Insurance policy identification method and device based on large model, equipment and medium
Through the policy identification method based on large-model, property insurance risk assessment is automated, which solves the problem of manual assessment in the prior art that manual assessment is time-consuming and labor-intensive and inaccurate results, and achieves more efficient and accurate risk assessment results.
Patent Information
- Application Number
- CN202510301133.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-13
- Publication Date
- 2025-06-27
AI Technical Summary
In the prior art, property insurance risk assessment relies on manual exploration and evaluation, which is time-consuming and labor-intensive. Especially in the case of complex and high-risk operations in the medical field, it is easy to ignore important risks, resulting in the assessment results that are not comprehensive and accurate enough.
The policy identification method based on the big model is adopted. By obtaining the data of the property insurance target submitted by the user and user information, matching the auxiliary evaluation data, calculating the cost risk level, determining the risk feature extraction model, extracting risk feature data, screening key feature data, and using the user risk calculation model for evaluation, obtaining the user risk assessment results.
It realizes automated risk assessment, reduces human intervention, improves the efficiency and accuracy of risk assessment, and can more accurately identify the risk level of the insured object, extracts key information related to risks, and comprehensively considers risk factors related to users.
Smart Images
Figure CN120219090A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence technology, and particularly to a method, device, equipment and medium for policy identification based on a large model Background Art Property insurance inquiry and quotation refers to the process in which a customer consults an insurance company and obtains an insurance quotation, including two stages: property inquiry and quotation. In the inquiry stage, the customer mainly asks the insurance company about relevant information such as the price, coverage, and claims process of insurance products. In the quotation stage, the insurance company needs to formulate and provide a detailed insurance cost plan and corresponding safeguard measures according to the customer's needs and risk situation. In this process, the insurance company needs to conduct a risk assessment on the insured information such as the insurance object information and customer credit information provided by the customer, so as to formulate a reasonable insurance cost plan and corresponding safeguard measures according to the risk assessment results
[0002] In the medical field, in order to cope with possible accidental losses, medical institutions usually need to insure property such as their key medical equipment, building facilities, and inventory drugs. However, in the prior art, the risk assessment work mainly relies on the staff of the insurance company to conduct on-site inspections and risk assessments based on the received insured information. This process is both time-consuming and laborious. Especially when the operating environment of medical institutions is complex, the types of equipment are numerous, and it involves high-risk fields such as patient safety and medical quality, the difficulty and time consumption of manual assessment increase significantly, and it may be limited by the experience and knowledge of the staff, resulting in the omission of some important risks, making the results of risk assessment not comprehensive and accurate enough
[0003] Therefore, how to improve the efficiency of risk assessment while ensuring the accuracy of the assessment results has become an urgent problem to be solved Summary of the Invention
[0004] The present invention provides a method, device, computer equipment and medium for policy identification based on a large model to solve the problem of how to improve the efficiency of risk assessment while ensuring the accuracy of the assessment results
[0005] In a first aspect, a method for policy identification based on a large model is provided, including: Obtain the property insurance object data, coverage data submitted by the user and the user information of the corresponding user, and match the corresponding auxiliary assessment data from a preset database according to the property insurance object data Calculate the cost risk level according to the property insurance object data and the auxiliary assessment data, and determine the corresponding risk feature extraction model according to the cost risk level Input the user information into the risk feature extraction model and output risk feature data Match the associated information from the preset knowledge graph according to the user information, and screen the risk feature data according to the association relationship between the associated information and the user information to obtain the key feature data; Use the user risk calculation model to evaluate the key feature data to obtain the user risk assessment result.
[0006] In a second aspect, a policy recognition device based on a large model is provided, including: An acquisition module, configured to acquire the property insurance subject data, the coverage data submitted by the user, and the user information of the corresponding user, and match the corresponding auxiliary evaluation data from the preset database according to the property insurance subject data; A first evaluation module, configured to calculate the cost risk level according to the property insurance subject data and the auxiliary evaluation data, and determine the corresponding risk feature extraction model according to the cost risk level; An extraction module, configured to input the user information into the risk feature extraction model and output risk feature data; A screening module, configured to match the associated information from the preset knowledge graph according to the user information, and screen the risk feature data according to the association relationship between the associated information and the user information to obtain the key feature data; A second evaluation module, configured to use the user risk calculation model to evaluate the key feature data to obtain the user risk assessment result.
[0007] In a third aspect, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the policy recognition method based on the large model in the first aspect are implemented.
[0008] In a fourth aspect, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the policy recognition method based on the large model in the first aspect are implemented.
[0009] In the solution implemented by the above-mentioned large model-based policy recognition method, device, equipment and medium, by obtaining the property insurance subject data, coverage data submitted by the user and the user information corresponding to the user, according to the property insurance subject data, the corresponding auxiliary evaluation data is matched from the preset database, according to the property insurance subject data and the auxiliary evaluation data, the cost risk level is calculated, according to the cost risk level, the corresponding risk feature extraction model is determined, the user information is input into the risk feature extraction model, the risk feature data is output, according to the user information, the associated information is matched from the preset knowledge graph, and according to the association relationship between the associated information and the user information, the risk feature data is screened to obtain the key feature data, and the user risk calculation model is used to evaluate the key feature data to obtain the user risk assessment result.
[0010] Among them, through the risk feature extraction model and the user risk calculation model, the risk assessment of the property subject data and user information submitted by the user is automatically realized, reducing manual intervention and improving the efficiency of risk assessment. By using the auxiliary evaluation data as a reference to calculate the cost risk level of the insurance subject data, the risk level of the insurance subject can be more accurately identified; based on the cost risk level, using the corresponding risk feature extraction model for risk feature extraction can more effectively extract the key information related to the risk and improve the pertinence of risk assessment; and based on the preset knowledge graph, screening the extracted risk feature data can more comprehensively consider the risk factors related to the corresponding user, thereby improving the efficiency of risk assessment while ensuring the accuracy of the assessment result. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the description of the embodiments of the present invention. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.
[0012] Figure 1 It is a schematic application environment diagram of a large model-based policy recognition method provided in Embodiment 1 of the present invention; Figure 2 It is a schematic flowchart of a large model-based policy recognition method provided in Embodiment 2 of the present invention; Figure 3 It is a schematic flowchart of a large model-based policy recognition method provided in Embodiment 3 of the present invention; Figure 4 It is a schematic flowchart of a large model-based policy recognition method provided in Embodiment 4 of the present invention; Figure 5Schematic flowchart of a policy recognition method based on a large model provided in Embodiment 5 of the present invention; Figure 6 Schematic structural diagram of a policy recognition device based on a large model provided in Embodiment 6 of the present invention; Figure 7 Schematic structural diagram of a computer device provided in Embodiment 7 of the present invention. Detailed implementation manners
[0013] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.
[0014] The policy recognition method based on a large model provided in Embodiment 1 of the present invention can be applied in an application environment such as Figure 1 wherein, the server communicates with the client, the server provides a policy recognition service based on a large model, and the client triggers a policy recognition task based on a large model to the server. Among them, the client includes, but is not limited to, a palm computer, a desktop computer, a notebook computer, an ultra-mobile personal computer (UMPC), a netbook, a cloud computer device, a personal digital assistant (PDA), and other devices. The computer device corresponding to the server can be implemented by an independent server or a server cluster composed of multiple servers.
[0015] Such as Figure 2 shown, it is a schematic flowchart of a policy recognition method provided in Embodiment 2 of the present invention, including the following steps: Step S201: Obtain the property insurance subject data, the coverage data submitted by the user, and the user information of the corresponding user, and match the corresponding auxiliary evaluation data from the preset database according to the property insurance subject data.
[0016] In this embodiment, the property target data may refer to the property insured by the user and its related data. For example, in a medical application scenario, the property insured by the property target data may be medical equipment, and the corresponding property target data may include relevant information such as the equipment type and specifications of the medical equipment, the purchase value and depreciation situation, and the maintenance and repair records. The coverage data may refer to the specific coverage content and conditions that the user hopes the insurance company will provide for the property insured by the user. For example, the coverage data may include the coverage time of the property that the user hopes to insure. The user information may refer to the basic information and background information related to the user, which is used for the insurance company to understand the user's identity, credit status, and risk tolerance. For example, the user information may include basic identity information, credit records, and insurance history. The auxiliary evaluation data may refer to the data used to evaluate the value of the property corresponding to the property target data, which is matched from a preset database. For example, in a medical application scenario, if the property insured by the property target data is medical equipment, the auxiliary evaluation data may include the standard purchase price data, standard maintenance and repair cost data, and standard operation and consumption cost data of the corresponding medical equipment.
[0017] Specifically, the property target data submitted by the user can be parsed to determine the property insured by the property target data, and the corresponding auxiliary evaluation data can be matched from the preset database according to the property insured by the property target data.
[0018] Step S202: Calculate the cost risk level according to the property insurance target data and the auxiliary evaluation data, and determine the corresponding risk feature extraction model according to the cost risk level.
[0019] Step S203: Input the user information into the risk feature extraction model and output the risk feature data.
[0020] In this embodiment, the cost risk level may refer to the value risk of the property insured corresponding to the property target data. The risk feature extraction model may refer to a model that has been trained for risk feature data extraction. For example, the risk feature extraction model may be a recurrent neural network model, a long short-term memory network model, etc. The risk feature data may refer to the data reflecting the user's risk status. For example, the risk feature data may include the user's credit score, debt situation, and income stability situation.
[0021] Specifically, according to the auxiliary evaluation data, the value of the property corresponding to the property insurance target data can be evaluated to obtain the evaluation cost. According to the evaluation cost and in combination with the preset cost risk level table, the cost risk level corresponding to the property insurance target data can be determined. According to different cost risk levels, the corresponding risk feature extraction model can be determined, and the user information can be input into the risk feature extraction model. The risk feature extraction model extracts risk features from the user information and outputs risk feature data.
[0022] Among them, the evaluation cost can refer to an economic quantification index used to measure the value of the property corresponding to the property insurance target data. This index is based on currency units. The preset cost risk level table can refer to a table that pre - sets the mapping relationship between the evaluation cost and the cost risk level.
[0023] Optionally, according to the auxiliary evaluation data, a comprehensive evaluation can also be performed on the value, type, usage environment, maintenance status, potential losses, etc. of the property corresponding to the property target data. According to the comprehensive evaluation result, the value risk of the property corresponding to the property target data can be determined, and thus the corresponding cost risk level can be determined.
[0024] Step S204: According to the user information, associated information is matched from the preset knowledge graph. According to the association relationship between the associated information and the user information, the risk feature data is screened to obtain key feature data.
[0025] Step S205: Use the user risk calculation model to evaluate the key feature data to obtain the user risk assessment result.
[0026] In this embodiment, the preset knowledge graph can refer to a graph - structured data that is pre - constructed for storing and representing various entities related to the user information and their mutual relationships. For example, the preset knowledge graph can include entity information (such as people, locations, organizations, and products, etc.), attribute information (such as the identity information of people, the scale of organizations, business scope, etc.), and relationship information between entities (such as the social relationship between users and others, the employment relationship between users and organizations, and the purchase relationship between users and products, etc.); the associated information can refer to the entity information related to the user information in the preset knowledge graph, and the association relationship can refer to the relationship information between the entity represented by the associated information and the entity represented by the user information; the key feature data can refer to the risk feature data with important influence, and the user risk calculation model can refer to a model that has been trained for user risk assessment. For example, the user risk calculation model can be a convolutional neural network model, etc. The user risk assessment result can refer to the result of risk assessment of the user information.
[0027] Specifically, according to the user information, match in a preset knowledge graph to obtain associated information. Based on the association relationship between the associated information and the user information, evaluate the risk characteristic data. According to the evaluation result, screen out the key characteristic data, and input the key characteristic data into the user risk calculation model. Through the user risk calculation model, conduct a risk assessment on the user information and output the user risk assessment result.
[0028] Optionally, after obtaining the user risk assessment result, also match the corresponding risk control strategy according to the user risk assessment result, and conduct risk control on the user according to the risk control strategy.
[0029] For example, according to the user risk assessment result, for high-risk users, more stringent underwriting reviews can be conducted to ensure that they meet the insurance company's underwriting conditions; for medium-risk users, appropriate risk warnings and suggestions for preventive measures can be provided to reduce potential risks; for low-risk users, more preferential insurance terms and rates can be offered to attract more high-quality customers.
[0030] In this embodiment, through the risk characteristic extraction model and the user risk calculation model, the risk assessment of the property target data and user information submitted by the user is automatically realized, reducing human intervention and improving the efficiency of risk assessment. By using the auxiliary assessment data as a reference to calculate the cost risk level of the insurance target data, the risk level of the insurance target can be more accurately identified; based on the cost risk level, using the corresponding risk characteristic extraction model to extract risk characteristics can more effectively extract key information related to risks and improve the pertinence of risk assessment; and based on the preset knowledge graph, screening the extracted risk characteristic data can more comprehensively consider the risk factors related to the corresponding user, thereby improving the efficiency of risk assessment while ensuring the accuracy of the assessment result.
[0031] As Figure 3 shown, it is a schematic flowchart of a method for identifying insurance policies based on a large model provided by Embodiment 3 of the present invention. The step of matching the associated information from the preset knowledge graph according to the user information in the above step S204 may include the following steps: Step S301: According to the user information, determine the entity node in the preset knowledge graph that represents the same entity as the user information corresponding user.
[0032] Step S302: In the preset knowledge graph, determine the k-hop neighborhood graph corresponding to the entity node.
[0033] Step S303: According to the user information, determine the matching node that matches the user information from the k-hop neighborhood graph.
[0034] Step S304: Determine the nodes other than the entity node and the matching node in the k-hop neighborhood graph as associated nodes, and form association information from all the associated nodes.
[0035] In this embodiment, the entity node may refer to a node that represents the same entity as the user corresponding to the user information. The k-hop neighborhood graph may refer to a graph centered on the entity node, including all neighbor nodes within its k-hop range and the connection relationships between these nodes. k is an integer greater than zero. The matching node may refer to a node that represents the same entity as the user information. The matching nodes include the entity node. The associated node may refer to a node in the k-hop neighborhood graph other than the entity node and the matching node.
[0036] Specifically, according to the user information, determine the matching nodes that match the user information and the entity nodes that represent the same entity as the user corresponding to the user information from the preset knowledge graph, and determine the k-hop neighborhood graph formed by all neighbor nodes within the k-hop range of the entity node and the connection relationships between these nodes. Determine the nodes other than the entity node and the matching node in the k-hop neighborhood graph as associated nodes, and form association information from all the associated nodes.
[0037] In this embodiment, by determining the k-hop neighborhood graph of the entity node corresponding to the user information from the preset knowledge graph, determining the associated nodes not included in the user information in the k-hop neighborhood graph as association information, and using the association information as a supplement to the user information to screen the extracted risk feature data, it is possible to more comprehensively consider the risk factors related to the corresponding user, thereby improving the efficiency of risk assessment while ensuring the accuracy of the assessment results.
[0038] As Figure 4 shown, it is a schematic flowchart of a policy recognition method based on a large model provided in Embodiment 4 of the present invention. In the above step S204, according to the association relationship between the association information and the user information, screening the risk feature data to obtain the key feature data may include the following steps: Step S401: For any associated node in the association information, analyze the association relationship between the feature represented by the associated node and the entity node.
[0039] Step S402: For any risk feature data, evaluate the risk feature data according to the association relationship to obtain a risk score.
[0040] Step S403: According to the risk score, screen all the risk feature data to obtain the key feature data.
[0041] In this embodiment, for any node in the preset knowledge graph, the features represented by the node may refer to the attribute information of the entity represented by the node, the entity information related to the entity represented by the node, and the relationship information between the two. The risk score may refer to a score representing the influence degree of risk characteristic data.
[0042] For example, for an associated node, the entity represented by the associated node is a medical institution. The features represented by the associated node are that the users (i.e., patients) associated with the entity node have received a treatment or examination in this medical institution, and the patients have not completed the payment according to the agreed time or requirements, that is, there is a situation of overdue payment. Then, based on this association relationship, a higher risk score can be given to the data representing the credit of the user in the risk characteristic data to indicate that the potential risk of the user's credit needs to be considered emphatically. Thus, all the risk characteristic data can be sorted or screened according to the risk score, and the risk characteristic data with a higher risk score is obtained as the key characteristic data.
[0043] In this embodiment, through the association relationship between the features represented by the associated node and the entity node, the risk characteristic data is evaluated to obtain a risk score. According to the risk score, the key characteristic data is screened, and the association information is used as a supplement to the user information to screen the extracted risk characteristic data, which can more comprehensively consider the risk factors related to the corresponding user. Thus, while improving the efficiency of risk assessment, the accuracy of the assessment result is also ensured.
[0044] As Figure 5 shown, it is a schematic flowchart of a policy recognition method based on a large model provided in Embodiment 5 of the present invention. After obtaining the user risk assessment result in the above step S205, the following steps may further be included: Step S501: According to the user risk assessment result, adjust the expense risk level to obtain the adjusted expense risk level.
[0045] Step S502: According to the user risk assessment result, the adjusted expense risk level, and the protection scope data, determine the premium rate.
[0046] Step S503: Calculate the premium expense according to the premium rate and the assessment expense.
[0047] In this embodiment, the premium rate may refer to the insurance rate, that is, the fee ratio charged by the insurance company to assume specific risks. The premium expense may refer to the insurance premium, that is, the fee paid by the user to obtain insurance protection.
[0048] Specifically, the cost risk level can be adjusted according to the user risk assessment result. That is, if the user risk assessment result shows that the user is a high-risk user, the cost risk level is correspondingly increased to raise the risk degree of the insured object data. The user risk assessment result, the adjusted cost risk level, and the coverage data are comprehensively considered to determine the premium rate, and the premium rate is multiplied by the assessment cost to obtain the premium cost.
[0049] Optionally, after obtaining the user risk assessment result, a risk assessment report can also be generated according to the user information, the user risk assessment result, the coverage data, the property object data, the adjusted cost risk level, the premium rate, and the premium cost.
[0050] In this embodiment, by adjusting the cost risk level according to the user risk assessment result, determining the premium rate based on the adjusted cost risk level, the user risk assessment result, and the coverage, and obtaining the premium cost based on the premium rate and the assessment cost, the accuracy of risk assessment is improved, and thus the accuracy of the calculated premium cost is improved.
[0051] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.
[0052] As Figure 6 shown, a policy recognition device based on a large model provided in Embodiment VI of the present invention is corresponding to the policy recognition method based on a large model in the above embodiment. As Figure 6 shown, the policy recognition device based on a large model includes an acquisition module 61, a first evaluation module 62, an extraction module 63, a screening module 64, and a second evaluation module 66. The detailed description of each functional module is as follows: The acquisition module 61 is configured to acquire the property insurance object data, the coverage data, and the user information of the corresponding user submitted by the user, and match the corresponding auxiliary evaluation data from a preset database according to the property insurance object data; The first evaluation module 62 is configured to calculate the cost risk level according to the property insurance object data and the auxiliary evaluation data, and determine the corresponding risk feature extraction model according to the cost risk level; The extraction module 63 is configured to input the user information into the risk feature extraction model and output risk feature data; The screening module 64 is configured to match the associated information from a preset knowledge graph according to the user information, and screen the risk feature data according to the association relationship between the associated information and the user information to obtain key feature data; The second evaluation module 66 is used to evaluate the key feature data using the user risk calculation model to obtain the user risk evaluation result.
[0053] Optionally, the above screening module 64 includes: The first matching unit is used to determine, according to the user information, an entity node in the preset knowledge graph that represents the same entity as the user information corresponding user representation; The sub-graph determination unit is used to determine, in the preset knowledge graph, a k-hop neighborhood graph corresponding to the entity node, where k is an integer greater than zero; The second matching unit is used to determine, according to the user information, a matching node that matches the user information from the k-hop neighborhood graph; The formation unit is used to determine, from the k-hop neighborhood graph, nodes other than the entity node and the matching node as associated nodes, and form all the associated nodes into the associated information.
[0054] Optionally, the above screening module 64 includes: The analysis unit is used to analyze, for any associated node in the associated information, the association relationship between the feature represented by the associated node and the entity node; The score determination unit is used to evaluate any risk feature data according to the association relationship to obtain a risk score; The score screening unit is used to screen all the risk feature data according to the risk score to obtain the key feature data.
[0055] Optionally, the above first evaluation module 62 includes: The third evaluation unit is used to evaluate the value of the property corresponding to the property insurance subject data according to the auxiliary evaluation data to obtain an evaluation cost; The level determination unit is used to determine the cost risk level corresponding to the property insurance subject data according to the evaluation cost in combination with a preset cost risk level table.
[0056] Optionally, the risk assessment device further includes: The adjustment module is used to adjust the cost risk level according to the user risk evaluation result to obtain an adjusted cost risk level; The premium rate determination module is used to determine the premium rate according to the user risk evaluation result, the adjusted cost risk level, and the coverage data; The cost calculation module is used to calculate the premium cost according to the premium rate and the evaluation cost.
[0057] Optionally, the risk assessment device further includes: A report generation module, configured to generate a risk assessment report according to the user information, the user risk assessment result, the protection scope data, the property target data, the adjusted cost risk level, the premium rate, and the premium cost.
[0058] Optionally, the risk assessment device further includes: A policy matching module, configured to match a corresponding risk control policy according to the user risk assessment result; A control module, configured to perform risk control on the user according to the risk control policy.
[0059] For the specific limitations of the policy recognition device based on the large model, reference can be made to the limitations of the policy recognition method based on the large model in the above text, which will not be elaborated here. Each module in the above policy recognition device based on the large model can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in the processor in the computer device in hardware form or independent of it, or stored in the memory in the computer device in software form, so that the processor can call and execute the operations corresponding to the above modules.
[0060] As Figure 7 shown, it is a schematic structural diagram of a computer device provided in Embodiment 7 of the present invention. The computer device in this embodiment includes: at least one processor ( Figure 7 only one is shown in the figure), a memory, and a computer program stored in the memory and executable on at least one processor. When the processor executes the computer program, it implements the steps in any of the above embodiments of the policy recognition method based on the large model.
[0061] The computer device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that Figure 7 this is only an example of a computer device and does not constitute a limitation on the computer device. The computer device may include more or fewer components than shown in the figure, or combine some components, or different components. For example, it may further include a network interface, a display screen, and an input device, etc.
[0062] The so-called processor may be a CPU, and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.
[0063] The memory includes a readable storage medium, internal memory, etc. Among them, the internal memory may be the memory of the computer device, and the internal memory provides an environment for the operation of the operating system and computer-readable instructions in the readable storage medium. The readable storage medium may be the hard disk of the computer device, and in some other embodiments, it may also be an external storage device of the computer device. For example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device. Further, the memory may also include both the internal storage unit of the computer device and the external storage device. The memory is used to store the operating system, application programs, boot loaders, data, and other programs, such as the program code of computer programs. The memory may also be used to temporarily store the data that has been output or will be output.
[0064] Those skilled in the art can clearly understand that, for the convenience and conciseness of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be allocated to different functional units and modules as needed, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiments can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of the functional units and modules are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working processes of the units and modules in the above-mentioned device can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above-mentioned method embodiments of this application, a computer program can be used to instruct the relevant hardware to complete. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can at least include: any entity or device that can carry the computer program code, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disc, etc. In some jurisdictions, according to legislation and patent practice, the computer-readable medium cannot be an electrical carrier signal and a telecommunication signal.
[0065] All or part of the processes in the above-mentioned method embodiments of this application can also be completed by a computer program product. When the computer program product runs on a computer device, it enables the computer device to execute and implement the steps in the above-mentioned method embodiments.
[0066] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0067] Those of ordinary skill in the art will realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. A professional technician can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of this application.
[0068] In the embodiments provided in this application, it should be understood that the disclosed device / computer device and method can be implemented in other ways. For example, the device / computer device embodiments described above are merely illustrative. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in electrical, mechanical or other forms.
[0069] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0070] The above embodiments are only used to illustrate the technical solutions of this application, rather than to limit them; although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included in the protection scope of this application. The non-company software tools or components that appear in the embodiments of this application are only introduced by way of example and do not represent actual use.
Claims
1. A large model-based insurance policy identification method, characterized in that: include: Obtain the property insurance target data, coverage data and user information of the corresponding user submitted by the user, and match the corresponding auxiliary assessment data from the preset database according to the property insurance target data; Calculating the expense risk level according to the property insurance target data and the auxiliary assessment data, and determining the corresponding risk feature extraction model according to the expense risk level; Inputting the user information into the risk feature extraction model and outputting risk feature data; According to the user information, matching the associated information from the preset knowledge graph, and screening the risk feature data according to the association relationship between the associated information and the user information to obtain key feature data; The key feature data is evaluated using a user risk calculation model to obtain a user risk assessment result.
2. The insurance policy identification method based on a large model as claimed in claim 1, characterized in that: The matching of related information from a preset knowledge graph according to the user information includes: According to the user information, determining from a preset knowledge graph an entity node representing the same entity as the user corresponding to the user information; In the preset knowledge graph, determine a k-hop neighborhood graph corresponding to the entity node, where k is an integer greater than zero; According to the user information, determining a matching node that matches the user information from the k-hop neighborhood graph; Nodes other than the entity node and the matching node are determined as associated nodes from the k-hop neighborhood graph, and all associated nodes are formed into the associated information.
3. The insurance policy identification method based on a large model as claimed in claim 2, characterized in that: The step of screening the risk feature data according to the association relationship between the association information and the user information to obtain key feature data includes: For any associated node in the associated information, analyzing the associated relationship between the feature represented by the associated node and the entity node; For any risk characteristic data, the risk characteristic data is evaluated according to the association relationship to obtain a risk score; According to the risk score, all risk feature data are screened to obtain the key feature data.
4. The insurance policy identification method based on a large model as claimed in claim 1, characterized in that: The step of calculating the expense risk level according to the property insurance target data and the auxiliary assessment data includes: According to the auxiliary evaluation data, the value of the property corresponding to the property insurance target data is evaluated to obtain an evaluation fee; According to the assessment fee, combined with a preset fee risk level table, the fee risk level corresponding to the property insurance target data is determined.
5. The insurance policy identification method based on a large model as claimed in claim 4, characterized in that: After obtaining the user risk assessment result, the method further includes: According to the user risk assessment result, the expense risk level is adjusted to obtain an adjusted expense risk level; Determining the premium rate according to the user risk assessment result, the adjusted expense risk level and the coverage data; The premium cost is calculated based on the premium rate and the assessment fee.
6. The insurance policy identification method based on a large model as claimed in claim 5, characterized in that: After obtaining the user risk assessment result, the method further includes: A risk assessment report is generated based on the user information, the user risk assessment result, the coverage data, the property target data, the adjusted expense risk level, the premium rate and the premium expense.
7. The insurance policy identification method based on a large model as claimed in claim 1, characterized in that: After obtaining the user risk assessment result, the method further includes: According to the user risk assessment results, matching corresponding risk management strategies; Perform risk management on the user according to the risk management strategy.
8. A large-model-based insurance policy recognition device, characterized in that: include: An acquisition module is used to acquire the property insurance target data, coverage data and user information of the corresponding user submitted by the user, and match the corresponding auxiliary assessment data from the preset database according to the property insurance target data; A first evaluation module is used to calculate the expense risk level according to the property insurance target data and the auxiliary evaluation data, and determine the corresponding risk feature extraction model according to the expense risk level; An extraction module, used for inputting the user information into the risk feature extraction model and outputting risk feature data; A screening module, used to match the associated information from a preset knowledge graph according to the user information, and screen the risk feature data according to the association relationship between the associated information and the user information to obtain key feature data; The second evaluation module is used to evaluate the key feature data using the user risk calculation model to obtain a user risk evaluation result.
9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the insurance policy identification method based on a large model as described in any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the insurance policy identification method based on a large model as described in any one of claims 1 to 7 are implemented.