Method and device for determining personal injury claim settlement mode based on artificial intelligence
Through artificial intelligence technology, legal, medical and insurance models are used to process human injury data, screen and merge claims methods, and solve the problems of inefficient and inaccurate human injury claims methods in the existing technology, achieving a more efficient and accurate claim method Sure.
Patent Information
- Application Number
- CN202411214902.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-06
AI Technical Summary
In the prior art, insurance claims makers are less efficient in determining the human injury claims method and have a single information, resulting in the determined claims method being inaccurate enough.
Using an artificial intelligence-based approach, obtaining claims for historically similar events by obtaining data on human injury events and inputting them into pre-trained legal, medical and insurance models, and filtering and combining them to determine a more accurate claim method. .
The efficiency and accuracy of human injury claims are improved, and a more suitable claim method is determined to take into account different aspects of law, medical care and insurance.
Smart Images

Figure CN119090656A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and in particular to a method and device for determining a personal injury claims settlement method based on artificial intelligence. Background Art
[0002] Insurance can provide protection for users. In related technologies, the claim method for personal injury incidents can be determined by claims adjusters. However, the efficiency of determining the claim method by claims adjusters is low. In addition, for the same claim time, the information held by claims adjusters is relatively single, resulting in the claim method determined by the claims adjusters not being suitable for personal injury incidents, that is, the determined claim method is not accurate. Summary of the invention
[0003] The embodiments of the present application provide a method and device for determining a personal injury claims settlement method based on artificial intelligence, which can improve the efficiency and accuracy of determining the claims settlement method.
[0004] The technical solution of the embodiment of the present application is implemented as follows:
[0005] The present application embodiment provides a method for determining a personal injury claims settlement method based on artificial intelligence, the method comprising:
[0006] Acquire personal injury data of a personal injury event, wherein the personal injury data includes accident data of the personal injury event, object data of the personal injury event, and injury condition data of the personal injury event;
[0007] Inputting the personal injury data into a pre-trained legal model to obtain a first claim settlement method determined based on a historical legal event of the same type as the personal injury event;
[0008] Inputting the personal injury data into a pre-trained medical model to obtain a second claim settlement method determined based on the object data and target data, wherein the target data is obtained after screening the accident data and the injury data;
[0009] Inputting the personal injury data into a pre-trained insurance model to obtain a third claim settlement method determined based on a historical insurance event whose similarity to the personal injury data is greater than a preset first similarity threshold;
[0010] The first claim settlement method, the second claim settlement method and the third claim settlement method are screened and / or combined to obtain a target claim settlement method.
[0011] An embodiment of the present application provides a device for determining a personal injury claims settlement method based on artificial intelligence, and the device includes:
[0012] An acquisition module, used to acquire the personal injury data of the personal injury event, wherein the personal injury data includes the accident data of the personal injury event, the object data of the personal injury event and the injury data of the personal injury event;
[0013] A determination module, configured to input the personal injury data into a pre-trained legal model to obtain a first claim settlement method determined based on a historical legal event of the same type as the personal injury event;
[0014] The determination module is further used to input the personal injury data into a pre-trained medical model to obtain a second claim settlement method determined based on the object data and the target data, wherein the target data is obtained after screening the accident data and the injury data;
[0015] The determination module is further used to input the personal injury data into a pre-trained insurance model to obtain a third claim settlement method determined based on a historical insurance event whose similarity to the personal injury data is greater than a preset first similarity threshold;
[0016] The determination module is also used to screen and / or merge the first claim settlement method, the second claim settlement method and the third claim settlement method to obtain a target claim settlement method.
[0017] An embodiment of the present application provides an electronic device, the electronic device comprising:
[0018] Memory for storing computer executable instructions or computer programs;
[0019] The processor is used to implement the method for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiment of the present application when executing the computer executable instructions or computer programs stored in the memory.
[0020] An embodiment of the present application provides a computer-readable storage medium storing a computer program or computer-executable instructions, which is used to implement the method for determining the personal injury claims settlement method based on artificial intelligence provided by the embodiment of the present application when executed by a processor.
[0021] An embodiment of the present application provides a computer program product, including a computer program or computer executable instructions. When the computer program or computer executable instructions are executed by a processor, the method for determining the personal injury claims settlement method based on artificial intelligence provided by the embodiment of the present application is implemented.
[0022] The embodiments of the present application have the following beneficial effects:
[0023] The method for determining the personal injury claims method based on artificial intelligence provided in the embodiment of the present application can process personal injury data through pre-trained legal models, medical models, and insurance models, and thus can obtain a first claims method determined based on historical legal events of the same type as the personal injury event, a second claims method determined based on object data and target data, and a third claims method determined based on historical insurance events whose similarity to the personal injury data is greater than a preset first similarity threshold. The first claims method, the second claims method, and the third claims method can then be screened and / or merged to obtain the target claims method.
[0024] That is to say, the embodiments of the present application can respectively determine the claims settlement methods corresponding to the law, medical treatment and insurance, so as to obtain a target claims settlement method that integrates the law, medical treatment and insurance. Compared with the claims settlement method determined by the claims adjusters, it can combine the three different claims settlement methods of law, medical treatment and insurance to determine a more accurate target claims settlement method for personal injury incidents, and can improve the efficiency of determining the target claims settlement method for personal injury incidents. That is, the method for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiments of the present application can improve the efficiency and accuracy of determining the claims settlement method. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 It is a structural diagram of a system for determining a personal injury claims settlement method based on artificial intelligence provided in an embodiment of the present application;
[0026] Figure 2 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application;
[0027] Figure 3 This is a flowchart of a method for determining a personal injury claim settlement method based on artificial intelligence provided in an embodiment of the present application. Figure 1 ;
[0028] Figure 4 It is a schematic diagram of the structure of the legal model provided in the embodiment of the present application;
[0029] Figure 5 This is a schematic diagram of the structure of the medical model provided in the embodiment of the present application. Figure 1 ;
[0030] Figure 6 This is a schematic diagram of the structure of the medical model provided in the embodiment of the present application. Figure 2 ;
[0031] Figure 7 This is a schematic diagram of the structure of the medical model provided in the embodiment of the present application. Figure 3 ;
[0032] Figure 8 It is a structural diagram of the insurance model provided in the embodiment of the present application;
[0033] Fig. 9 This is a flowchart of a method for determining a personal injury claim settlement method based on artificial intelligence provided in an embodiment of the present application. Figure 2 . DETAILED DESCRIPTION
[0034] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.
[0035] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.
[0036] In the following description, the terms "first\second\third" involved are merely used to distinguish similar target objects and do not represent a specific ordering of the target objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described here can be implemented in an order other than that illustrated or described here.
[0037] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.
[0038] Unless otherwise defined, all technical and scientific terms used in the embodiments of the present application have the same meanings as those commonly understood by those skilled in the art. The terms used in the embodiments of the present application are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.
[0039] The relevant data collection and processing in the embodiments of this application should be strictly in accordance with the requirements of relevant laws and regulations when applied in examples, obtain the informed consent or separate consent of the personal information subject, and carry out subsequent data use and processing within the scope of authorization of laws and regulations and the personal information subject.
[0040] The embodiments of the present application provide a method, device, electronic device, computer-readable storage medium and computer program product for determining a personal injury claims settlement method based on artificial intelligence, so as to improve the efficiency and accuracy of determining the claims settlement method.
[0041] See also Figure 1 , Figure 1 is a structural diagram of a system for determining a personal injury claims settlement method based on artificial intelligence provided in an embodiment of the present application, Figure 1 The AI-based personal injury claims determination system 100 shown can support an AI-based personal injury claims determination application. The terminal 400 is connected to the server 200 via the network 300. The network 300 can be a wide area network or a local area network, or a combination of the two.
[0042] In the case where the user needs to determine the personal injury claim method, in response to a claim method determination instruction triggered by the graphical interface 410 - 1 of the terminal 400 , the terminal 400 may send a claim method determination request to the server 200 .
[0043] After receiving the claim settlement method determination request, the server 200 can obtain personal injury data of the personal injury event in response to the claim settlement method determination request, wherein the personal injury data includes accident data of the personal injury event, object data of the personal injury event, and injury data of the personal injury event.
[0044] Then, the personal injury data can be input into a pre-trained legal model to obtain a first claim settlement method determined based on a historical legal event of the same type as the personal injury event; the personal injury data can be input into a pre-trained medical model to obtain a second claim settlement method determined based on the object data and the target data, wherein the target data is obtained after screening the accident data and the injury data; the personal injury data can be input into a pre-trained insurance model to obtain a third claim settlement method determined based on a historical insurance event whose similarity to the personal injury data is greater than a preset first similarity threshold. The first claim settlement method, the second claim settlement method, and the third claim settlement method are screened and / or combined to obtain a target claim settlement method.
[0045] That is to say, the embodiments of the present application can respectively determine the claims settlement methods corresponding to the law, medical treatment and insurance, so as to obtain a target claims settlement method that integrates the law, medical treatment and insurance. Compared with the claims settlement method determined by the claims adjusters, it can combine the three different claims settlement methods of law, medical treatment and insurance to determine a more accurate target claims settlement method for personal injury incidents, and can improve the efficiency of determining the target claims settlement method for personal injury incidents. That is, the method for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiments of the present application can improve the efficiency and accuracy of determining the claims settlement method.
[0046] The electronic device provided in the embodiment of the present application is introduced below. The electronic device provided in the embodiment of the present application can be implemented as a terminal, a server, or a combination of the two. Among them, the terminal can be various types of terminals such as a laptop computer, a tablet computer, a desktop computer, a set-top box, a smart phone, a smart speaker, a smart watch, a smart TV, and a car terminal.
[0047] The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms. The terminal and the server can be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiments of the present application.
[0048] See also Figure 2 , Figure 2 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, Figure 2 The electronic device shown includes: at least one processor 410, a memory 450, at least one network interface 420 and a user interface 430. The various components in the electronic device are coupled together via a bus system 440. It is understood that the bus system 440 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 440 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, the bus system 440 is not shown in FIG. Figure 2 Various buses are labeled as bus system 440 .
[0049] The processor 410 can be an integrated circuit chip with signal processing capabilities, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., where the general-purpose processor can be a microprocessor or any conventional processor, etc.
[0050] The user interface 430 includes one or more output devices 431 that enable presentation of media content, including one or more speakers and / or one or more visual display screens. The user interface 430 also includes one or more input devices 432, including user interface components that facilitate user input, such as a keyboard, mouse, microphone, touch screen display, camera, other input buttons and controls.
[0051] The memory 450 may be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid-state memory, hard disk drives, optical disk drives, etc. The memory 450 may optionally include one or more storage devices that are physically remote from the processor 410.
[0052] The memory 450 includes a volatile memory or a non-volatile memory, and may also include both volatile and non-volatile memories. The non-volatile memory may be a read-only memory (ROM), and the volatile memory may be a random access memory (RAM). The memory 450 described in the embodiments of the present application is intended to include any suitable type of memory.
[0053] In some embodiments, memory 450 can store data to support various operations, examples of which include programs, modules, and data structures, or a subset or superset thereof, as exemplarily described below.
[0054] Operating system 451, including system programs for processing various basic system services and performing hardware-related tasks, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks;
[0055] A network communication module 452, used to reach other electronic devices via one or more (wired or wireless) network interfaces 420, exemplary network interfaces 420 include: Bluetooth, Wireless Compatibility Certification (WiFi), and Universal Serial Bus (USB), etc.;
[0056] a presentation module 453 for enabling presentation of information via one or more output devices 431 (e.g., display screen, speaker, etc.) associated with the user interface 430 (e.g., a user interface for operating peripherals and displaying content and information);
[0057] The input processing module 454 is used to detect one or more user inputs or interactions from one of the one or more input devices 432 and translate the detected inputs or interactions.
[0058] In some embodiments, the device for determining the personal injury claims method based on artificial intelligence provided in the embodiments of the present application can be implemented in a software manner. Figure 2The device 455 for determining the compensation method for personal injury based on artificial intelligence stored in the memory 450 is shown, which can be software in the form of a program and a plug-in, etc., and includes the following software modules: an acquisition module 4551 and a determination module 4552. These modules are logical, and therefore can be arbitrarily combined or further split according to the functions implemented. The functions of each module will be described below.
[0059] In other embodiments, the device for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiments of the present application can be implemented in hardware. As an example, the device for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiments of the present application can be a processor in the form of a hardware decoding processor, which is programmed to execute the method for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiments of the present application. For example, the processor in the form of a hardware decoding processor can adopt one or more application specific integrated circuits (Application Specific Integrated Circuit, ASIC), digital signal processors (Digital Signal Processor, DSP), programmable logic devices (Programmable Logic Device, PLD), complex programmable logic devices (Complex Programmable Logic Device, CPLD), field programmable gate arrays (Field-Programmable Gate Array, FPGA) or other electronic components.
[0060] In some embodiments, the terminal or server can implement the method for determining the personal injury claims method based on artificial intelligence provided by the embodiment of the present application by running various computer executable instructions or computer programs. For example, computer executable instructions can be microprogram-level commands, machine instructions or software instructions. The computer program can be a native program or software module in the operating system; it can be a native application (APPlication, APP), that is, a program that needs to be installed in the operating system to run, such as an insurance-related APP; it can also be a small program that can be embedded in any APP, that is, a program that only needs to be downloaded to the browser environment to run. In short, the above-mentioned computer executable instructions can be instructions in any form, and the above-mentioned computer program can be an application, module or plug-in in any form.
[0061] The following is an introduction to the method for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiment of the present application. As mentioned above, the electronic device that implements the method for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiment of the present application can be a terminal, a server, or a combination of the two.
[0062] See also Figure 3 , Figure 3 This is a flowchart of a method for determining a personal injury claim settlement method based on artificial intelligence provided in an embodiment of the present application. Figure 1 , the following combination Figure 3 And taking the server as an example, the method for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiment of the present application is explained.
[0063] In step 101, personal injury data of a personal injury event is obtained.
[0064] When the user needs to determine the claim settlement method for a personal injury incident, the user can first obtain the personal injury data of the personal injury incident, wherein a personal injury incident refers to an incident in which a target object (which can be the insured or a third party) suffers personal injury due to the occurrence of an accident during the validity period of the insurance contract. The user can be the target object (i.e., the insured or a third party), other objects that have an associated relationship with the target object but are not injured, or claims adjusters of the insurance company. For example, the associated relationship can be a relative relationship, a friend relationship, etc.
[0065] The personal injury data may be in the form of voice, video, image, text, etc. The personal injury data may include the accident data of the personal injury event, the object data of the personal injury event, and the injury data of the personal injury event. The accident data may include one or more of the data representing the accident location, the data representing the accident time, the data representing the accident type, the data representing the accident level, and the data used to describe the situation of the accident. The accident data may be different for different accidents.
[0066] Object data refers to the data of the target object involved in the human injury incident. The target object may be a person, an animal, etc. The object data may include one or more of the following: data corresponding to the target object identifier, data representing the target object's age, data representing the target object's gender, data describing the target object's height, data describing the target object's weight, and data representing the region to which the target object belongs. Object data may also include data of other target objects that are associated with the target object.
[0067] Injury data refers to data corresponding to the injury suffered by the target object caused by the human injury event. The injury data may include one or more of data corresponding to the injury identification, data describing the severity of the injury, data used to describe the injury, data representing the treatment method of the injury, and data representing the event causing the injury. Through step 101, accurate human injury data can be obtained.
[0068] In step 102, the personal injury data is input into a pre-trained legal model to obtain a first claim settlement method determined based on a historical legal event of the same type as the personal injury event.
[0069] In some embodiments, step 102 may be implemented by inputting personal injury data into a pre-trained legal model, and the legal model may output a first claim settlement method determined based on historical legal events of the same type as the personal injury event.
[0070] The first claim settlement method includes one or more first claim settlement sub-methods, and the first claim settlement sub-methods may be a claim amount, a payment method for the claim amount, a care method for the target object in the personal injury incident, a method for observing the injury of the target object in the personal injury incident, etc., which may be set in accordance with actual conditions. In other words, the first claim settlement method may include one or more of the claim amount, a payment method for the claim amount, a care method for the target object in the personal injury incident, and a method for observing the injury of the target object in the personal injury incident.
[0071] Among them, the payment method for the claim amount may include the channel for paying the claim amount, the time for paying the claim amount, the target object for paying the claim amount, etc. The care method for personal injury incidents may include whether to care for the target object in the personal injury incident, the content of care for the target object in the personal injury incident, the number of times the care is provided, the purpose of the care, etc. The injury observation method for the target object in the personal injury incident may include whether to observe the injury of the target object in the personal injury incident, the number of times the injury is observed, the purpose of the injury observation, etc.
[0072] In some embodiments, the type of personal injury incident may be classified according to one or more of the degree of injury of the target object, the field of the personal injury incident, the cause of the accident, the level of the accident, the nature of the injury of the target object, and the injured part of the target object, wherein the degree of injury of the target object includes light injury and severe injury, etc. The field of the personal injury incident may be production safety accident, fire accident, road traffic accident, etc.
[0073] The causes of the accident may include object impact accidents, vehicle injury accidents, mechanical injury accidents, etc. The nature of the target object's injury may include contusion, rolling injury, crushing injury, etc. The target object's injured parts may include hands, face, legs, etc., which can be set according to the actual situation.
[0074] The type of personal injury incident can correspond to a legal model. After obtaining the personal injury data, keywords can be extracted from the personal injury data to obtain keywords that characterize the type of personal injury incident, so as to obtain the legal model corresponding to the keywords. By inputting the personal injury data into the legal model corresponding to the keywords, the first claim settlement method output by the legal model corresponding to the keywords can be obtained, that is, the first claim settlement method determined based on historical legal events of the same type as the personal injury incident can be obtained.
[0075] In some embodiments, any legal model may include a first feature extraction layer, a first feature fusion layer, and a claim settlement method acquisition layer. In some embodiments, the first feature extraction layer may include a first text feature extraction layer and a first image feature extraction layer. The claim settlement method acquisition layer may include a first amount prediction layer, a legal action prediction layer, and a method matching layer.
[0076] For example, see Figure 4 , Figure 4 It is a structural diagram of the legal model provided in an embodiment of the present application. The legal model 410 may include a first text feature extraction layer 401, a first image feature extraction layer 402, a first feature fusion layer 403, a first amount prediction layer 404, a legal action prediction layer 405 and a method matching layer 406.
[0077] The personal injury data may include text personal injury data and image personal injury data, wherein the text personal injury data may be personal injury data in text form, and in some embodiments, the text personal injury data may also be obtained by performing text extraction on voice, video, image, etc. The first text feature extraction layer may perform feature extraction on the text personal injury data to obtain a first text feature.
[0078] The image personal injury data may be personal injury data in image form. In some embodiments, the image personal injury data may also be obtained by selecting key frames in a video. The first image feature extraction layer may perform feature extraction on the image personal injury data to obtain a first image feature.
[0079] The first feature fusion layer can fuse the first text feature and the first image feature to obtain a first fusion feature, wherein the way of fusing the first text feature and the first image feature can be feature splicing, weighted summation, etc., which can be specifically set according to actual usage requirements.
[0080] After obtaining the fusion feature, the first amount prediction layer can be a regression network, and the first amount prediction layer can output the first predicted amount based on the first fusion feature. The legal action prediction layer can be a classification network, and the legal action prediction layer can output the probability of executing a legal action for the personal injury incident based on the first fusion feature.
[0081] After obtaining the first predicted amount and the probability of taking legal action against the personal injury incident, the method matching layer can output a first claim settlement method based on the first predicted amount and the probability of taking legal action against the personal injury incident. The first claim settlement method includes a claim settlement sub-method associated with the probability of taking legal action against the personal injury incident and a claim settlement sub-method not associated with the probability of taking legal action against the personal injury incident.
[0082] The following is an explanation of the training method of any type of legal model. In some embodiments, historical personal injury events, historical claim amounts of historical personal injury events, whether legal actions were taken for historical personal injury events, the first historical claim method after legal actions were taken for historical personal injury events, and the second historical claim method for historical personal injury events can be obtained.
[0083] The historical personal injury events are input into the legal model to be trained. The legal model can extract the first historical text feature and the second historical image feature of the historical personal injury events. The first historical text feature and the first historical image feature are fused to obtain the first historical fusion feature. Furthermore, the second predicted amount can be obtained based on the first historical fusion feature. The parameters of the first amount prediction layer can be updated based on the difference between the second predicted amount and the historical claim amount.
[0084] The probability of executing legal action for personal injury incidents can also be obtained based on the second fusion feature, and the difference between legal action prediction layers can be updated based on the difference between the probability of executing legal action for personal injury incidents and whether legal action was executed in historical personal injury incidents.
[0085] After obtaining the second predicted amount and the probability of taking legal action for the personal injury incident, the predicted claim method can be obtained based on the second predicted amount and the probability of taking legal action for the personal injury incident, and the parameters of the legal action prediction layer can be updated according to the difference between the predicted claim method and the target historical claim method.
[0086] The target historical claim settlement method is determined based on the first historical claim settlement method and the second historical claim settlement method. In some embodiments, the union of the first historical claim settlement method and the second historical claim settlement method can be used as the target historical claim settlement method. In some embodiments, the first historical claim settlement method and the second historical claim settlement method can also be manually merged to obtain the target historical claim settlement method.
[0087] In some embodiments, any one of the first historical claim settlement method and the second historical claim settlement method can also be used as the target historical claim settlement method. In some embodiments, the predicted claim settlement method can also be compared with the first historical claim settlement method to obtain a first similarity, and the predicted claim settlement method can be compared with the second historical claim settlement method to obtain a second similarity, and then the historical claim settlement method corresponding to the larger similarity between the first similarity and the second similarity can be used as the target historical claim settlement method.
[0088] In some embodiments, when the probability of taking legal action against the personal injury incident is greater than a preset probability threshold, the first claim settlement sub-method includes a claim settlement sub-method associated with the probability of taking legal action against the personal injury incident.
[0089] Among them, the degree of correspondence between the claim sub-method associated with the probability of taking legal action against personal injury incidents and the claim sub-method unrelated to the probability of taking legal action against personal injury incidents can be different. For example, the claim sub-method is the claim amount, and the claim amount associated with the probability of taking legal action against personal injury incidents is greater than the claim amount unrelated to the probability of taking legal action against personal injury incidents.
[0090] Accordingly, during the training process, for a claim settlement sub-method associated with the probability of executing a legal action for a personal injury incident, a first difference between the claim settlement sub-method and the first historical claim settlement method can be obtained, and a second difference between a claim settlement sub-method unrelated to the probability of executing a legal action for a personal injury incident and the second historical claim settlement method can be obtained, and then the parameters of the legal action prediction layer can be updated based on the first difference and the second difference. Through step 102, the first claim settlement method at the legal level can be quickly and accurately obtained.
[0091] In step 103, the personal injury data is input into a pre-trained medical model to obtain a second claim settlement method determined based on the object data and the target data.
[0092] Among them, the second claim settlement method may include one or more second claim settlement sub-methods, and the second claim settlement sub-method may be the claim amount, the payment method for the claim amount, the medical method for the target object in the personal injury incident, the care method for the target object in the personal injury incident, the injury observation method for the target object in the personal injury incident, etc., which can be set specifically according to the actual situation.
[0093] That is, the second claim settlement method may include one or more of the claim amount, the payment method for the claim amount, the medical treatment method for the target object in the personal injury incident, the care method for the target object in the personal injury incident, and the injury observation method for the target object in the personal injury incident. Among them, the medical treatment method for the target object in the personal injury incident may be the duration of the medical treatment, the items involved in the medical treatment, the location of the medical treatment, etc.
[0094] In some embodiments, the medical model includes a first medical sub-model, a second medical sub-model and a third medical sub-model. The first medical sub-model, the second medical sub-model and the third medical sub-model are described below respectively.
[0095] In some embodiments, the accident data and the injury data may be input into the first medical sub-model to obtain target data after screening based on the matching degree between the accident data and the data, wherein the target data is obtained after screening the accident data and the injury data.
[0096] The first medical sub-model may include a second feature extraction layer, a classification layer, and a screening layer, for example, see Figure 5, Figure 5 This is a schematic diagram of the structure of the medical model provided in the embodiment of the present application. Figure 1 , the first medical sub-model 510 may include a second feature extraction layer 501 , a matching layer 502 , and a screening layer 503 .
[0097] The second feature extraction layer 501 is used to extract features from the accident data to obtain accident features. The second feature extraction layer 501 is also used to extract features from the injury data to obtain injury features. The matching layer 502 is used to obtain the matching degree between the accident features and the injury features. The screening layer 503 is used to screen out target data whose matching degree is greater than a preset matching degree threshold. When the matching degree is greater than the preset matching degree threshold, it can be explained that the accident data and the injury data match each other. Correspondingly, when the matching degree is less than or equal to the preset matching degree threshold, it can be explained that the accident data and the injury data do not match each other.
[0098] For accident data and injury data whose matching degree is less than or equal to the preset matching degree threshold, if the accident data contains injury data whose matching degree is greater than the preset matching degree threshold, the injury data whose matching degree is less than or equal to the preset matching degree threshold can be deleted.
[0099] If the accident data does not contain injury data with a matching degree greater than a preset matching degree threshold, the accident data can be marked for manual review, or the target object can be prompted to supplement the injury data based on the accident data, thereby improving the integrity of the injury data. After obtaining the supplemented injury data, the supplemented injury data and the accident data can be input into the first medical sub-model. For details, please refer to the above instructions and will not be repeated here.
[0100] The training method of the first medical sub-model is described below. In some embodiments, historical accident data, historical injury data and historical matching degree may be obtained, wherein the historical matching degree is the matching degree between the historical accident data and the historical injury data.
[0101] The historical accident data and the historical injury data are input into the first medical sub-model to be trained. The first medical sub-model can extract the first historical accident features and the first historical injury features according to the historical accident data and the historical injury data, and then obtain the matching degree between the first historical accident features and the first historical injury features, and then update the matching layer according to the difference between the historical matching degree and the matching degree between the first historical accident features and the first historical injury features.
[0102] The first medical sub-model can output the historical target data, and update the parameters of the screening layer according to the difference between the historical target data and the historical target accident data and the historical target injury data, so as to obtain the first medical sub-model. The historical target accident data and the historical target injury data are data with a historical matching degree greater than a preset matching degree threshold.
[0103] In some embodiments, after the target data is acquired, the target data may be input into the second medical sub-model to obtain a medical treatment method determined based on historical medical events whose similarity to the target data is greater than a second similarity threshold.
[0104] In some embodiments, the second medical sub-model may include a third feature extraction layer, a second feature fusion layer, and a medical method acquisition layer. In some embodiments, the third feature extraction layer may include a second text feature extraction layer and a second image feature extraction layer.
[0105] See also Figure 6 , Figure 6 This is a schematic diagram of the structure of the medical model provided in the embodiment of the present application. Figure 2 The second medical sub-model 610 includes a third feature extraction layer, a second feature fusion layer 603 and a medical method acquisition layer 604 . The third feature extraction layer includes a second text feature extraction layer 601 and a second image feature extraction layer 602 .
[0106] The target data may include text target data and image target data, wherein the text target data may be target data in text form, and in some embodiments, the text target data may also be obtained by performing text extraction on voice, video, image, etc. The second text feature extraction layer may perform feature extraction on the text target data to obtain a second text feature.
[0107] The image target data may be target data in the form of an image. In some embodiments, the image target data may also be obtained by selecting key frames in a video. The first image extraction layer may perform feature extraction on the image target data to obtain second image features.
[0108] The second feature fusion layer can fuse the second text feature and the second image feature to obtain a second fusion feature, wherein the second text feature and the second image feature can be fused in a manner of feature splicing, weighted summation, etc., which can be specifically set according to actual usage requirements.
[0109] Furthermore, the medical method acquisition layer includes multiple neurons, each neuron corresponds to a candidate medical method, and the medical method acquisition layer can determine the matching degree between the second fusion feature and each candidate medical method based on the second fusion feature, so that the candidate treatment method with the highest matching degree can be used as the medical method output by the second medical sub-model.
[0110] In some embodiments, each neuron may also correspond to a candidate care method, that is, a candidate medical method may have a corresponding candidate care method, and accordingly, the candidate care method corresponding to the candidate treatment method with the highest matching degree may also be used as the output of the second medical sub-model, that is, the second medical sub-model may output the medical method and the care method. Of course, the injury observation method may also be output, which may be set according to actual use requirements. For the description of outputting the injury observation method, the above content may be replaced with the same content, which will not be elaborated here.
[0111] The following is an explanation of the training method of the second medical sub-model. Historical medical events can be obtained, including historical accident data and historical injury data. The historical accident data and historical injury data are input into the second medical sub-model to be trained. The second medical sub-model can obtain second historical text features and second historical image features.
[0112] By fusing the second historical text feature and the second historical image feature, a second historical fusion feature can be obtained, and then the matching degree between the second fusion feature and each candidate historical medical treatment method can be obtained, so that the candidate historical treatment method with the highest matching degree can be used as the historical treatment method. Therefore, the parameters of the second medical sub-model can be updated based on the difference between the historical treatment method and the treatment method corresponding to the historical medical event.
[0113] In some embodiments, after obtaining the medical treatment method, the medical treatment method and the object data can be input into the third medical sub-model to obtain a medical claim amount predicted by combining the medical treatment method and the object data. In some embodiments, the amount range corresponding to the medical treatment method can be determined based on the age of the target object and / or the region to which the target object belongs, so that any amount in the range can be used as the medical claim amount.
[0114] In some embodiments, the third medical sub-model may include a fourth feature extraction layer, a third feature fusion layer, and a second amount prediction layer, see Figure 7 , Figure 7 This is a schematic diagram of the structure of the medical model provided in the embodiment of the present application. Figure 3 The third medical sub-model 710 may include a fourth feature extraction layer 701 , a third feature fusion layer 702 , and a second amount prediction layer 703 .
[0115] The fourth feature extraction layer is used to extract the characteristics of the medical method and the characteristics of the object data. The third feature fusion layer is used to fuse the characteristics of the medical method and the characteristics of the object data to obtain the fourth fused feature. The second amount prediction layer can output the medical claim amount based on the third fused feature.
[0116] The following is an explanation of the training method of the third medical sub-model. The historical medical methods and historical object data are obtained, and the historical medical methods and historical object data are input into the third medical sub-model to be trained. The third medical sub-model can extract the characteristics of the historical medical methods and the characteristics of the historical object data, and then the characteristics of the historical medical methods and the characteristics of the historical object data can be fused to obtain the third historical fusion characteristics. The predicted claim amount is output based on the third historical fusion characteristics. The parameters of the third medical sub-model can be updated according to the difference between the predicted claim amount and the actual claim amount, so as to obtain the third medical sub-model. Through step 103, the second claim method at the legal level can be quickly and accurately obtained.
[0117] In step 104, the personal injury data is input into a pre-trained insurance model to obtain a third claim settlement method determined based on historical insurance events whose similarity to the personal injury data is greater than a preset first similarity threshold.
[0118] In some embodiments, step 104 can be implemented in the following manner: personal injury data can be input into a pre-trained insurance model to obtain a third claim settlement method determined based on historical insurance events whose similarity to the personal injury data is greater than a preset first similarity threshold.
[0119] Among them, the third claim settlement method includes one or more third claim settlement sub-methods, and the third claim settlement sub-methods can be the claim amount, the payment method for the claim amount, the care method for the target object in the personal injury incident, the injury observation method for the target object in the personal injury incident, the insurance recommendation method for the target object in the personal injury incident, etc., which can be set according to actual conditions. In other words, the first claim settlement method may include one or more of the claim amount, the payment method for the claim amount, the care method for the target object in the personal injury incident, the injury observation method for the target object in the personal injury incident, and the insurance recommendation method for the target object in the personal injury incident.
[0120] Among them, the insurance recommendation method for the target object in the personal injury incident may include a supplementary insurance method for the insurance corresponding to the personal injury incident, a replacement insurance method for the insurance corresponding to the personal injury incident, and an insurance method recommended for the target object, which can be set specifically according to actual usage.
[0121] In some embodiments, the insurance model includes a fifth feature extraction layer, a fourth feature fusion layer, a third amount prediction layer, and a method generation layer. Figure 8 , Figure 8 It is a structural diagram of the insurance model provided in an embodiment of the present application. The insurance model 810 includes a fifth feature extraction layer 801, a fourth feature fusion layer 802, a third amount prediction layer 803 and a method generation layer 804.
[0122] The first feature extraction layer may include a third text feature extraction layer and a third image feature extraction layer, and the third text feature extraction layer may perform feature extraction on text injury data to obtain a third text feature. The third image feature extraction layer may perform feature extraction on image injury data to obtain a third image feature.
[0123] The third text feature extraction layer is also used to extract features of the insurance information corresponding to the personal injury incident to obtain insurance features, wherein the insurance information can be identification information of the insurance purchased by the target object in the personal injury incident, and the insurance information can also be item information involved in the insurance purchased by the target object in the personal injury incident that needs to be claimed.
[0124] The fourth feature fusion layer can fuse the third text feature, the third image feature and the insurance feature to obtain a fourth fused feature. The third amount prediction layer is used to predict the amount based on the fourth fused feature to obtain the insurance claim amount.
[0125] After obtaining the insurance claim amount, the insurance claim amount, object data and additional information can be input into the method generation layer, so as to obtain the third claim method. The additional information can be preset by the user. For example, the additional information can be a care activity launched by the insurance company. It can be determined whether the target object needs to participate in the care activity based on the object data. If the target object can participate, the care method can be obtained based on the corresponding relationship between the insurance claim amount and the care activity.
[0126] For another example, the attachment information may also be an insurance claim amount threshold. When the insurance claim amount is greater than the insurance claim amount threshold, the injury observation method corresponding to the interval is determined from the correspondence between the pre-set candidate object data and the candidate injury observation methods.
[0127] For another example, the additional information may be the target object's preference information for insurance items included in the object data, and then an insurance recommendation method for the target object in the personal injury incident may be generated based on the insurance claim amount and the preference information.
[0128] The following is an explanation of the training method of the insurance model. Historical insurance events, historical insurance claim amounts and historical additional information can be obtained, and the historical insurance events can be input into the insurance model to be trained. The third historical text feature, the second historical image feature and the historical insurance feature of the historical insurance event can be extracted, and the third historical text feature, the second historical image feature and the historical insurance feature can be fused to obtain the fourth historical fusion feature.
[0129] Based on the fourth historical fusion feature, the amount is predicted to obtain the predicted insurance claim amount. Based on the difference between the predicted insurance claim amount and the historical insurance claim amount, the parameters of the third amount prediction layer are updated. The historical additional information and the predicted insurance claim amount are input into the method generation layer, which belongs to the predicted claim method. Based on the difference between the predicted claim method and the claim method corresponding to the historical insurance event, the parameters of the method generation layer are updated, so that the insurance model can be obtained. Through step 104, the third claim method at the legal level can be quickly and accurately obtained.
[0130] In step 105, the first claim settlement method, the second claim settlement method and the third claim settlement method are screened and / or combined to obtain a target claim settlement method.
[0131] In some embodiments, step 105 is implemented in the following manner: the first claim settlement method, the second claim settlement method, and the third claim settlement method all include at least one type of claim settlement sub-method. Specifically, the first claim settlement method includes one or more of a claim amount, a payment method for the claim amount, a care method for a target object in a personal injury event, and an injury observation method for a target object in a personal injury event.
[0132] The second claim settlement method includes one or more of the claim amount, the payment method for the claim amount, the medical treatment method for the target object in the personal injury incident, the care method for the target object in the personal injury incident, and the injury observation method for the target object in the personal injury incident.
[0133] The third claim settlement method includes one or more of the claim amount, the payment method for the claim amount, the care method for the target object in the personal injury incident, the injury observation method for the target object in the personal injury incident, and the insurance recommendation method for the target object in the personal injury incident. In other words, the first claim settlement method, the second claim settlement method, and the third claim settlement method include claim sub-methods of the same type and claim sub-methods of different types.
[0134] See also Fig. 9 , Fig. 9 This is a flowchart of a method for determining a personal injury claim settlement method based on artificial intelligence provided in an embodiment of the present application. Figure 2 , the following combination Fig. 9 The steps shown are Figure 3 Step 105 is shown for explanation.
[0135] In step 1051, a processing method for the claim sub-method is determined based on the object data, the type corresponding to the first claim settlement method, the type corresponding to the second claim settlement method, and the type corresponding to the third claim settlement method.
[0136] In some embodiments, it can be determined whether the type corresponding to the first claim method, the type corresponding to the second claim method, and the type corresponding to the third claim method include an amount type, wherein the claim sub-method corresponding to the amount type is the claim amount.
[0137] If there is an amount type among the types corresponding to the first claim method, the types corresponding to the second claim method, and the types corresponding to the third claim method, that is, the first claim method includes the claim amount, the second claim method includes the claim amount, and the third claim method includes the claim amount, then the calculation method corresponding to the object data can be determined based on the object data and the mapping relationship between the pre-set candidate object data and the candidate calculation method.
[0138] In some embodiments, the object data may include the region where the target object is located, the age of the target object, the health status of the target object, etc. Different regions may be set with different calculation methods, and the calculation method corresponding to the region where the target object is located may be determined.
[0139] Different age intervals can be set with different calculation methods, and the calculation method corresponding to the age of the target object can be determined. Different health conditions correspond to different calculation methods, and the calculation method corresponding to the health condition of the target object can be determined. It can be seen that the calculation method corresponding to the object data includes one of the calculation method corresponding to the area where the target object is located, the calculation method corresponding to the age of the target object, and the calculation method corresponding to the health condition of the target object.
[0140] When multiple different calculation methods are determined, the target calculation method can be determined according to the pre-set weights from the calculation method corresponding to the area where the target object is located, the calculation method corresponding to the age of the target object, and the calculation method corresponding to the health status of the target object, and the target calculation method can be used as the calculation method corresponding to the object data, wherein the calculation method can be determined as the processing method.
[0141] In some embodiments, if the number of types corresponding to the first claim settlement method, the type corresponding to the second claim settlement method, and the type corresponding to the third claim settlement method are all multiple, the grouping method corresponding to the object data can be determined from pre-set candidate object data and candidate grouping methods based on the object data.
[0142] In some embodiments, the object data may include the region where the target object is located, the age of the target object, the health status of the target object, etc. Different regions may be set with different grouping methods, and the grouping method corresponding to the region where the target object is located may be determined.
[0143] Different age intervals can be set with different grouping methods, and the grouping method corresponding to the age of the target object can be determined. Different health conditions correspond to different grouping methods, and the grouping method corresponding to the health condition of the target object can be determined. It can be seen that the grouping method corresponding to the object data includes one of the grouping method corresponding to the area where the target object is located, the grouping method corresponding to the age of the target object, and the grouping method corresponding to the health condition of the target object.
[0144] When multiple different grouping methods are determined, a target grouping method can be determined from the grouping methods corresponding to the region where the target object is located, the grouping methods corresponding to the age of the target object, and the grouping methods corresponding to the health status of the target object according to the preset weights, and the target grouping method is used as the grouping method corresponding to the object data, wherein the grouping method can be determined as a processing method. The calculation method and / or the grouping method can be determined as the processing method, that is, the processing method can include the calculation method and / or the grouping method.
[0145] In some embodiments, the grouping method may be to take the claim amount as one group and the remaining claim sub-methods as another group. In some embodiments, the grouping method may also be to take the claim amount as the first group, the care method for the target object in the personal injury incident as the second group, and the remaining claim sub-methods as the third group. The specific setting may be made according to actual use requirements.
[0146] In some embodiments, each claim sub-method may have a pre-set importance level, and each claim sub-method may be grouped according to the importance level. For the same type of claim sub-method, the importance level belonging to the first claim method may be different from the importance level belonging to the second claim method. Correspondingly, for the same type of claim sub-method, the importance level belonging to the first claim method may be different from the importance level belonging to the third claim method.
[0147] In some embodiments, the importance of the claims sub-method may be output by the corresponding model, that is, the importance may be considered from the legal level, medical level, and insurance level respectively, so as to obtain the importance of the claims sub-method at different levels.
[0148] In step 1052, based on the target processing method, the claim sub-methods included in the first claim settlement method, the claim sub-methods included in the second claim settlement method, and the claim sub-methods included in the third claim settlement method are screened and / or merged to obtain the target claim settlement method.
[0149] In some embodiments, step 1052 can be implemented in the following ways. In a first implementation method, based on a calculation method, a first target claim sub-method is determined from the claim sub-methods included in the first claim method, the claim sub-methods included in the second claim method, and the claim sub-methods included in the third claim method.
[0150] In some embodiments, the calculation method may include one or more weights, and the weight may be any one or more weights of the claim amount included in the first claim settlement method, the claim amount included in the second claim settlement method, and the claim amount included in the third target claim settlement method.
[0151] Based on the calculation method, the claim amount can be used as the first target claim sub-method from the claim sub-methods included in the first claim method, the claim sub-methods included in the second claim method, and the claim sub-methods included in the third claim method. Furthermore, the target claim method can be determined based on the calculation method and the first target claim sub-method.
[0152] In some embodiments, the weight included in the calculation method can be used, and the claim amount with a weight greater than a preset weight threshold can be used as the first target claim sub-method, and then the target claim method can be determined based on the calculation method and the first target claim sub-method. The first target claim sub-method includes one or more of the claim amount included in the first claim method, the claim amount included in the second claim method, and the claim amount included in the third target claim method.
[0153] In some embodiments, after determining the calculation method corresponding to the object data, the calculation method can be used to calculate one or more of the claim amounts included in the first claim method, the claim amount included in the second claim method, and the claim amount included in the third target claim method to obtain the target claim amount.
[0154] In some embodiments, if the first target claim settlement sub-method only includes one of the claim amounts included in the first claim settlement method, the claim amount included in the second claim settlement method, and the claim amount included in the third target claim settlement method, the product of the weight and the claim amount can be used as the target claim amount.
[0155] In some embodiments, if the first target claim settlement sub-method includes multiple claims included in the first claim settlement method, the second claim settlement method, and the third target claim settlement method, the calculation method also includes an algorithm for the claims included in the first claim settlement method, the claims included in the second claim settlement method, and the claims included in the third target claim settlement method, for example, the algorithm may be addition, so that the target claim settlement amount can be calculated.
[0156] In some embodiments, if the first claim settlement method, the second claim settlement method, and the third claim settlement method do not include other claim settlement sub-methods, the target claim amount can be used as the target claim settlement method. In some embodiments, if the type corresponding to the first claim settlement method, the type of the second claim settlement method, and the type of the third claim settlement method are all different, the first claim settlement method, the second claim settlement method, and the third claim settlement method can be combined to obtain the target claim settlement method.
[0157] For example, after obtaining the target claim amount, the first claim method also includes a payment method for the claim amount, the second claim method also includes a care method for the target object in the personal injury incident, and the third claim method also includes an insurance recommendation method for the target object in the personal injury incident. The payment method for the claim amount, the care method for the target object in the personal injury incident, the insurance recommendation method for the target object in the personal injury incident, and the target claim amount can be used as target claim methods.
[0158] For another example, the first claim settlement method also includes the claim amount, the second claim settlement method also includes the care method for the target object in the personal injury incident, and the third claim settlement method also includes the insurance recommendation method for the target object in the personal injury incident. The claim amount, the care method for the target object in the personal injury incident, and the insurance recommendation method for the target object in the personal injury incident can be used as target claim settlement methods.
[0159] The second implementation method is to group the claim sub-methods included in the first claim settlement method, the claim sub-methods included in the second claim settlement method, and the claim sub-methods included in the third claim settlement method based on the grouping method to obtain a target group, which includes the second target claim sub-method.
[0160] For the second target claim settlement sub-method in the target group, the second target claim settlement sub-method is screened and / or merged to obtain the target claim settlement method. In some embodiments, the target group is obtained by grouping the various claim settlement sub-methods according to their importance, wherein the second target claim settlement sub-method is a claim settlement sub-method whose importance is greater than a preset importance threshold.
[0161] For the claim sub-methods whose importance is greater than the preset importance threshold, that is, for the second claim sub-method, the number of second claim sub-methods is multiple. If the types corresponding to the second claim sub-methods are different, then multiple second claim sub-methods can be used as target claim methods. In this way, the second target claim methods can be merged.
[0162] In some embodiments, if there are second claim sub-methods of the same type, for the second claim sub-methods of the same type, the second claim sub-method with the highest importance can be selected, thereby screening the second target claim sub-method and obtaining the target claim method.
[0163] In some embodiments, for the second claim settlement sub-method of the same type, if the second claim settlement sub-method has a number of times, the second claim settlement sub-method with the largest or smallest number of times can be selected, thereby screening the second target claim settlement sub-method to obtain the target claim settlement method. The number of times can be the number of times of care, the number of times of injury observation, etc.
[0164] In some embodiments, when the grouping method is to group the claim amount as one group and the remaining claim sub-methods as another group, the remaining claim sub-methods can be the target grouping, and for the remaining claim sub-methods, one or more types of claim sub-methods can be randomly selected to obtain the target claim sub-method.
[0165] In some embodiments, the number of claim sub-methods (second target claim sub-methods) under each type can be determined. If there are multiple claim sub-methods, the same information can be extracted from the multiple claim sub-methods, and the extracted same information can be used as the target claim method. In this way, the second target claim method can be screened.
[0166] The third implementation method can combine the first implementation method and the second implementation method, that is, for the first target claim sub-method (claim amount), the first target claim sub-method can be determined by the calculation method and the first target claim sub-method. And for the second claim sub-method, the second target claim sub-method can be screened and / or merged to obtain the second target claim sub-method, and then the target claim sub-method can be determined based on the first target claim sub-method and the second target claim sub-method. Through step 105, different implementation methods can be used to determine the precise target claim sub-method.
[0167] In some embodiments, a method for determining a target claim amount based on a calculation method and a claim amount can be achieved through an amount model. After obtaining the claim amount included in the first claim method, the claim amount included in the second claim method, and the claim amount included in the third target claim method, the claim amount included in the first claim method, the claim amount included in the second claim method, and the claim amount included in the third target claim method, as well as object data, can be input into the amount model. The amount model can extract features of the object data, thereby determining a calculation method based on the features of the object data, thereby calculating the claim amount included in the first claim method, the claim amount included in the second claim method, and the claim amount included in the third target claim method through the calculation method, and outputting the target claim amount.
[0168] In some embodiments, the object data includes decision information of whether the target object takes legal action for the personal injury incident, and the first claim settlement method includes a claim settlement sub-method associated with the probability of taking legal action for the personal injury incident.
[0169] exist Figure 3 After step 102 shown, that is, after obtaining the first claim settlement method, the first claim settlement method can be adjusted according to the decision information. If the decision information indicates that no legal action will be taken for the personal injury incident, the decision information will be added to the first claim settlement method, and the claim sub-method associated with the probability in the first claim settlement method will be deleted to obtain an updated first claim settlement method.
[0170] If the decision information represents the execution of a legal decision on a personal injury incident, and the probability of executing a legal action on a personal injury incident predicted by the legal model is less than a preset threshold, the decision information is added to the first claim settlement method, and the weight of the claim settlement sub-method associated with the probability in the first claim settlement method is increased to obtain an updated first claim settlement method.
[0171] If the decision information represents the execution of a legal decision on a personal injury incident, and the probability of executing a legal action on a personal injury incident predicted by the legal model is greater than a preset threshold, then the weight of the claim sub-method associated with the probability in the first claim method is increased to obtain an updated first claim method.
[0172] Accordingly, the updated first claim settlement method, the second claim settlement method, and the third claim settlement method are screened and / or merged to obtain a target claim settlement method. For instructions on obtaining the target claim settlement method, please refer to Figure 3 The description of step 105 is shown in the figure, and the description of step 105 is equivalently replaced and will not be repeated here.
[0173] The embodiments of the present application can respectively determine the claims settlement methods corresponding to the law, medical treatment and insurance, so as to obtain a target claims settlement method that integrates the law, medical treatment and insurance. Compared with the claims settlement method determined by the claims adjusters, it can combine the three different claims settlement methods of law, medical treatment and insurance to determine a more accurate target claims settlement method for personal injury incidents, and can improve the efficiency of determining the target claims settlement method for personal injury incidents. That is, the method for determining the personal injury claims settlement method based on artificial intelligence provided in the embodiments of the present application can improve the efficiency and accuracy of determining the claims settlement method.
[0174] In some embodiments, after determining the target claim settlement method, the target object's satisfaction information for the target claim settlement method can be obtained, and new injury data can be obtained. The satisfaction information can be obtained through questionnaires, manual return visits, feedback from target objects, etc. The new injury data can be injury data corresponding to different time points, and can be obtained through manual return visits, feedback from target objects, etc.
[0175] In order to further improve the accuracy of the claim settlement method, the target claim settlement method can be adjusted based on the satisfaction information and the new injury data to obtain a new target claim settlement method. In some embodiments, a first new claim settlement sub-method can be selected from the claim settlement sub-methods that do not belong to the target claim settlement method in the first claim settlement method, the second claim settlement method, and the third claim settlement method according to the satisfaction information, and the first new claim settlement sub-method is added to the target claim settlement method.
[0176] In some embodiments, according to the new injury information, a second new claim sub-method can be selected from the claim sub-methods that do not belong to the target claim sub-method, and the second new claim sub-method can replace the claim sub-methods that belong to the second claim sub-method in the target claim sub-method. Thus, a new target claim sub-method can be obtained by adding the first new claim sub-method and replacing it with the second new claim sub-method.
[0177] In some embodiments, the satisfaction information includes adjustment information of the claim settlement sub-method, so that the target claim settlement method can be adjusted based on the adjustment information, so that a new target claim settlement method can be obtained. In some embodiments, the injury data in the personal injury data can be replaced with new injury data to obtain new personal injury data, and the new personal injury data can be input into the legal model, the medical model, and the insurance model respectively to obtain a new target claim settlement method. In this way, the target claim settlement method can be updated and a more accurate claim settlement method can be obtained.
[0178] The following is a description of an exemplary structure of the implementation of the device 455 for determining the personal injury claim settlement method based on artificial intelligence provided in the embodiment of the present application as a software module. In some embodiments, for example Figure 2 As shown, the software modules stored in the device 455 for determining the personal injury claim settlement method based on artificial intelligence in the memory 450 may include:
[0179] The acquisition module 4551 is used to acquire the personal injury data of the personal injury event, wherein the personal injury data includes the accident data of the personal injury event, the object data of the personal injury event and the injury data of the personal injury event.
[0180] A determination module 4552, configured to input the personal injury data into a pre-trained legal model to obtain a first claim settlement method determined based on a historical legal event of the same type as the personal injury event;
[0181] The determination module 4552 is further used to input the personal injury data into a pre-trained medical model to obtain a second claim settlement method determined based on the object data and the target data, wherein the target data is obtained after screening the accident data and the injury data;
[0182] The determination module 4552 is further used to input the personal injury data into a pre-trained insurance model to obtain a third claim settlement method determined based on a historical insurance event whose similarity to the personal injury data is greater than a preset first similarity threshold;
[0183] The determination module 4552 is also used to screen and / or merge the first claim settlement method, the second claim settlement method and the third claim settlement method to obtain a target claim settlement method.
[0184] In some embodiments, the first claim settlement method, the second claim settlement method, and the third claim settlement method each include at least one type of claim settlement sub-method;
[0185] Determination module 4552 is also used to determine the processing method for the claim sub-method based on the object data, the type corresponding to the first claim method, the type corresponding to the second claim method and the type corresponding to the third claim method; based on the target processing method, the claim sub-methods included in the first claim method, the claim sub-methods included in the second claim method and the claim sub-methods included in the third claim method are screened and / or merged to obtain the target claim method.
[0186] In some embodiments, the determination module 4552 is also used to determine the calculation method corresponding to the object data based on the object data and the mapping relationship between the pre-set candidate object data and the candidate calculation method if there is an amount type among the types corresponding to the first claim method, the second claim method and the third claim method; if the number of types corresponding to the first claim method, the second claim method and the third claim method is multiple, then based on the object data, determine the grouping method corresponding to the object data from the pre-set candidate object data and the candidate grouping methods; and determine the calculation method and / or the grouping method as the processing method for the claim sub-method.
[0187] In some embodiments, the target processing method includes a calculation method and / or a grouping method; the determination module 4552 is further used to implement at least one of the following:
[0188] Based on the calculation method, determining a first target claim settlement sub-method from the claim settlement sub-methods included in the first claim settlement method, the claim settlement sub-methods included in the second claim settlement method, and the claim settlement sub-methods included in the third claim settlement method; based on the calculation method and the first target claim settlement sub-method, determining a target claim settlement method;
[0189] Based on the grouping method, the claim sub-methods included in the first claim method, the claim sub-methods included in the second claim method and the claim sub-methods included in the third claim method are grouped to obtain a target group; the target group includes a second target claim sub-method; for the second target claim sub-method in the target group, the second target claim sub-method is screened and / or merged to obtain a target claim method.
[0190] In some embodiments, the medical model includes a first medical sub-model, a second medical sub-model and a third medical sub-model. The determination module 4552 is further used to input the accident data and the injury data into the first medical sub-model to obtain target data after screening based on the matching degree of the accident data and the injury data;
[0191] Inputting the target data into the second medical sub-model to obtain a medical treatment method determined based on historical medical events whose similarity to the target data is greater than a second similarity threshold;
[0192] Inputting the medical treatment method and the object data into the third medical sub-model to obtain a medical claim amount predicted by combining the medical treatment method and the object data;
[0193] The medical treatment method and the medical claim amount are used as the second claim method.
[0194] In some embodiments, the object data includes decision information of whether the target object takes legal action against the personal injury incident, and the first claim settlement method includes a claim settlement sub-method associated with the probability of taking legal action against the personal injury incident;
[0195] Determination module 4552 is also used for, after the personal injury data is input into a pre-trained legal model to obtain a first claim settlement method determined based on a historical legal event of the same type as the personal injury event, if the decision information indicates that no legal action will be taken with respect to the personal injury event, then adding the decision information to the first claim settlement method and deleting the claim sub-method associated with the probability in the first claim settlement method to obtain an updated first claim settlement method; screening and / or merging the updated first claim settlement method, the second claim settlement method and the third claim settlement method to obtain a target claim settlement method.
[0196] In some embodiments, the determination module 4552 is also used to obtain the satisfaction information of the target object for the target claim settlement method after screening and / or merging the first claim settlement method, the second claim settlement method and the third claim settlement method to obtain the target claim settlement method, and obtain new injury data; based on the satisfaction information and the new injury data, adjust the target claim settlement method to obtain a new target claim settlement method.
[0197] The embodiment of the present application provides a computer program product, which includes a computer program or a computer executable instruction, and the computer program or the computer executable instruction is stored in a computer-readable storage medium. The processor of the electronic device reads the computer executable instruction from the computer-readable storage medium, and the processor executes the computer executable instruction, so that the electronic device executes the method for determining the personal injury claim settlement method based on artificial intelligence described in the embodiment of the present application.
[0198] The present application embodiment provides a computer-readable storage medium, in which computer executable instructions or computer programs are stored. When the computer executable instructions or computer programs are executed by a processor, the processor will be caused to execute the method for determining the personal injury claims method based on artificial intelligence provided by the present application embodiment, for example, Figure 3 A method for determining personal injury claims based on artificial intelligence is shown.
[0199] In some embodiments, the computer-readable storage medium may be a memory such as RAM, ROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; or may be various devices including one or any combination of the above memories.
[0200] In some embodiments, computer executable instructions may be in the form of a program, software, software module, script or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine or other unit suitable for use in a computing environment.
[0201] As an example, computer-executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file that stores other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files storing one or more modules, subroutines, or code portions).
[0202] As an example, computer executable instructions may be deployed to be executed on one electronic device, or on multiple electronic devices located at one site, or on multiple electronic devices distributed at multiple sites and interconnected by a communication network.
[0203] To sum up, the embodiments of the present application can respectively determine the claims methods corresponding to the law, medical treatment and insurance, so as to obtain the target claims method that integrates the law, medical treatment and insurance. Compared with the claims method determined by the claims adjusters, it can combine the three different claims methods of law, medical treatment and insurance to determine a more accurate target claims method for personal injury incidents, and can improve the efficiency of determining the target claims method for personal injury incidents. That is, the method for determining the personal injury claims method based on artificial intelligence provided by the embodiments of the present application can improve the efficiency and accuracy of determining the claims method.
[0204] The above is only an embodiment of the present application and is not intended to limit the protection scope of the present application. Any modifications, equivalent substitutions and improvements made within the spirit and scope of the present application are included in the protection scope of the present application.
Claims
1. A method for determining personal injury claims based on artificial intelligence, It is characterized in that The method comprises: Acquire personal injury data of a personal injury event, wherein the personal injury data includes accident data of the personal injury event, object data of the personal injury event, and injury condition data of the personal injury event; Inputting the personal injury data into a pre-trained legal model to obtain a first claim settlement method determined based on a historical legal event of the same type as the personal injury event; Inputting the personal injury data into a pre-trained medical model to obtain a second claim settlement method determined based on the object data and target data, wherein the target data is obtained after screening the accident data and the injury data; Inputting the personal injury data into a pre-trained insurance model to obtain a third claim settlement method determined based on a historical insurance event whose similarity to the personal injury data is greater than a preset first similarity threshold; The first claim settlement method, the second claim settlement method and the third claim settlement method are screened and / or combined to obtain a target claim settlement method.
2. The method according to claim 1, It is characterized in that The first claim settlement method, the second claim settlement method and the third claim settlement method each include at least one type of claim settlement sub-method; The screening and / or merging of the first claim settlement method, the second claim settlement method and the third claim settlement method to obtain a target claim settlement method includes: Determine a processing method for the claim settlement sub-method based on the object data, the type corresponding to the first claim settlement method, the type corresponding to the second claim settlement method, and the type corresponding to the third claim settlement method; Based on the target processing method, the claim sub-methods included in the first claim settlement method, the claim sub-methods included in the second claim settlement method, and the claim sub-methods included in the third claim settlement method are screened and / or merged to obtain the target claim settlement method.
3. The method according to claim 2, It is characterized in that The determining, based on the object data, the type corresponding to the first claim settlement method, the type corresponding to the second claim settlement method, and the type corresponding to the third claim settlement method, a processing method for the claim settlement sub-method includes: If there is an amount type among the types corresponding to the first claim settlement method, the types corresponding to the second claim settlement method, and the types corresponding to the third claim settlement method, determining the calculation method corresponding to the object data based on the object data and the pre-set mapping relationship between the candidate object data and the candidate calculation methods; If the number of types corresponding to the first claim settlement method, the number of types corresponding to the second claim settlement method, and the number of types corresponding to the third claim settlement method are all multiple, determining, based on the object data, a grouping method corresponding to the object data from pre-set candidate object data and candidate grouping methods; The calculation method and / or the grouping method are determined as a processing method for the claim settlement sub-method.
4. The method according to claim 2, It is characterized in that The target processing method includes a calculation method and / or a grouping method; Based on the target processing method, the claim settlement sub-methods included in the first claim settlement method, the claim settlement sub-methods included in the second claim settlement method, and the claim settlement sub-methods included in the third claim settlement method are screened and / or merged to obtain a target claim settlement method corresponding to at least one of the following: Based on the calculation method, determining a first target claim settlement sub-method from the claim settlement sub-methods included in the first claim settlement method, the claim settlement sub-methods included in the second claim settlement method, and the claim settlement sub-methods included in the third claim settlement method; based on the calculation method and the first target claim settlement sub-method, determining a target claim settlement method; Based on the grouping method, the claim sub-methods included in the first claim method, the claim sub-methods included in the second claim method and the claim sub-methods included in the third claim method are grouped to obtain a target group; the target group includes a second target claim sub-method; for the second target claim sub-method in the target group, the second target claim sub-method is screened and / or merged to obtain a target claim method.
5. The method according to claim 1, It is characterized in that The medical model includes a first medical sub-model, a second medical sub-model and a third medical sub-model; The step of inputting the personal injury data into a pre-trained medical model to obtain a second claim settlement method determined based on the object data and the target data includes: Inputting the accident data and the injury data into the first medical sub-model to obtain target data after screening based on the matching degree between the accident data and the injury data; Inputting the target data into the second medical sub-model to obtain a medical treatment method determined based on historical medical events whose similarity to the target data is greater than a second similarity threshold; Inputting the medical treatment method and the object data into the third medical sub-model to obtain a medical claim amount predicted by combining the medical treatment method and the object data; The medical treatment method and the medical claim amount are used as the second claim method.
6. The method according to any one of claims 1 to 5, It is characterized in that The object data includes decision information of whether the target object takes legal action for the personal injury incident, and the first claim settlement method includes a claim settlement sub-method associated with the probability of taking legal action for the personal injury incident; After inputting the personal injury data into a pre-trained legal model to obtain a first claim settlement method determined based on a historical legal event of the same type as the personal injury event, the method further includes: If the decision information indicates that no legal action will be taken with respect to the personal injury incident, the decision information is added to the first claim settlement method, and the claim settlement sub-method associated with the probability in the first claim settlement method is deleted to obtain an updated first claim settlement method; The screening and / or merging of the first claim settlement method, the second claim settlement method and the third claim settlement method to obtain a target claim settlement method includes: The updated first claim settlement method, the second claim settlement method and the third claim settlement method are screened and / or combined to obtain a target claim settlement method.
7. The method according to any one of claims 1 to 5, It is characterized in that After screening and / or merging the first claim settlement method, the second claim settlement method, and the third claim settlement method to obtain a target claim settlement method, the method further includes: Obtaining target object's satisfaction information regarding the target claim settlement method, and obtaining new injury data; Based on the satisfaction information and the new injury data, the target claim settlement method is adjusted to obtain a new target claim settlement method.
8. A device for determining the compensation method for personal injury based on artificial intelligence, It is characterized in that The device comprises: An acquisition module, used to acquire the personal injury data of the personal injury event, wherein the personal injury data includes the accident data of the personal injury event, the object data of the personal injury event and the injury data of the personal injury event; A determination module, configured to input the personal injury data into a pre-trained legal model to obtain a first claim settlement method determined based on a historical legal event of the same type as the personal injury event; The determination module is further used to input the personal injury data into a pre-trained medical model to obtain a second claim settlement method determined based on the object data and the target data, wherein the target data is obtained after screening the accident data and the injury data; The determination module is further used to input the personal injury data into a pre-trained insurance model to obtain a third claim settlement method determined based on a historical insurance event whose similarity to the personal injury data is greater than a preset first similarity threshold; The determination module is also used to screen and / or merge the first claim settlement method, the second claim settlement method and the third claim settlement method to obtain a target claim settlement method.
9. An electronic device, It is characterized in that The electronic device comprises: Memory for storing computer executable instructions or computer programs; A processor is used to implement the method for determining the personal injury claims settlement method based on artificial intelligence as described in any one of claims 1 to 7 when executing the computer executable instructions or computer programs stored in the memory.
10. A computer-readable storage medium storing computer-executable instructions or a computer program, It is characterized in that When the computer executable instructions or computer program are executed by the processor, the method for determining the personal injury claims settlement method based on artificial intelligence as described in any one of claims 1 to 7 is implemented.
