Multi-modal medical insurance data processing method and device

Abstract

The invention discloses a multi-modal medical insurance data processing method and device. The method comprises the steps of collecting original data; determining an application scene of the original data, wherein the application scene comprises a medical insurance basic portrait scene, a medical insurance service process portrait scene and a medical insurance service efficiency portrait scene; classifying the original data according to the application scene of the original data; determining an analysis model and a data structure according to the application scene; carrying out deep mining on basic information and insurance information of insured persons based on the analysis model; and reprocessing the classification result according to a data structure, wherein the reprocessing comprises selection of effective fields and dimension reduction processing of partial data fields. According to the method, required personalized services can be provided for the insured people more accurately, a scientific research and exploration tool based on machine learning can be provided for employees of medical institutions, and scientific information support can be provided for enterprises for production, circulation, sales and the like of medical products.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and particularly relates to a method and device for processing multi-modal medical insurance data. Background Art

[0002] Traditional analysis of the behaviors of medical insurance participants mainly uses statistical data for group large number analysis and prediction, without conducting personalized analysis and demand prediction on their behaviors, thus restricting the accuracy and personalization of medical security services. Summary of the Invention

[0003] An object of the present invention is to provide a method and device for processing multi-modal medical insurance data, which can solve the technical problem in the prior art that there is no personalized analysis and demand prediction for the behaviors of medical insurance participants.

[0004] According to a first aspect of the present invention, there is provided a method for processing multi-modal medical insurance data, including:

[0005] Collecting original data, where the original data includes medical insurance data and external medical insurance data;

[0006] Determining the application scenarios of the original data, where the application scenarios include medical insurance basic portrait scenarios, medical insurance service process portrait scenarios, and medical insurance service efficiency portrait scenarios;

[0007] Classifying the original data according to the application scenarios of the original data, evaluating the classification results to obtain an evaluation result, and the evaluation result is used to guide the collection of the original data;

[0008] Determining an analysis model and a data structure according to the application scenarios;

[0009] Deeply mining the basic information and insurance utilization information of the insured based on the analysis model;

[0010] Re-processing the classification results according to the data structure, including selecting effective fields and performing dimensionality reduction processing on some data fields.

[0011] Optionally, when the application scenario is a medical insurance basic portrait scenario, the classifying the original data according to the application scenarios of the original data includes:

[0012] In the case where the application scenario is a medical insurance basic portrait scenario, classifying the original data according to the basic biological characteristics and social characteristics of the insured, the basic health characteristics of the insured, and the basic characteristics of medical institutions;

[0013] The basic biological characteristics of the insured include age, gender, and ethnicity;

[0014] The social characteristics include personal annual income, family annual income, work unit, previous insurance types, and insurance duration;

[0015] The basic health characteristics of the insured include outpatient records, inpatient diagnosis records, previous medical treatment experiences, and previous medication experiences;

[0016] The basic characteristics of the medical institutions include the levels of medical institutions, departments in operation, and the usage of medical insurance funds.

[0017] Optionally, the application scenario is the scenario of the portrait of the medical insurance service process. Classifying the original data according to the application scenario of the original data includes:

[0018] In the case where the application scenario is the medical insurance service process scenario, classifying the original data according to the medical insurance handling service process data and the medical treatment process data;

[0019] The medical insurance handling service process data includes the data of handling various medical insurance service items;

[0020] The medical treatment process data includes the examination conditions, medication conditions, surgical conditions, and nursing conditions.

[0021] Optionally, the application scenario is the scenario of the portrait of the medical insurance service efficiency. Classifying the original data according to the application scenario of the original data includes:

[0022] In the case where the application scenario is the scenario of the portrait of the medical insurance service efficiency, classifying the original data according to the medical treatment effect data;

[0023] The medical treatment effect data includes medical treatment costs, medical insurance reimbursement costs, discharge methods, medical treatment effects, disease prognoses, outcomes, the corresponding portrait data of medical insurance service demanders after receiving services, and the corresponding portrait data of medical insurance service providers after providing services.

[0024] Optionally, the method further includes:

[0025] Adjusting the parameters of the analysis model according to the actual scenario requirements and data quality;

[0026] Selecting the part of the data with low dispersion degree according to the calculation results of the analysis model to form a standard database, and the standard database is used for decision-making assistance in restricting behavior standards;

[0027] Select some data with too high dispersion degree according to the calculation results of the analysis model to form a discrete database, which is used to correct the calculation bias generated by the standard database, judge abnormal behaviors, classify abnormal behaviors into super-optimal behaviors and super-inferior behaviors according to different dimensions of efficiency, and the super-inferior behaviors are used for early warning of abnormal data. After the number of super-optimal behaviors exceeds a predetermined threshold, the corresponding standard database is replaced by the super-optimal behaviors.

[0028] Optionally, the analysis model is XGBoost, and the in-depth mining of the basic information and insurance utilization information of the insured includes:

[0029] Form a prediction model by combining multiple decision trees;

[0030] Evaluate the prediction effect of the model by minimizing a differentiable multi-class cross-entropy;

[0031] Add a new decision tree in each iteration to correct the prediction error of the previous decision tree;

[0032] Introduce a regularization term to prevent overfitting;

[0033] Extract and model through feature engineering to predict the subsequent behaviors and self-health risks of the insured.

[0034] Optionally, the method further includes:

[0035] Predict the willingness of the insured to continue participating in insurance, health management, and insurance utilization needs by profiling the basic characteristics of the insured, and perform abnormal data screening;

[0036] Combined with the basic characteristic portrait data of the insured, recommend medical insurance services, recommend treatment paths, evaluate treatment effects, predict costs, predict hospital stay lengths, and give early warnings of abnormal insurance utilization information by profiling the treatment path and treatment effect.

[0037] According to the second aspect of the present invention, there is provided an apparatus for applying the processing method of a multi-modal medical insurance data according to the first aspect of the present invention, including:

[0038] A data acquisition module for acquiring original data, where the original data includes medical insurance data and external medical insurance data;

[0039] A data processing module for determining the application scenarios of the original data, where the application scenarios include medical insurance basic portrait scenarios, medical insurance service process portrait scenarios, and medical insurance service efficiency portrait scenarios, classify the original data according to the application scenarios of the original data, evaluate the classification results to obtain an evaluation result, and the evaluation result is used to guide the acquisition of the original data;

[0040] A data modeling and analysis module, which is used to determine an analysis model and a data structure according to the application scenario, deeply mine the basic information and insurance usage information of the insured based on the analysis model, and reprocess the classification result according to the data structure, including the selection of valid fields and the dimensionality reduction processing of some data fields.

[0041] According to the third aspect of the present invention, there is provided an electronic device, including a processor and a memory, where the memory stores a program or instruction that can run on the processor, and when the program or instruction is executed by the processor, the steps of a method for processing multi-modal medical insurance data as described in the first aspect of the present invention are implemented.

[0042] According to the fourth aspect of the present invention, there is provided a readable storage medium, where a program or instruction is stored on the readable storage medium, and when the program or instruction is executed by a processor, the steps of a method for processing multi-modal medical insurance data as described in the first aspect of the present invention are implemented.

[0043] The present invention has made multi-modal improvements to the existing medical insurance data structure to make it more adaptable to the data requirements of various machine learning algorithms. Through data empowerment, the present invention explores more medical insurance service scenarios, saves the service costs and decision-making costs of medical insurance departments and medical institutions, and at the same time provides a decision-making basis. The present invention also introduces machine learning algorithms to achieve precise analysis of medical insurance and related data.

[0044] The present invention can provide more precise personalized services required by the insured, can provide scientific research and exploration tools based on machine learning for medical institution practitioners, and can provide scientific information support for enterprises in the production, circulation, and sales of medical products. Description of the Drawings

[0045] Figure 1 is a flowchart of a method for processing multi-modal medical insurance data in an embodiment of the present invention.

[0046] Figure 2 is a flowchart of a method for processing multi-modal medical insurance data in another embodiment of the present invention. Detailed Embodiments

[0047] Now, various exemplary embodiments of the present invention will be described in detail with reference to the drawings. It should be noted that: unless otherwise specifically stated, the relative arrangements, numerical expressions, and numerical values of the components and steps described in these embodiments do not limit the scope of the present invention.

[0048] The following description of at least one exemplary embodiment is actually merely illustrative and in no way limits the present invention and its application or use.

[0049] Known technologies, methods, and devices for those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, such technologies, methods, and devices shall be regarded as part of the specification. In all examples shown and discussed herein, any specific values should be construed as merely exemplary and not as a limitation. Thus, other examples of the exemplary embodiments may have different values.

[0050] It should be noted that like reference numerals and letters denote like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.

[0051] In the description and claims of the present invention, features related to the terms "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, unless otherwise specified, "a plurality of" means two or more. In addition, "and / or" in the specification and claims means at least one of the connected objects, and the character " / " generally means an "or" relationship between the related objects before and after.

[0052] This embodiment provides a multi-modal medical insurance data personalized intelligent prediction and analysis method, including steps 1100-1500.

[0053] Step 1100: Collect raw data, where the raw data includes medical insurance data and external medical insurance data.

[0054] The medical insurance data includes the basic information of insured persons, medical insurance settlement data, etc. The medical insurance data may also include other data, which is selected according to actual needs.

[0055] The external medical insurance data includes hospital His data, health examination data, etc. The external medical insurance data may also include other data, which is selected according to actual needs.

[0056] Step 1200: Determine the application scenarios of the raw data, where the application scenarios include the medical insurance basic portrait scenario, the medical insurance service process portrait scenario, and the medical insurance service efficiency portrait scenario.

[0057] As Figure 1 and Figure 2 shown, through data empowerment, the application scenarios are determined. The medical insurance basic portrait scenario includes portraits of medical insurance service demanders (insured persons, insured units, insured groups with spatio-temporal consistency, etc.), and portraits of medical insurance service providers (medical insurance side, fixed-point medical and pharmaceutical institutions side, drug and consumable supply side, etc.). The medical insurance service demanders in the medical insurance basic portrait scenario include insured persons, insured units, insured groups with spatio-temporal consistency, etc. For a certain employer, the employees of the employer are insured persons, the employer is an insured unit, and the medical insurance of all employees of the employer is uniformly paid by the employer.

[0058] The scenarios of the medical insurance service process portrait include the service selection portrait of medical insurance service demanders and the service supply portrait of medical insurance service providers.

[0059] The scenarios of the medical insurance service efficiency portrait include the portraits of medical insurance service demanders (insured persons, insured units, etc.) and the portraits of medical insurance service providers (medical insurance side, designated medical and pharmaceutical institutions side, drug and consumable supply side, etc.).

[0060] Step 1300: Classify the original data according to the application scenario of the original data, evaluate the classification result to obtain an evaluation result, and the evaluation result is used to guide the collection of the original data.

[0061] Since the medical insurance data contains too many types of data, and the types of data required for different scenarios are also different, it is necessary to classify the original data according to the application scenario to meet the requirements of the application scenario.

[0062] In one implementation, when the application scenario is the medical insurance basic portrait scenario, the classifying the original data according to the application scenario of the original data includes:

[0063] When the application scenario is the medical insurance basic portrait scenario, classify the original data according to the basic biological characteristics and social characteristics of the insured person, the basic health characteristics of the insured person, and the basic characteristics of the medical and pharmaceutical institutions;

[0064] The basic biological characteristics of the insured person include age, gender, and ethnicity;

[0065] The social characteristics include personal annual income, family annual income, work unit, previous insurance types, and insurance time;

[0066] The basic health characteristics of the insured person include outpatient records, inpatient diagnosis records, previous medical treatment experiences, and previous medication experiences;

[0067] The basic characteristics of the medical and pharmaceutical institutions include the level of the medical and pharmaceutical institutions, the departments offered, and the use of medical insurance funds.

[0068] In one implementation, when the application scenario is the medical insurance service process portrait scenario, the classifying the original data according to the application scenario of the original data includes:

[0069] When the application scenario is the medical insurance service process scenario, classify the original data according to the medical insurance handling service process data and the diagnosis and treatment process data;

[0070] The medical insurance handling service process data includes various medical insurance service item handling data;

[0071] The medical treatment process data includes the examination situation, medication situation, surgical situation, and nursing situation.

[0072] The medical insurance service process refers to the process of the insured using medical insurance, including the insured going to the hospital to use medical insurance for diagnosis and treatment, and the insured going to the pharmacy to use medical insurance to purchase drugs. During the medical insurance service process, the insured needs to go through various medical insurance service items, and in this case, various medical insurance service item handling data is generated.

[0073] During the process of the insured receiving medical treatment in the hospital, medical treatment process data will be generated, such as various examinations and medications for the insured. Depending on the actual condition, surgery, hospitalization, nursing, etc. may also be required.

[0074] In this embodiment, when the application scenario is the medical insurance service efficiency portrait scenario, the classification of the original data according to the application scenario of the original data includes:

[0075] When the application scenario is the medical insurance service efficiency portrait scenario, classify the original data according to the medical treatment effect data;

[0076] The medical treatment effect data includes medical treatment expenses, medical insurance reimbursement expenses, discharge methods, medical treatment effects, disease prognosis, outcome, corresponding portrait data after the medical insurance service demanders receive services, and corresponding portrait data after the medical insurance service providers provide services.

[0077] The medical insurance service efficiency portrait scenario mainly faces the situation after the insured receives diagnosis and treatment in the hospital. The expenses of the insured receiving diagnosis and treatment in the hospital will be settled through medical insurance when the insured leaves the hospital. At this time, a series of medical treatment effect data such as medical treatment expenses, medical insurance reimbursement expenses, discharge methods, and medical treatment effects will be generated.

[0078] Step 1400: Determine the analysis model and data structure according to the application scenario.

[0079] Multiple different models can be selected for the analysis model according to needs, such as XGBoost, neural network, logistic regression, etc. The data structure is the data structure required for this application scenario.

[0080] Step 1500: Deeply mine the basic information and insurance usage information of the insured based on the analysis model.

[0081] Step 1600: Determine the data structure, and reprocess the classified data according to the data structure requirements, including the selection of valid fields and the dimensionality reduction processing of some data fields.

[0082] When making service selection recommendations or service effect judgments that require relevant portraits of service demanders, the relevant portrait details can be dimensionally reduced and compressed in a statistical manner, which can save a large amount of computing power.

[0083] In this embodiment, the method further includes adjusting the parameters of the analysis model according to actual scenario requirements and data quality;

[0084] Select a part of the data with low dispersion degree according to the calculation result of the analysis model to form a standard database;

[0085] Select a part of the data with too high dispersion degree according to the calculation result of the analysis model to form a discrete database.

[0086] In the analysis process, first, adjust the parameters according to actual scenario requirements and data quality to obtain higher prediction accuracy and calculation efficiency. Second, according to the calculation results, select a part of the data series with low dispersion degree to form a standard database. The standard database can further save future computing power and can also be used for decision-making assistance in restricting behavior standards, etc. Select a part of the data columns with too high dispersion degree to form a discrete database. The discrete database can be used to enrich the diversity of behavior data, correct the calculation bias generated by the standard database, and judge abnormal behaviors (which can be divided into super-optimal and super-inferior according to different dimensions of efficiency). The data between the standard library and the discrete library can be converted into each other according to actual needs (super-inferior behaviors are used as early warnings for abnormal data, and super-optimal behaviors can replace the corresponding standard library data after accumulating enough).

[0087] In this embodiment, the analysis model is XGBoost. The in-depth mining of the basic information and insurance utilization information of the insured based on the analysis model includes: forming a prediction model by combining multiple decision trees; evaluating the prediction effect of the model by minimizing a differentiable multi-class cross-entropy; adding a new decision tree in each iteration to correct the prediction error of the previous decision tree; introducing a regularization term to prevent overfitting and ensure the generalization ability of the model on complex data sets; predicting the subsequent behaviors and self-health risks of the insured through feature engineering extraction and modeling.

[0088] In this embodiment, the method further includes: predicting the willingness of the insured to continue insurance, health management, and insurance utilization needs by portraying the basic characteristics of the insured, and screening abnormal data;

[0089] Combined with the portrait data of the basic characteristics of the insured, through portraying the diagnosis and treatment path and the diagnosis and treatment effect, conduct medical insurance service recommendations, diagnosis and treatment path recommendations, curative effect evaluation, cost prediction, hospitalization duration prediction, and early warning of abnormal insurance utilization information.

[0090] The basic characteristics of the insured include basic biological characteristics and social characteristics, such as age, gender, personal annual income, work unit, previous insurance types, etc. Predict the willingness of the insured to continue participating in insurance based on their basic characteristics. For example, insured persons with previous insurance type of employee medical insurance usually have a stronger willingness to continue participating in insurance. At the same time, it is also possible to further judge by combining the income of the insured and the situation of their work unit. For example, if there are relatively high risks in the work unit, it indicates that the labor relationship between the insured and the work unit may be terminated, which may lead to the insured discontinuing the payment of medical insurance.

[0091] By depicting the diagnosis and treatment path and the diagnosis and treatment effect, reflect the historical diagnosis and treatment data of the insured, such as the hospital selected by the insured, the diagnosis and treatment effect in that hospital, the diagnosis and treatment expenses, etc. Combine the basic characteristic portrait data of the insured, such as the personal annual income of the insured, and recommend suitable medical institutions. It is also possible to conduct curative effect evaluation, cost prediction, hospitalization duration prediction, etc., to provide personalized medical services for the insured and select the most suitable medical institution.

[0092] This embodiment introduces a device that applies the processing method of a multi-modal medical insurance data described in any embodiment of the present invention, including:

[0093] A data acquisition module, used to acquire raw data, where the raw data includes medical insurance data and medical insurance external data;

[0094] A data processing module, used to determine the application scenarios of the raw data, where the application scenarios include medical insurance basic portrait scenarios, medical insurance service process portrait scenarios, and medical insurance service efficiency portrait scenarios. Classify the raw data according to the application scenarios of the raw data, evaluate the classification results, and obtain evaluation results, where the evaluation results are used to guide the acquisition of raw data;

[0095] A data modeling and analysis module, used to determine an analysis model and a data structure according to the application scenario, deeply mine the basic information and insurance usage information of the insured based on the analysis model, and reprocess the classification results according to the data structure, including the selection of valid fields and the dimensionality reduction processing of some data fields.

[0096] This embodiment introduces an electronic device, including a processor and a memory. The memory stores a program or instruction that can run on the processor. When the program or instruction is executed by the processor, it implements the steps of a processing method of a multi-modal medical insurance data described in any embodiment of the present invention.

[0097] This embodiment introduces a readable storage medium. A program or instruction is stored on the readable storage medium. When the program or instruction is executed by a processor, it implements the steps of a processing method of a multi-modal medical insurance data described in any embodiment of the present invention.

[0098] Although some specific embodiments of the present invention have been described in detail by way of examples, those skilled in the art should understand that the above examples are for illustrative purposes only and not for limiting the scope of the present invention. Those skilled in the art should understand that the above embodiments can be modified without departing from the scope and spirit of the present invention. The scope of the present invention is defined by the appended claims.

[0099] Those of ordinary skill in the art can realize that the modules and algorithm steps described in connection with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0100] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the devices and equipment described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0101] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or modules can be in electrical, mechanical or other forms.

[0102] The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they can be located in one place or distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the embodiments of the present invention.

[0103] In addition, the various functional modules in the embodiments of the present invention can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0104] When the above-mentioned function is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical discs.

[0105] The above description is only a preferred embodiment of the present application and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the present application is not limited to the technical solution formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the technical solution formed by mutually replacing the above features with the (but not limited to) technical features having similar functions disclosed in the present application.

[0106] It should be understood that the magnitude of the sequence numbers of the steps in the inventive content and embodiments of the present invention does not absolutely mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention. For the purpose of illustration and description, the foregoing description of the implementation of the present disclosure has been given. The foregoing description is not exhaustive and is not intended to limit the present disclosure to the exact form disclosed. According to the above teachings, various deformations and modifications may exist, or various deformations and modifications may be obtained from the practice of the present disclosure. These embodiments are selected and described to illustrate the principles of the present disclosure and its practical applications, so that those skilled in the art can utilize the present disclosure in various embodiments and various modifications suitable for the specific purpose of the conceived idea.

Claims

1. A method for processing multimodal medical insurance data, characterized in that: include: Collecting original data, wherein the original data includes medical insurance data and medical insurance external data; Determine the application scenarios of the original data, wherein the application scenarios include basic medical insurance portrait scenarios, medical insurance service process portrait scenarios, and medical insurance service effectiveness portrait scenarios; Classifying the original data according to the application scenarios of the original data, evaluating the classification results to obtain evaluation results, and the evaluation results are used to guide the collection of original data; Determine the analysis model and data structure according to the application scenario; Based on the analysis model, the basic information and insurance information of the insured are deeply mined; The classification result is reprocessed according to the data structure, including the selection of valid fields and the dimensionality reduction of some data fields.

2. The method according to claim 1, characterized in that The application scenario is a basic medical insurance portrait scenario, and the classification of the original data according to the application scenario of the original data includes: When the application scenario is a basic medical insurance profiling scenario, the raw data is classified according to the basic biological and social characteristics of the insured person, the basic health characteristics of the insured person, and the basic characteristics of the medical institution; The basic biological characteristics of the insured person include age, gender, and ethnicity; The social characteristics include personal annual income, family annual income, work unit, past insurance type, and insurance period; The basic health characteristics of the insured person include outpatient records, hospitalization diagnosis records, past medical treatment experience, and past medication experience; The basic characteristics of the medical institution include the grade of the medical institution, the departments in charge, and the use of medical insurance funds.

3. The method according to claim 1, characterized in that The application scenario is a medical insurance service process portrait scenario, and the classification of the original data according to the application scenario of the original data includes: In the case where the application scenario is a medical insurance service process scenario, the original data is classified according to the medical insurance handling service process data and the diagnosis and treatment process data; The medical insurance service processing data includes various medical insurance service item processing data; The diagnosis and treatment process data includes examination results, medication results, surgery results, and nursing results.

4. The method according to claim 1, characterized in that: The application scenario is a medical insurance service effectiveness portrait scenario, and the classification of the original data according to the application scenario of the original data includes: In the case where the application scenario is a medical insurance service effectiveness profiling scenario, the raw data is classified according to the diagnosis and treatment effect data; The diagnosis and treatment effect data include diagnosis and treatment expenses, medical insurance reimbursement expenses, discharge methods, diagnosis and treatment effects, disease prognosis, outcomes, corresponding portrait data of medical insurance service demanders after receiving services, and corresponding portrait data of medical insurance service providers after providing services.

5. The method according to claim 1, characterized in that The method further comprises: Adjusting parameters of the analysis model according to actual scenario requirements and data quality; According to the calculation results of the analysis model, some data with low discreteness are screened out to form a standard database, and the standard database is used for decision-making assistance of constraining behavior standards; According to the calculation results of the analysis model, some data with too high discreteness are screened out to form a discrete database. The discrete database is used to correct the calculation bias generated by the standard database, judge abnormal behaviors, and divide abnormal behaviors into super-optimal behaviors and super-inferior behaviors according to different dimensions of performance. The super-inferior behaviors are used for abnormal data early warning. After the number of super-optimal behaviors exceeds a predetermined threshold, the super-optimal behaviors are used to replace the corresponding standard database.

6. The method according to claim 1, characterized in that The analysis model is XGBoost, and the in-depth mining of the basic information and insurance information of the insured person based on the analysis model includes: Form a prediction model by combining multiple decision trees; The prediction performance of the model is evaluated by minimizing a differentiable multi-class cross entropy. Add new decision trees in each iteration to correct the prediction errors of the previous decision trees; Introduce regularization terms to prevent overfitting; Through feature engineering extraction and modeling, the subsequent behavior of the insured and their own health risks are predicted.

7. The method according to claim 1, characterized in that The method further comprises: By profiling the basic characteristics of the insured, predicting the insured's willingness to continue insurance, health management and insurance needs, and screening abnormal data; Combined with the basic characteristic profile data of the insured person, through the profiling of the treatment pathway and the treatment effect, we make medical insurance service recommendations, treatment pathway recommendations, efficacy evaluation, cost forecasts, hospitalization duration forecasts, and abnormal insurance usage information warnings.

8. A device using the method for processing multimodal medical insurance data according to any one of claims 1 to 7, characterized in that: include: A data collection module, used to collect original data, wherein the original data includes medical insurance data and medical insurance external data; A data processing module is used to determine the application scenarios of the original data, wherein the application scenarios include medical insurance basic portrait scenarios, medical insurance service process portrait scenarios, and medical insurance service efficiency portrait scenarios, classify the original data according to the application scenarios of the original data, evaluate the classification results, and obtain evaluation results, which are used to guide the collection of original data; The data modeling and analysis module is used to determine the analysis model and data structure according to the application scenario, conduct in-depth mining of the basic information and insurance information of the insured person based on the analysis model, and reprocess the classification results according to the data structure, including the selection of valid fields and dimensionality reduction of some data fields.

9. An electronic device, characterized in that: It includes a processor and a memory, the memory stores programs or instructions that can be run on the processor, and when the program or instructions are executed by the processor, the steps of a method for processing multimodal medical insurance data as described in any one of claims 1 to 7 are implemented.

10. A readable storage medium, characterized in that: The readable storage medium stores programs or instructions, and when the programs or instructions are executed by the processor, the steps of a method for processing multimodal medical insurance data as described in any one of claims 1 to 7 are implemented.

Images