 0. For example, when a mobile device first appears in the amendment application of the account, and the device has frequently associated with multiple accounts in recent period of time, the value of the initiation device indicator is 1.

[0110] In step S306 of some embodiments, vector construction refers to arranging the promotion time indicator, field switching indicator, initiation address indicator, initiation time indicator and initiation device indicator in a fixed order. For example, when the values of the promotion time indicator, field switching indicator, initiation address indicator, initiation time indicator and initiation device indicator are 1, 0, 1, 0, 1 in turn, the value of the application operation risk indicator vector is [1, 0, 1, 0, 1].

[0111] In step S205 of some embodiments, vector construction refers to splicing each indicator in the premium adjustment risk indicator, liability adjustment risk indicator, application batch risk indicator and application operation risk indicator in a preset order. For example, when the premium adjustment risk indicator is 1, the liability adjustment risk indicator is 0, the application batch risk indicator is 1, the promotion time indicator is 1, the field switching indicator is 0, the initiation address indicator is 1, the initiation time indicator is 0, and the initiation device indicator is 1, the final risk operation evaluation vector is

[0112] [1, 0, 1, 1, 0, 1, 0, 1].

[0113] Please refer to Figure 4In some embodiments, step S103 can include, but is not limited to, steps S401 to S404:

[0114] In step S401, reference user information is obtained, and user clustering is performed according to the information of the applicant and the reference user information to obtain a target clustering area; wherein the information of the applicant is located in the target clustering area.

[0115] In step S402, user information of the target clustering area is obtained to obtain target user information, and historical amendment application information of the target user information is obtained.

[0116] In step S403, the historical amendment application information and a preset application abnormal threshold index are compared to obtain abnormal amendment application information.

[0117] In step S404, the abnormal amendment application information is vectorized to obtain a risk portrait evaluation vector.

[0118] The steps S401 to S404 shown in the embodiments of the present application construct a user clustering model based on reference user information, identify target users in the same clustering area as the information of the applicant, and then obtain historical amendment application information of the target users and establish a group behavior benchmark. On this basis, by comparing the amendment application information with the historical amendment behavior of the group and combining the preset application abnormal threshold index, abnormal amendment application information deviating from the normal behavior mode is screened out. The abnormal amendment application information is vectorized to construct a risk portrait evaluation vector representing the degree of deviation of the applicant's behavior. Thus, the embodiments of the present application identify potential risks in user behavior based on the behavior similarity between user groups.

[0119] In step S401 of some embodiments, the reference user information is a historical user data set. User clustering refers to a process of dividing applicants with similar behavior characteristics into the same category based on the information of the applicant and the reference user information through feature vectorization and similarity calculation. The implementation of user clustering is as follows: first, the information of the applicant and the reference user information are uniformly converted into standardized feature vectors; then, K-Means algorithm is used for aggregation analysis based on Euclidean distance; and finally, a target clustering area containing the current applicant is generated. For example, if the clustering algorithm calculation result shows that the current applicant is in the same cluster as a group of users with amendment records in the past three years, an average premium fluctuation range exceeding 20%, and a family protection product configuration, then the cluster is the target clustering area.

[0120] In step S402 of some embodiments, the user information of the target clustering area refers to the information of all users in the same cluster as the current applicant in the clustering result. Based on the user information of the target clustering area, the historical amendment application information of these users is extracted.

[0121] In step S403 of some embodiments, the preset application anomaly threshold indicator is a set of field sensitivity rules constructed based on historical revision application information within the target cluster area, and is used to identify whether there are abnormal operations on sensitive fields in the current policy revision application information. The threshold indicator is generated by first counting the modification frequency of each field in the historical revision application information within the target cluster area, identifying fields with extremely low modification frequencies in the overall revision behavior, and marking these fields as sensitive fields; then, for each sensitive field, a judgment indicator related to the field modification behavior is pre-set to form a corresponding anomaly threshold. For example, if the proportion of revisions in the "payment frequency" field in the historical data is extremely low, then this field is classified as a sensitive field; further setting the preset application anomaly threshold indicator for this field to "changing from annual payment to monthly payment and the insured is over 60 years old" is considered an anomaly. The current policy revision application information is compared with these preset indicators, and any field that meets the abnormal conditions constitutes abnormal revision application information.

[0122] In step S404 of some embodiments, the risk profile assessment vector is a structured vector representation of the result of determining anomalies in all fields involved in the policy revision application based on a preset application anomaly threshold. If a field is not considered a sensitive field, or is marked as a sensitive field but does not meet the preset application anomaly threshold, it is assigned a value of 1. If a field is marked as a sensitive field but meets the preset application anomaly threshold, it is assigned a value of 0. The risk profile assessment vector is then generated by sorting the fields.

[0123] For example, the set field order includes five fields: payment frequency, insured amount, coverage, policy duration, and beneficiary type. The payment frequency and coverage fields are sensitive fields. In the current policy modification application, the payment frequency field meets its abnormality threshold, while the coverage field does not meet its corresponding threshold. The remaining fields are not marked as sensitive fields. Therefore, the risk profile assessment vector is [0, 1, 1, 1, 1].

[0124] See also Figure 5 In some embodiments, the policy modification application information includes at least one policy field and a policy field modification record of the policy field. Step S104 includes but is not limited to steps S501 to S503:

[0125] Step S501: Obtain sample modification application information and sample claim records; wherein the sample modification application information includes at least one sample field modification record, and the sample field modification record corresponds to the policy field modification record;

[0126] Step S502: Perform self-attention calculations on the sample field modification records and the sample claim records using a preset attention model to determine the sample claim association value of each sample field modification record, and use the sample claim association value of the sample field modification record as the field claim association value of the policy field modification record corresponding to the sample field modification record;

[0127] Step S503: performing aggregation calculation based on the field claim association value, policy fields, and historical claim records to obtain a risk claim assessment vector.

[0128] In steps S501 to S503 shown in the embodiment of the present application, sample correction application information and sample claim records are obtained; self-attention calculations are performed on the sample field modification records and sample claim records through a preset attention model to determine the sample claim association value of each sample field modification record, and the sample claim association value of the sample field modification record is used as the field claim association value of the policy field modification record corresponding to the sample field modification record. Aggregate calculations are performed based on the field claim association value, policy field, and historical claim records to obtain a risk claim assessment vector. In this way, the embodiment of the present application can achieve risk characterization at the field level, focusing not only on the correction field itself, but also on the claim probability trend hidden behind it, effectively improving the fine-grained ability of correction identification and the pre-risk warning effect, and enhancing the depth of risk identification and processing accuracy.

[0129] In step S501 of some embodiments, the sample field modification record refers to the modification record for each field in the sample amendment application information. For example, in a historical amendment application, the policyholder changed the "Critical Illness Coverage" field from "Not Included" to "Included" and also modified the amount in the "Sum Insured" field. The correspondence between the sample field modification record and the policy field modification record refers to the fields that appeared in the sample amendment application information corresponding to the fields that appeared in the policy amendment application information.

[0130] In step S502 of some embodiments, the role of the attention model is to determine the weight of the field modification record in the claim, so that the field with a strong correlation with the claim behavior obtains a higher correlation value. For example, the attention model can adopt a bidirectional Transformer structure, whose input is a set of sample field modification records and the corresponding claim label sequence. The self-attention mechanism calculates the correlation value of each field modification record to the historical claim results, and outputs a correlation score for measuring the impact of the field on the claim risk. If the model determines that the modification of "coverage liability-major illness" has a significant co-occurrence relationship with multiple high-value claims, the sample claim correlation value of the field modification record will be higher than other fields, and will be assigned a higher correlation value.

[0131] See also Figure 6In some embodiments, step S503 includes but is not limited to steps S601 to S604:

[0132] Step S601: extract fields from the claim record text according to a preset field extraction model to obtain matching fields;

[0133] Step S602: assigning a first value to a field in the policy field that matches the matching field, and assigning a second value to a field in the policy field that does not match the matching field;

[0134] Step S603: Perform weighted calculation based on the field claim association value and the value in the policy field to determine the risk claim assessment value of each policy field;

[0135] Step S604: perform vectorization processing on each risk claim assessment value to obtain a risk claim assessment vector.

[0136] In steps S601 to S604 shown in the embodiment of the present application, field extraction is performed on the claim record text according to a preset field extraction model to obtain matching fields. The fields in the policy field that match the matching fields are assigned a first value, and the fields in the policy field that do not match the matching fields are assigned a second value. A weighted calculation is performed based on the field claim association value and the value in the policy field to determine the risk claim assessment value of each policy field. Vectorization is performed based on each risk claim assessment value to obtain a risk claim assessment vector. In this way, the embodiment of the present application improves the mapping capability of the policy field to the claim semantic scenario and realizes the association between structured fields and claim history.

[0137] In step S601 of some embodiments, the field extraction model is a model structure used to identify key content that has a semantic correspondence with the revised fields from the historical claim record text. The function of the field extraction model is to convert unstructured claim text information into standard fields that can be compared with the policy field structure through semantic understanding. The field extraction model is implemented by using a deep network structure with text semantic modeling capabilities to embed the claim record text, and then perform field mapping on the keywords in the claim segment through field tags or nested query mechanisms.

[0138] For example, the field extraction model can adopt a dual-channel BERT-CRF joint structure, with one channel inputting a set of revised field names and the other channel inputting the claim record text. Through a multi-round interactive attention mechanism, keywords or phrases in the text that are semantically related to the revised field names are captured, and matching field results are output. For example, the claim content "The customer applied for hospitalization allowance due to an accidental fracture" is identified as matching fields "Guarantee Liability - Accidental Medical Treatment" and "Guarantee Liability - Hospitalization Allowance".

[0139] In step S602 of some embodiments, the fields in the insurance policy field that are consistent with the matching fields at the semantic level are assigned a first value, such as 1.0, indicating a complete match; the fields in the insurance policy field that are not covered by the matching fields but may still have an implicit semantic association are assigned a second value, such as 0.3, indicating a weak match. By introducing a non-zero weak match value, all fields retain a basic participation weight in subsequent risk assessment, avoiding the problem of fields losing risk representation ability due to structural assignment of zero.

[0140] In step S603 of some embodiments, the risk claim assessment value is a numerical result formed after weighting each insurance policy field, used to measure the potential risk response level of the field in the historical claim scenario.

[0141] For example, in a certain embodiment, the insurance policy field "guarantee responsibility-accidental medical" is assigned a structural value of 1.0, and the corresponding field claim association value is 0.72, so its risk claim assessment value is 0.72x1.0=0.72; the insurance policy field "guarantee responsibility-critical illness guarantee" is not identified as a matching field and is assigned a structural value of 0.3, and the corresponding field claim association value is 0.85, so its risk claim assessment value is 0.85x0.3=0.255. This weighting method effectively combines the semantic matching degree and historical risk weight, improving the granularity and discrimination ability of field risk measurement.

[0142] In step S604 of some embodiments, the risk claim assessment vector is a structured vector formed by arranging and concatenating the risk claim assessment values of all insurance policy fields in a predetermined order.

[0143] Please refer to Figure 7 In some embodiments, step S105 can include but is not limited to steps S701-S703:

[0144] Step S701, vector splicing the risk operation assessment vector, the risk portrait assessment vector and the risk claim assessment vector to obtain a spliced assessment vector;

[0145] Step S702, risk rating the spliced assessment vector through a predetermined random forest model to obtain a risk level label;

[0146] Step S703, screening the application processing scheme according to the risk level label to obtain a target processing scheme.

[0147] The steps S701 to S703 shown in the embodiment of the present application perform vector splicing on the risk operation evaluation vector, the risk portrait evaluation vector and the risk claim evaluation vector to obtain a spliced evaluation vector. The spliced evaluation vector is subjected to risk rating by a preset random forest model to obtain a risk rating label. The application processing scheme is screened according to the risk rating label to obtain a target processing scheme. In this way, the embodiment of the present application realizes automatic screening and standardized control of the processing scheme.

[0148] In step S702 of some embodiments, the random forest model is a classification model used to determine the risk rating of the spliced evaluation vector. The risk rating label includes three levels of low risk, medium risk and high risk, which correspond to different risk levels.

[0149] In step S703 of some embodiments, the application processing scheme is a pre-defined processing scheme. The application processing scheme includes: if the risk rating label is low risk, no processing is selected and direct passing is performed; if the risk rating label is medium risk, a data supplement mechanism is triggered, for example, when a certain type of disease protection responsibility is added in the amendment application, health examination materials or medical proof documents related to the protection content need to be supplemented; and if the risk rating label is high risk, the application amendment information is submitted to an artificial audit queue for intervention determination by professional auditors.

[0150] Please refer to Figure 8 The embodiment of the present application also provides a processing device for a policy amendment application, which can implement the above-mentioned processing method for the policy amendment application. The device comprises:

[0151] The acquisition data module 801 is configured to acquire the policy amendment application information of a target user, acquire the policyholder information and historical claim records of the target user;

[0152] The operation evaluation module 802 is configured to perform a policy modification operation risk evaluation according to the policy amendment application information to obtain a risk operation evaluation vector;

[0153] The deviation evaluation module 803 is configured to perform a user portrait deviation evaluation according to the policyholder information and the policy amendment application information to obtain a risk portrait evaluation vector;

[0154] The anti-fraud evaluation module 804 is configured to perform a claim anti-fraud evaluation according to the policy amendment application information and the historical claim records to obtain a risk claim evaluation vector;

[0155] The scheme screening module 805 is configured to screen a preset application processing scheme based on the risk operation evaluation vector, the risk portrait evaluation vector and the risk claim evaluation vector to obtain a target processing scheme;

[0156] The application processing module 806 is used to process the policy amendment application information according to the target processing solution.

[0157] The specific implementation of the device for processing the policy amendment application is substantially the same as the specific embodiment of the method for processing the policy amendment application described above, and will not be described in detail herein.

[0158] The present application also provides an electronic device comprising a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method for processing an insurance policy revision application. The electronic device can be any smart terminal, such as a tablet computer or an in-vehicle computer.

[0159] See also Figure 9 , Figure 9 The hardware structure of an electronic device according to another embodiment is shown. The electronic device includes:

[0160] The processor 901 can be implemented as a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0161] The memory 902 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 902 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program codes are stored in the memory 902, and the processor 901 calls and executes the method for processing the policy amendment application in the embodiments of this application.

[0162] Input / output interface 903, used to implement information input and output;

[0163] Communication interface 904, used to implement communication interaction between this device and other devices, which can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WiFi, Bluetooth, etc.);

[0164] Bus 905 , which transmits information between various components of the device (e.g., processor 901 , memory 902 , input / output interface 903 , and communication interface 904 );

[0165] The processor 901, the memory 902, the input / output interface 903, and the communication interface 904 are communicatively connected with each other through the bus 905.

[0166] The application also provides a computer readable storage medium, which stores a computer program. The computer program is executed by a processor to implement the policy amendment application processing method.

[0167] The memory is a non-transitory computer readable storage medium, which can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include a memory remotely arranged relative to the processor, and these remote memories can be connected to the processor through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0168] The policy amendment application processing method, the policy amendment application processing device, the electronic device, and the storage medium provided by the application obtain policy amendment application information, applicant information, and historical claim records; perform risk operation evaluation according to the policy amendment application information to obtain a risk operation evaluation vector; perform user portrait deviation evaluation according to the applicant information and the policy amendment application information to obtain a risk portrait evaluation vector; perform claim anti-fraud evaluation according to the policy amendment application information and the historical claim records to obtain a risk claim evaluation vector; filter a preset application processing scheme based on the risk operation evaluation vector, the risk portrait evaluation vector, and the risk claim evaluation vector to obtain a target processing scheme; and perform application processing on the policy amendment application information according to the target processing scheme. In this way, the application introduces a multi-dimensional evaluation mechanism to comprehensively process the policy amendment application information, the applicant information, and the historical claim records, jointly evaluates the risk of policy modification operation, the user portrait deviation, and the claim anti-fraud risk, and then filters a processing scheme matching the risk level based on multiple risk evaluation results, thereby improving the pertinence and rationality of the processing strategy, and finally processes the amendment application according to the target processing scheme. The application can accurately judge the rationality of the amendment application based on the cooperative evaluation in different risk dimensions, significantly reduces the misjudgment rate caused by manual subjective judgment, and improves the standardization and accuracy of the policy amendment processing.

[0169] The embodiments described in the embodiments of this application are intended to more clearly illustrate the technical solutions of the embodiments of this application and do not constitute a limitation on the technical solutions provided by the embodiments of this application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0170] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0171] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, i.e., they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment.

[0172] Those skilled in the art will appreciate that all or some of the steps in the methods, systems, and functional modules / units in the devices disclosed above may be implemented as software, firmware, hardware, or appropriate combinations thereof.

[0173] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0174] It should be understood that, in the application, "at least one" refers to one or more, and "multiple" refers to two or more. "And / or" is used to describe the association relationship of the associated objects, which means that there can be three relationships, for example, "A and / or B" can represent three cases of only A, only B, and A and B existing at the same time, wherein A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects before and after it. "At least one of the following" or similar expressions means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b or c can represent a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, and c can be single or multiple.

[0175] In several embodiments provided in the application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the above units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed mutual ones can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0176] The units described above as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0177] In addition, the functional units in each embodiment of the application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0178] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, essentially or in other words, the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes multiple instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program storage media.

[0179] The preferred embodiments of the embodiments of the present application are described above with reference to the accompanying drawings, and are not limited to the scope of the embodiments of the present application. Any modifications, equivalent replacements and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of the present application shall be within the scope of the embodiments of the present application.

Claims

1. A method for processing an insurance policy amendment application, characterized in that: The method comprises: Obtain the target user's insurance policy modification application information, and obtain the target user's policyholder information and historical claims records; Performing a risk assessment of the policy modification operation based on the policy modification application information to obtain a risk operation assessment vector; Performing a user profile deviation assessment based on the policyholder information and the policy revision application information to obtain a risk profile assessment vector; Performing a claim anti-fraud assessment based on the policy revision application information and the historical claim records to obtain a risk claim assessment vector; Screening a preset application processing solution based on the risk operation assessment vector, the risk profile assessment vector, and the risk claim assessment vector to obtain a target processing solution; The policy revision application information is processed according to the target processing solution.

2. The method according to claim 1, characterized in that The policy modification application information includes insurance amount adjustment data, coverage change data, modification application frequency data, and application operation attribute data; performing a policy modification operation risk assessment based on the policy modification application information to obtain a risk operation assessment vector includes: Matching the insured amount adjustment data with a preset insured amount adjustment indicator to obtain an insured amount adjustment risk indicator; Matching the guarantee responsibility change data with the preset responsibility adjustment index to obtain the responsibility adjustment risk index; Matching the application revision frequency data with a preset adjustment threshold indicator to obtain an application batch risk indicator; Matching the application operation attribute data with a preset operation risk indicator to obtain an application operation risk indicator; A vector is constructed based on the insurance amount adjustment risk indicator, the liability adjustment risk indicator, the application batch risk indicator and the application operation risk indicator to obtain the risk operation assessment vector.

3. The method according to claim 2, characterized in that The application operation attribute data includes step advancement time data, sensitive field switching frequency, application initiation address data, application initiation time data, and application initiation device data; the application operation attribute data is matched with a preset operation risk indicator to obtain an application operation risk indicator, including: Matching the step advancement time data with the preset operational risk indicator to obtain an advancement time indicator; Matching the sensitive field switching frequency with the preset operation risk indicator to obtain a field switching indicator; Matching the application initiation address data with the preset operational risk indicator to obtain an initiation address indicator; Matching the application initiation time data with the preset operational risk indicator to obtain an initiation time indicator; Matching the application initiating device data with the preset operational risk indicator to obtain an initiating device indicator; A vector is constructed according to the advancement time indicator, the field switching indicator, the initiation address indicator, the initiation time indicator, and the initiation device indicator to obtain the application operation risk indicator.

4. The method according to claim 1, wherein The user profile deviation assessment is performed based on the policyholder information and the policy modification application information to obtain a risk profile assessment vector, including: Acquire reference user information, and perform user clustering based on the policyholder information and the reference user information to obtain a target clustering area; wherein the policyholder information is located in the target clustering area; Obtaining user information of a target cluster area, obtaining target user information, and obtaining historical revision application information of the target user information; Comparing the policy revision application information with the historical revision application information and the preset application abnormality threshold index to obtain abnormal revision application information; The abnormal correction application information is vectorized to obtain the risk profile assessment vector.

5. The method according to claim 1, wherein The policy modification application information includes at least one policy field and a policy field modification record of the policy field; performing a claim anti-fraud assessment based on the policy modification application information and the historical claim records to obtain a risk claim assessment vector includes: Obtaining sample revision application information and sample claim records; wherein the sample revision application information includes at least one sample field modification record, and the sample field modification record corresponds to the policy field modification record; Performing self-attention calculation on the sample field modification record and the sample claim record through a preset attention model, determining a sample claim association value of each sample field modification record, and using the sample claim association value of the sample field modification record as the field claim association value of the policy field modification record corresponding to the sample field modification record; Aggregate calculation is performed based on the field claim association value, the policy field and the historical claim records to obtain the risk claim assessment vector.

6. The method according to claim 5, characterized in that The historical claim records include claim record texts. The aggregate calculation is performed based on the field claim association value, the policy field, and the historical claim records to obtain the risk claim assessment vector, including: Performing field extraction on the claim record text according to a preset field extraction model to obtain matching fields; Assigning a first value to a field in the policy field that matches the matching field, and a second value to a field in the policy field that does not match the matching field; Perform weighted calculation based on the claim association value of the field and the value in the policy field to determine the risk claim assessment value of each policy field; Vectorization processing is performed on each of the risk claim assessment values ​​to obtain the risk claim assessment vector.

7. The method according to any one of claims 1 to 6, characterized in that The step of screening a preset application processing solution based on the risk operation assessment vector, the risk profile assessment vector, and the risk claim assessment vector to obtain a target processing solution includes: Performing vector splicing on the risk operation assessment vector, the risk profile assessment vector, and the risk claim assessment vector to obtain a spliced ​​assessment vector; Performing risk rating on the spliced ​​assessment vector using a preset random forest model to obtain a risk level label; The application processing solutions are screened according to the risk level labels to obtain the target processing solution.

8. A device for processing insurance policy amendment applications, characterized in that: The device comprises: The data acquisition module is used to obtain the target user's insurance policy modification application information, the target user's policyholder information and historical claims records; An operation assessment module, configured to perform a risk assessment of the policy modification operation based on the policy modification application information to obtain a risk operation assessment vector; a deviation evaluation module, configured to evaluate the deviation of the user profile based on the policyholder information and the policy amendment application information, and obtain a risk profile evaluation vector; an anti-fraud assessment module, configured to perform a claim anti-fraud assessment based on the policy revision application information and the historical claim records to obtain a risk claim assessment vector; A solution screening module, configured to screen preset application processing solutions based on the risk operation assessment vector, the risk profile assessment vector, and the risk claim assessment vector to obtain a target processing solution; The application processing module is used to process the insurance policy amendment application information according to the target processing solution.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the method for processing an insurance policy amendment application according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for processing an insurance policy amendment application according to any one of claims 1 to 7 is implemented.

Images