Method for processing information of medical insurance business

Abstract

The embodiment of the specification provides an information processing method of medical insurance business, wherein the information processing method of medical insurance business comprises the following steps: acquiring attribute distribution information of medical insurance business in association with a business attribute dimension, thereby creating a simulated user set and determining medical insurance business simulation information of a simulated user in the set; inputting the medical insurance business simulation information into a medical insurance business prediction model to obtain user guarantee information of the simulated user; generating global guarantee information according to the user guarantee information, and creating a medical insurance business update strategy according to the medical insurance business information and the global guarantee information. The simulated user is created through the attribute distribution information, and is applied to guarantee behavior prediction of the medical insurance business, so that the guarantee situation of the medical insurance business can be more reasonably predicted, and the creation of the business update strategy on this basis can enable the medical insurance business to provide medical insurance services for users on the premise of balancing income and expenditure, so that relevant factors that have an impact on the medical insurance business can be effectively controlled.

Description

Technical Field

[0001] The embodiments in this specification relate to the field of information processing technology, and in particular to information processing methods for medical insurance business. Background Technology

[0002] As a crucial component of social security services, medical insurance is a vital measure to safeguard users' personal safety. Forecasting the revenue and expenditure of medical insurance services enables timely adjustments to policies, preventing deficits and maintaining a balance between income and expenditure. Currently, most macro-level decisions regarding medical insurance operations are based on simple, traditional linear one-to-one relationships, lacking comprehensive consideration. Furthermore, they fail to account for qualitative data and factors lacking specific datasets in medical scenarios, resulting in low accuracy in revenue and expenditure forecasting and hindering precise policy adjustments. Therefore, an effective solution is urgently needed to address these issues. Summary of the Invention

[0003] In view of this, embodiments of this specification provide an information processing method for medical insurance business. One or more embodiments of this specification also relate to an information processing method, an information processing apparatus, a computing device, a computer-readable storage medium, and a computer program, to address the technical deficiencies existing in the prior art.

[0004] According to a first aspect of the embodiments of this specification, an information processing method for medical insurance services is provided, including:

[0005] Obtain attribute distribution information related to medical insurance business in the business attribute dimension;

[0006] A set of simulated users is created based on the attribute distribution information, and the simulated medical insurance business information of the simulated users in the set of simulated users is determined.

[0007] The simulated medical insurance business information is input into the medical insurance business prediction model for processing to obtain the user insurance information of the simulated users in the simulated user set.

[0008] Global protection information is generated based on the user protection information, and a medical insurance service update strategy is created according to the medical insurance service information corresponding to the medical insurance service and the global protection information.

[0009] According to a second aspect of the embodiments of this specification, an information processing method is provided, comprising:

[0010] Obtain the attribute distribution information associated with the protection business in the business attribute dimension;

[0011] A set of simulated users is created based on the attribute distribution information, and the service guarantee simulation information of the simulated users in the set of simulated users is determined.

[0012] The simulated information of the protection business is input into the protection business prediction model for processing to obtain the user protection information of the simulated users in the simulated user set.

[0013] Global protection information is generated based on the user protection information, and a protection service update strategy is created according to the protection service information corresponding to the protection service and the global protection information.

[0014] According to a third aspect of the embodiments of this specification, an information processing apparatus is provided, comprising:

[0015] The acquisition module is configured to acquire attribute distribution information associated with the business attribute dimension for the protection business;

[0016] The determination module is configured to create a set of simulated users based on the attribute distribution information, and determine the service guarantee simulation information of the simulated users in the set of simulated users;

[0017] The processing module is configured to input the security service simulation information into the security service prediction model for processing, and obtain the user security information of the simulated users in the simulated user set;

[0018] The creation module is configured to generate global protection information based on the user protection information, and to create a protection service update strategy according to the protection service information corresponding to the protection service and the global protection information.

[0019] According to a fourth aspect of the embodiments of this specification, a computing device is provided, comprising:

[0020] Memory and processor;

[0021] The memory is used to store computer-executable instructions, and the processor is used to implement the steps of any of the above-described information processing methods when executing the computer-executable instructions.

[0022] According to a fifth aspect of the embodiments of this specification, a computer-readable storage medium is provided that stores computer-executable instructions, which, when executed by a processor, implement the steps of the information processing method described above.

[0023] According to a sixth aspect of the embodiments of this specification, a computer program is provided, wherein when the computer program is executed in a computer, it causes the computer to perform the steps of the above-described information processing method.

[0024] This manual provides a method for processing information related to medical insurance operations. To improve the accuracy of adjustments to medical insurance operations, the method first obtains the attribute distribution information associated with medical insurance operations across the business attribute dimension. This information is then used to create a set of simulated users, and the simulated medical insurance information for each user within the set is determined. This simulated information is then input into a medical insurance prediction model to obtain the simulated user's coverage information. Global coverage information is generated based on this user coverage information, and a medical insurance update strategy is created according to both the medical insurance information and the global coverage information. By creating simulated users through attribute distribution information and applying this to predict medical insurance behavior, the method can more accurately predict the coverage status of medical insurance operations. Based on this, the creation of a business update strategy allows medical insurance operations to provide services to users while maintaining a balance between revenue and expenditure, thus effectively controlling factors that influence medical insurance operations. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating an information processing method for medical insurance services provided in one embodiment of this specification;

[0026] Figure 2-1 This is a schematic diagram of an information processing method provided in one embodiment of this specification;

[0027] Figure 2-2 This is a flowchart illustrating an information processing method provided in one embodiment of this specification;

[0028] Figure 3 This is a flowchart of model training in an information processing method provided in one embodiment of this specification.

[0029] Figure 4 This is a flowchart illustrating the processing procedure of an information processing method provided in one embodiment of this specification.

[0030] Figure 5 This is a schematic diagram of the structure of an information processing device provided in one embodiment of this specification;

[0031] Figure 6 This is a structural block diagram of a computing device provided in one embodiment of this specification. Detailed Implementation

[0032] Many specific details are set forth in the following description to provide a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.

[0033] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a,” “described,” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0034] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0035] First, the terms and concepts used in one or more embodiments of this specification will be explained.

[0036] ABM (Agent Based Model) is a computational model used to simulate the actions and interactions of autonomous intelligent agents (independent individuals or collective groups, such as organizations or teams), demonstrating their interactions within the overall system through relationships between agents. ABM integrates methods from game theory, complex systems, emergence, evolutionary computation, and stochastic theory. A typical ABM model includes: a certain number of "agents"; relationships between these agents; and a framework simulating the behavior and interactions of the agents. ABM clarifies the causal relationships of the simulated individuals or objects in time and space. An "agent" can represent a person, wild animal, vehicle, land parcel, or other discrete objects.

[0037] This specification provides an information processing method for medical insurance business. This specification also relates to an information processing method, an information processing device, a computing device, a computer-readable storage medium, and a computer program, which will be described in detail in the following embodiments.

[0038] Figure 1 A flowchart illustrating an information processing method for medical insurance services according to an embodiment of this specification is shown, specifically including the following steps:

[0039] Step S102: Obtain the attribute distribution information associated with medical insurance business in the business attribute dimension.

[0040] Step S104: Create a set of simulated users based on the attribute distribution information, and determine the simulated medical insurance business information of the simulated users in the set of simulated users;

[0041] Step S106: Input the medical insurance business simulation information into the medical insurance business prediction model for processing to obtain the user insurance information of the simulated users in the simulated user set;

[0042] Step S108: Generate global protection information based on the user protection information, and create a medical insurance service update strategy according to the medical insurance service information corresponding to the medical insurance service and the global protection information.

[0043] Medical insurance generally refers to basic medical insurance, a social insurance system established to compensate workers for economic losses caused by the risk of illness. Through contributions from employers and individuals, a medical insurance fund is established. When insured individuals incur medical expenses due to illness, the medical insurance institution provides them with certain financial compensation; it is abbreviated as medical insurance.

[0044] In practical applications, medical insurance funds, as a crucial component of users' medical security, can save them significant medical expenses. However, issues such as the waste of medical insurance funds and simplistic decision-making processes can impact users due to the complexity of real-world scenarios, leading to losses for stakeholders and even preventing truly needy patients from receiving adequate protection. Currently, most macro-level decisions regarding the operation of medical insurance funds are based on simple, traditional linear one-to-one relationships, lacking comprehensive consideration of multi-dimensional factors. Traditional machine learning modeling also fails to consider qualitative data and factors without data sets in medical scenarios. Traditional linear models cannot account for the effects of individual responses to external influences on the group, or the resulting non-linear relationships. Therefore, an effective solution is urgently needed to address these problems.

[0045] Therefore, by acquiring attribute distribution information related to medical insurance business in the business attribute dimension, a set of simulated users is created, and the simulated medical insurance business information of the simulated users in the set is determined. This simulated medical insurance business information is then input into a medical insurance business prediction model to obtain the user coverage information of the simulated users. Global coverage information is generated based on the user coverage information, and a medical insurance business update strategy is created according to the medical insurance business information and the global coverage information. This approach, which creates simulated users through attribute distribution information and applies it to predicting coverage behavior in medical insurance business, allows for more reasonable predictions of the coverage situation. Based on this, the creation of business update strategies enables medical insurance business to provide services to users while balancing revenue and expenditure, thus effectively controlling relevant factors affecting medical insurance business.

[0046] In scenarios where business operations are guaranteed, see [link / reference]. Figure 2-1 The schematic diagram illustrates the information processing method provided in this embodiment. To improve prediction accuracy and enable precise adjustments to support services, the method can, during the experimental preparation phase, first determine the inclusion criteria based on the experimental equipment, i.e., obtain the user set under the support service. Then, based on the user set, define screening conditions to determine the number of participants. Secondly, during the data preparation phase, considering the lack of uniform data standards, the participant data can be cleaned, integrated, and processed before undergoing data transformation and feature extraction to visualize the data distribution, thus obtaining data suitable for model training. Furthermore, to enhance data richness, data can be expanded through data mining. After obtaining standard data, the model preparation phase can then commence.

[0047] In the model preparation phase, the standard data can be split to obtain training data for model training and test data for model validation. The training data is then used to train the model, resulting in an intermediate model. The prediction accuracy of this intermediate model can then be verified using test data. If the model does not meet the requirements, training continues using the training data until a model that meets the training stopping criteria is obtained, resulting in the evaluation model, i.e., the business prediction model. In the simulation data phase, simulated population data related to the business can be generated by inputting variables from the business simulator. The relevant data corresponding to this simulated population data is then input into the model for processing, yielding prediction results for the current scenario. Finally, the prediction results are summarized to form visualized data. Based on this, business policy decisions are created, i.e., business update strategies are developed. This allows for precise adjustments to the business to ensure a balance between revenue and expenditure.

[0048] Figure 2-2 A flowchart of an information processing method according to an embodiment of this specification is shown, which specifically includes the following steps.

[0049] Step S202: Obtain the attribute distribution information associated with the business attribute dimension of the guarantee business.

[0050] Specifically, protection services refer to insurance businesses that provide protection services to users. These services establish a protection fund through individual contributions, allowing participating users to receive corresponding financial compensation after triggering the protection service. Examples include medical insurance. Furthermore, the protection fund's income and expenditure must be balanced to maintain the operation of the protection business and provide reasonable financial compensation to more users. Therefore, the balance of income and expenditure in protection services is fundamental to maintaining the lifespan of the protection business. Correspondingly, business attribute dimensions refer to the dimensions of information needed to predict the balance of income and expenditure in protection services. For example, in medical insurance, business attribute dimensions may include drug dimensions, diagnosis dimensions, treatment dimensions, years of medical insurance coverage, and gender dimensions. By integrating information from various dimensions, the relevant data of a user's medical insurance business can be clearly identified. Based on this, the relevant data of all users' medical insurance business can be coordinated, and then the balance of income and expenditure can be predicted. In other words, business attribute dimensions are the dimensions of information related to protection services.

[0051] Correspondingly, attribute distribution information specifically refers to the distribution information corresponding to the business attribute dimensions. Attribute distribution information can be used to simulate user sets. In other words, attribute distribution information can be understood as the attribute information of real users participating in insurance business in various business attribute dimensions, based on the real world. For example, in medical insurance business, if there is one patient with disease A out of 1000 people, the patient is 50 years old, male, weighs 60 kg, and uses medications A and B, it can be determined that this involves the age dimension (50 years), gender dimension (male), weight dimension (60 kg), and medication dimensions (A and B), and the probability of having disease A is one in a thousand. All of the above can be understood as the attribute distribution information corresponding to various attribute dimensions in medical insurance business. This information can be used to subsequently simulate users that are more similar to real-world scenarios, thereby improving the accuracy of insurance business revenue and expenditure prediction.

[0052] This embodiment uses medical insurance as an example to illustrate the information processing method. The information processing process in other business scenarios can be found in the same or similar descriptions in this embodiment, and will not be elaborated on further here.

[0053] Based on this, after obtaining the attribute distribution information associated with the protection business in the business attribute dimension, the attribute distribution of the protection business can be clearly defined. On this basis, users who are more similar to the real scenario can be simulated. Based on these users, the update strategy of the protection business can be determined, which can effectively improve the prediction accuracy.

[0054] Prior to this, considering the potentially complex business attribute dimensions involved in the insurance business, and that the attribute distribution information of some attribute dimensions might not affect the creation of the insurance business update strategy—for example, in the medical insurance business scenario, information on cross-regional medical insurance payments is also associated with medical insurance business, but this dimension's information does not affect the medical insurance revenue and expenditure balance, meaning that this dimension's information can be disregarded—it's possible to eliminate some dimensions that will not affect the insurance business before obtaining the attribute distribution information associated with the insurance business across business attribute dimensions. The relevant information can then be obtained from the associated dimensions for subsequent use. In this embodiment, the specific implementation method is as follows:

[0055] Obtain the initial business attribute dimension associated with the guaranteed business; process the initial business attribute dimension in response to the standard setting instruction to obtain the business attribute dimension.

[0056] Specifically, the initial business attribute dimension refers to each business attribute dimension among all business attribute dimensions of the associated guarantee business. Correspondingly, the standard setting instruction is an instruction submitted according to the current scenario requirements, used to filter out business attribute dimensions from all business attribute dimensions, and used to obtain information according to the selected dimensions in the future.

[0057] Based on this, before forecasting the balance of income and expenditure for the guarantee business, in order to reduce the impact of redundant dimension information, the initial business attribute dimensions associated with the guarantee business can be obtained first. Then, according to the needs of the current scenario, a standard setting instruction is submitted to remove the initial business attribute dimensions according to the standard setting instruction. Based on the removal result, the business attribute dimensions are determined, and step S204 is executed on this basis.

[0058] In other words, in order to accurately simulate real users and reduce the impact of redundant information during the simulation process, business attribute dimensions can be selected first based on business needs. Then, based on the business attribute dimensions, the attribute distribution information related to the business can be obtained. This information can be used to create simulated users and then perform break-even prediction on the business. This can effectively improve the accuracy of prediction and reduce the impact of redundant information.

[0059] In summary, by determining the business attribute dimensions in response to standard setting instructions, it is possible to avoid simulating information with low correlation to the business when simulating users, thereby reducing resource consumption and effectively improving information processing efficiency.

[0060] Step S204: Create a set of simulated users based on the attribute distribution information, and determine the service guarantee simulation information of the simulated users in the set of simulated users.

[0061] Specifically, after obtaining the attribute distribution information associated with each business attribute dimension, the attribute distribution information associated with each business attribute dimension can further characterize the distribution of real users in the assurance business. Based on this, a set of simulated users can be created according to the attribute distribution information to map the distribution of real users participating in the assurance business. Then, by determining the assurance business simulation information of the simulated users in the simulated user set, it is convenient to predict the expenditure of the assurance business by predicting and summarizing the assurance business simulation information, which is used to realize the creation of the assurance business update strategy.

[0062] Specifically, the simulated user set refers to a collection of simulated users, whose distribution is based on real-world data, thus more realistically mapping information about real-world users. Correspondingly, the simulated insurance business information refers to the information about insurance business associated with simulated users, which is also simulated based on real information. For example, in the context of medical insurance business, the simulated insurance business information of simulated users includes, but is not limited to, enrollment time, periodic payment amount, and medical insurance expense information. It should be noted that the simulated insurance business information of simulated users is closer to the real-world insurance business information of users participating in insurance business, and is used to realistically simulate actual users.

[0063] Furthermore, in the process of creating the simulated user set, considering that in actual application scenarios, the impact on the people participating in the protection business is not only affected by the service of the protection business, but also by factors such as the decision-making of the protection business, in order to create more realistic simulated users, the attribute distribution information associated with the protection business can be updated. In this embodiment, the specific implementation method is as follows:

[0064] Obtain business simulation parameters and decision simulation parameters associated with the protection service; update the attribute distribution information based on the business simulation parameters and the decision simulation parameters to obtain target attribute distribution information, and create the simulated user set according to the target attribute distribution information; create protection service simulation information for the simulated users in the simulated user set according to the target attribute distribution information.

[0065] Specifically, business simulation parameters refer to key simulation variables associated with the business assurance operations. These key simulation variables comprehensively represent important parameters of the business assurance operations, allowing adjustments to the distribution of business assurance services associated with simulated users when creating simulated users. Similarly, decision simulation parameters refer to important variables in the actual decision-making process related to the business assurance operations. These important variables in the actual decision-making process comprehensively represent parameters of the business assurance decisions, allowing adjustments to the demographic characteristics of simulated users associated with business assurance services when creating simulated users. Correspondingly, target attribute distribution information refers to the attribute distribution information obtained after updating the attribute distribution information associated with each business attribute dimension according to the business simulation parameters and decision simulation parameters.

[0066] Based on this, after obtaining the attribute distribution information of the associated business attribute dimensions, in order to improve the realism of the created simulated users, business simulation parameters and decision simulation parameters of the associated support business can be obtained. This allows for the identification of important variables in the actual decision-making of the support business based on the decision simulation parameters, and key simulated variables in the support business based on the business simulation parameters. Secondly, by updating the attribute distribution information using the business simulation parameters and decision simulation parameters, the target attribute distribution information affected by decisions and business can be obtained, which more closely matches the real business scenario. At this point, a set of simulated users can be constructed based on the target attribute distribution information, ensuring that the simulated users in the set are closer to real users. Finally, support business simulation information for each simulated user can be created according to the target attribute distribution information, which can then be used to synthesize the support business simulation information of all simulated users to predict the balance of payments for the support business.

[0067] Specifically, creating the protection business simulation information for each simulated user based on the target attribute distribution information means that the distribution of simulated users in each business attribute dimension can be determined based on the target attribute information. Each attribute dimension is the dimension to which the information related to the protection business belongs. Therefore, based on the target attribute distribution information, the information that the simulated user needs to be assigned in each business attribute dimension can be clearly defined. Based on this, the protection business simulation information for each simulated user can be created, which will facilitate subsequent revenue and expenditure prediction of the protection business.

[0068] In practical implementation, to ensure that simulated users more closely resemble real users and that the created assurance business simulation information is more realistic, an ABM (Agent Base Model) can be used to create the simulated users and their associated assurance business model information. In other words, based on the key simulation variables and important variables in the actual assurance business decision-making process when modeling the ABM agent base model associated with assurance business, a dynamic model is derived from the business relationships between the variables in the modeling and the assurance business. This dynamic model, or ABM agent base model associated with assurance business, is used to simulate and generate the set of simulated users associated with assurance business. The composition of the simulated user set and the assurance business simulation information are based on the distribution in the real world, and are simulated using a dynamic model combined with business simulation parameters and decision simulation parameters to ensure the realism of the simulated users.

[0069] For example, in the context of medical insurance operations, to accurately balance medical insurance revenue and expenditure, a simulated user set can be created to simulate the real-world population participating in medical insurance, used for predicting the balance of medical insurance revenue and expenditure. Based on this, according to important variables in actual medical insurance decision-making (demographic factors, aging factors, birth rate, population migration, macroeconomic factors, environmental variables, medical insurance policy adjustments, etc.) and key simulation variables in the modeling (user attributes, diseases, drugs, medical visits, etc.), a dynamic model relating these variables to diseases and medical insurance fund financing is derived, and combined with the natural annual population growth to simulate users participating in medical insurance operations. It is important to note that when simulating users participating in medical insurance operations, it is also necessary to refer to authoritative publications, academic journals, and research on converting the dynamic model into population distribution characteristics to simulate users that more closely resemble real users. Based on the simulation results, the simulated user set can be determined to include user U1 {years L1; male; monthly medical insurance payment R1 yuan; enrollment time T1; suffering from disease A1; using medication B1…}, user U2 {years L2; ​​male; monthly medical insurance payment R2 yuan; enrollment time T2; no disease; no medication…}, and user Un {years Ln; female; monthly medical insurance payment Rn yuan; enrollment time Tn; suffering from disease An; using medication Bn…} (where n is a positive integer greater than 0). That is, the simulated user set contains n simulated users participating in medical insurance, as well as relevant information about medical insurance business associated with each simulated user, to facilitate subsequent prediction of medical insurance revenue and expenditure balance.

[0070] In summary, by combining business simulation parameters and decision simulation parameters to create a set of simulated users and simulated security business information, it is possible to make the simulated users and their corresponding security business simulation information more closely resemble actual application scenarios during the simulation phase. This allows for a macro-level reflection of changes in security business under actual scenarios, facilitating accurate determination of the balance of payments for security business in the future.

[0071] Step S206: Input the security service simulation information into the security service prediction model for processing to obtain the user security information of the simulated users in the simulated user set.

[0072] Specifically, after obtaining the protection service simulation information for each simulated user, this information further reflects the user's participation in the protection service. As a service providing protection to users, the protection service maintains its operational cycle by ensuring a balance between revenue and expenditure. This ensures the provision of protection services to users with protection needs and manages fees paid by users without protection needs. Therefore, it is necessary to first determine the user protection information for each simulated user in the simulated user set. This facilitates the subsequent integration of user protection information from all simulated users, determining the global protection information related to the protection service, i.e., the protection service's expenditure information. Combining this with the protection service's fundraising information allows us to determine whether the current decision is balanced. Based on this, a protection service update strategy that meets the business requirements can be created.

[0073] Specifically, the insurance business prediction model refers to a model that predicts insurance information for each simulated user. The input to this model is the simulated insurance business information of the simulated user, and the output is the user's insurance information. It should be noted that the simulated insurance business information may include information about the simulated user across multiple business attribute dimensions. Therefore, before inputting it into the insurance information prediction model, it can be vectorized to obtain a vector representation of the associated simulated insurance business information, which can then be input into the model for processing. Correspondingly, the user insurance information specifically refers to the information regarding the insurance business providing insurance fund compensation to the simulated user.

[0074] Furthermore, considering that the protection service prediction model is the foundation for adjusting protection services, it can only truly reflect the user's protection information if the prediction accuracy of the protection service prediction model is high enough. Therefore, the protection service prediction model needs to be trained intensively to achieve sufficiently high predictive capabilities. In this embodiment, the training of the protection service prediction model is described in [reference needed]. Figure 3 As shown, steps S2062 to S2066 are included.

[0075] S2062, determine the set of business users associated with the protection service, and obtain the business association data of the business users in the set of business users.

[0076] Specifically, the business user set refers to the set of real users participating in the protection business. Correspondingly, the business-related data refers to the relevant data corresponding to the business users in the business user set when using the protection business. Its composition structure is the same as the protection business simulation information, and it also includes the user protection information of the business users, which is used to build sample pairs in the training model stage.

[0077] Therefore, in order to improve the prediction accuracy of the business forecasting model, real user data can be selected to complete the training during the model training phase. That is, first determine the set of business users associated with the business, and then obtain the business association data of the business users in the set of business users, so as to facilitate the subsequent construction of sample pairs based on this to complete the model training.

[0078] Furthermore, in determining the set of business users and obtaining their business-related data, considering that the data standards for different business users are not uniform in actual application scenarios, constructing samples based on these standards may affect model training. Therefore, after determining the set of business users and obtaining their business-related data, user filtering and data preprocessing can be performed. In this embodiment, the specific implementation method is as follows:

[0079] Determine the initial set of service users associated with the guaranteed service; filter the initial set of service users according to preset filtering conditions to obtain the service user set; obtain the initial service association data of the service users in the service user set; obtain the service association data of the service users in the service user set by preprocessing the initial service association data of the service users in the service user set.

[0080] Specifically, the initial business user set refers to the collection of business users that have not yet been filtered, including all business users who meet and do not meet the filtering criteria. Correspondingly, the filtering conditions refer to the criteria used to select standard users from the initial business user set. Standard users ensure the accuracy, completeness, and availability of relevant data. In other words, the preset filtering conditions can perform accuracy, completeness, and availability filtering on the initial business users, filtering out those with incomplete, inaccurate, or unusable data, and forming the remaining initial business users into the business user set. Correspondingly, the initial business-related data refers to the non-standardized business-related data corresponding to each business user, which may contain missing values, extreme values, etc.

[0081] Therefore, to construct higher-quality training and validation sets, we can first determine an initial set of business users to ensure business association. Then, we filter this initial set according to preset filtering conditions to remove users with missing data, inaccurate data, or unusable data, leaving the remaining initial users as the final business user set. Furthermore, after determining the business user set, we can obtain the initial business association data for each user. Then, we preprocess this initial business association data for each user to obtain standardized business association data for each user, facilitating subsequent use.

[0082] In summary, by filtering the initial set of business users and preprocessing the initial business-related data, the set of business users and their business-related data can be determined in a standardized way during the data preparation stage. This ensures higher data quality and allows for the training of models with higher prediction accuracy.

[0083] Furthermore, during the preprocessing of the initial business-related data, the data is actually standardized. In this embodiment, the specific implementation method is as follows:

[0084] Based on the knowledge graph associated with the protection services, the initial business association data of the business users in the business user set is standardized to obtain the intermediate business association data of the business users in the business user set; based on the natural language algorithm associated with the protection services, the intermediate business association data of the business users in the business user set is repaired to obtain the business association data of the business users in the business user set.

[0085] Specifically, the knowledge graph refers to a graph related to assurance business, containing triples with standard naming conventions within the assurance business scenario, and the knowledge graph is relevant to the assurance business domain. Correspondingly, the natural language algorithm refers to an algorithm capable of correcting non-standard naming or statements. Intermediate business-related data refers to data obtained after standardizing the initial business-related data using triples from the knowledge graph. Standardization refers to processing the data in the initial business-related data involving triples (entity-relationship-entity) in the knowledge graph according to the standard descriptions within the triples. Correspondingly, correction refers to using natural language algorithms to correct inaccurate or non-standard descriptions in the intermediate business-related data.

[0086] Based on this, after obtaining the initial business association data associated with business users, the initial business association data of each business user can be standardized according to the knowledge graph of the association guarantee business, so that the relevant descriptions in the initial business association data can have a unified standard, thus obtaining intermediate business association data; then, the intermediate business association data of each business user can be repaired according to the natural language algorithm of the association guarantee business, so that the relevant descriptions in the intermediate business association data can be correctly described, thereby obtaining the business association data of each business user in the business user set.

[0087] In practical applications, the process of standardizing initial business-related data according to the knowledge graph and repairing intermediate business-related data according to natural language algorithms can be set in a specific order according to actual needs. That is, the initial business-related data can be repaired first according to natural language algorithms, and then the repair results can be standardized according to the knowledge graph. In specific implementation, the settings can be set according to actual needs, and this embodiment does not impose any limitations.

[0088] In other words, in order to build higher-quality training and validation sets in the future, business users can be sorted and standardized based on the initial business association data in the real world that ensures business association. This allows for the selection of natural persons who meet the modeling standards for ensuring business needs and who also meet the standards for data integrity. Certain inclusion and exclusion criteria, i.e., filtering conditions, are set for the selected business users according to the modeling requirements, so that the data associated with business users in the determined set of business users has accuracy, completeness, and availability.

[0089] After determining a more standardized set of business users, we can use knowledge graphs and natural language algorithms related to the assurance business to perform necessary processing on the initial business association data of business users in the assurance business scenario, such as alignment, unification, standardization, missing values, and extreme values, for modeling preparation, thereby obtaining the standard business association data corresponding to each business user.

[0090] For example, if we determine that there are *m* users participating in medical insurance services, and then obtain the medical insurance data associated with each user, we perform accuracy, completeness, and availability checks on the medical insurance data associated with each user. We then determine that *s* users (m ≥ s) out of the *m* users meet these conditions, and these *s* users are then considered the standard user set. Further, we determine the medical insurance data for each user in the standard user set, and perform necessary processing such as alignment, unification, standardization, missing value handling, and extreme value handling according to medical knowledge graphs and natural language processing algorithms to obtain the standard medical insurance data for each user, which is used for subsequent modeling. Let's illustrate the standardization process with user *s1* out of the *s* users. That is, user *s1*'s medical insurance data is {Diagnosis - Sore throat; Medication - Anti-inflammatory drugs; Examination - Blood test; ... Medical insurance expenditure - R yuan}. At this point, the medical insurance data of user s1 is standardized according to the medical knowledge graph and natural language algorithm capabilities, and the standard medical insurance data of user s1 is obtained as {diagnosis result - tonsillitis; medicine - cephalosporin; examination - blood test; ... medical insurance expenditure - R yuan}, which is convenient for subsequent model training.

[0091] In summary, by combining knowledge graphs and natural language processing algorithms to standardize and repair the initial business-related data of business users, the data can be standardized, resulting in higher quality training and validation sets. This leads to the training of more accurate and better generalization-capable business prediction models.

[0092] S2064, construct a training set and a validation set based on the business association data of the business users in the business user set.

[0093] Specifically, after obtaining the business association data of the business users in the business user set, a training set and a validation set can be constructed based on this data. During the training phase, a service prediction model that meets the usage requirements can be trained using these sets. The training set specifically refers to the set used to train the initial service prediction model, and it contains a large amount of data. Correspondingly, the validation set specifically refers to the set used to verify the prediction accuracy of the service prediction model during training.

[0094] Step S2066: Train the initial service prediction model using the training set and the validation set until the service prediction model that meets the training stopping condition is obtained.

[0095] Specifically, after constructing the training and validation sets as described above, the initial business prediction model can be trained using these sets until it meets the training stopping condition. The model at this stopping point can then be used as the business prediction model for practical applications. The training stopping condition refers to the conditions that stop training the initial business prediction model, including but not limited to loss value comparison conditions, iteration count conditions, or validation conditions. Specifically, the loss value comparison condition involves calculating the loss value of the initial business prediction model and comparing it with a preset loss value threshold. Similarly, the iteration count condition involves determining the current number of training iterations and comparing them with a preset iteration count threshold. Likewise, the validation condition involves processing the data in the validation set with the current model to obtain prediction results and comparing them with standard results.

[0096] Furthermore, in order to improve the accuracy of model training when training the initial guarantee service prediction model, a sample set can be used to train the model, and a validation set can be used to validate the model, thereby obtaining a guarantee service prediction model that meets the current use case. In this embodiment, the specific implementation method is as follows:

[0097] The initial support service prediction model is trained using the training set to obtain an intermediate support service prediction model. The intermediate support service prediction model is then validated using the validation set to obtain model prediction information. If the intermediate support service prediction model meets the training stopping condition based on the model prediction information, it is adopted as the support service prediction model. If the training stopping condition is not met, the model continues to be trained using samples from the training set until the condition is met.

[0098] Based on this, after constructing the training set and validation set, the initial support business prediction model can be trained using the samples contained in the training set. The number of training iterations can be set according to actual needs. After meeting the required number of iterations, an intermediate support business prediction model can be obtained. At this point, the validation set can be used to validate the intermediate support business prediction model to obtain model prediction information, i.e., the prediction accuracy of the intermediate support business prediction model at the current stage. Based on the prediction accuracy, it can be determined whether the intermediate support business prediction model meets the training stopping condition. If it does, the intermediate support business prediction model can be used as the support business prediction model. If it does not meet the condition, the model can continue to be trained using the samples in the training set until the condition is met.

[0099] In practical applications, training a business prediction model involves iteratively training its predictive capabilities to obtain a model that meets the usage criteria. Specifically: First, the i-th training sample is selected from the training set and input into the initial business prediction model to obtain the i-th prediction result. Then, the j-th validation sample is selected from the validation set and input into the current stage of the initial business prediction model to obtain the j-th prediction result. Next, the j-th standard validation result corresponding to the j-th validation sample is extracted from the validation set and compared with the j-th prediction result. If they are inconsistent, the model does not meet the training stopping condition. At this point, i is incremented by 1, j is incremented by 1, and the process of inputting the i+1 training sample into the initial business prediction model to obtain the i-th prediction result is repeated until they are consistent, thus obtaining a business prediction model that meets the training stopping condition.

[0100] In summary, by combining the validation set and the training set to train the initial support business prediction model, the prediction accuracy of the initial support business prediction model can be monitored at all times during the training phase. This allows for timely termination of training when the stopping condition is met, which not only improves the model's prediction accuracy but also avoids overfitting, thus effectively ensuring the model's predictive capability.

[0101] Furthermore, before training the initial business prediction model using the training set, considering that in real-world application scenarios, due to time-series issues, the same data may have different prices at different times or locations, in order to ensure that the model has more accurate predictive capabilities, the model can be embedded using a relationship graph. In this embodiment, the specific implementation method is as follows:

[0102] Obtain business object information associated with the guarantee service, and related object information corresponding to the business object information; construct a relationship graph of the guarantee service based on the business object information and the related object information; adjust the parameters of the initial guarantee service prediction model based on the relationship graph, and perform the step of training the initial guarantee service prediction model using the training set to obtain an intermediate guarantee service prediction model based on the parameter adjustment results.

[0103] Specifically, business object information refers to the information corresponding to the business objects used in the insurance business. For example, in the medical insurance business scenario, the business object can be a drug. Correspondingly, associated object information refers to the information associated with the business object, including but not limited to drug prices and the diseases treated. Correspondingly, the relationship graph refers to a dynamic relationship graph constructed based on the business object information and associated object information, which is used to embed the initial insurance business prediction model so that the model can embed relevant information of associated business objects.

[0104] Based on this, before training the model, we can first obtain the business object information of the related guarantee business, as well as the related object information corresponding to the business object information. This allows us to construct a dynamic relationship graph based on the business object information and the related object information. Then, we can tune the parameters of the initial guarantee business prediction model based on the relationship graph, so as to embed the relevant information of the related business objects into the model, so that the model can have higher prediction accuracy in the subsequent training stage.

[0105] In the context of medical insurance business, drug prices vary across different scenarios (hospitals, pharmacies, communities, etc.) and patients' time-series medication use also varies, leading to changes in medical insurance expenditures. Therefore, to more accurately reflect drug prices each time, a dynamic relationship graph can be constructed based on the relationship between specific drugs and diseases in the medical insurance business scenario, as well as drug prices according to different hospital levels and consumption locations. This graph can be embedded into the initial insurance business prediction model during modeling. The initial insurance business prediction model can use a Transformer architecture model. This allows subsequent training of the initial insurance business prediction model with the above information to obtain an insurance business prediction model with higher prediction accuracy.

[0106] In summary, by creating a relationship graph based on business object information and related object information before model training, and embedding this graph into the model during modeling, the model can be optimized through computation. This can further improve the accuracy of model predictions and make the prediction results more consistent with reality.

[0107] Step S208: Generate global protection information based on the user protection information, and create a protection service update strategy according to the protection service information corresponding to the protection service and the global protection information.

[0108] Specifically, after obtaining the user protection information for each simulated user, global protection information related to the protection services can be constructed based on the user protection information. Then, protection service update strategies are created according to the global protection information and the protection service information corresponding to the protection services to update the protection services, so as to ensure that the protection services can maintain a balance between income and expenditure and provide users with better services.

[0109] Specifically, the overall protection information refers to the total user protection information obtained by integrating the user protection information of all simulated users, that is, the total user expenditure under the medical insurance scenario; correspondingly, the protection business information refers to the total financing information of the protection business at the current stage, that is, the actual financing information or planned financing information under the medical insurance scenario; correspondingly, the protection business update strategy refers to the strategy for updating the balance of income and expenditure, which includes, but is not limited to, strategies such as increasing the financing standard and reducing the reimbursement standard.

[0110] In practical applications, when creating a guarantee service update strategy, in order to make precise adjustments to the guarantee service, adjustments can be made based on region and time. That is, the guarantee service in a certain region within a certain time period can be adjusted to maintain its balance of income and expenditure.

[0111] Furthermore, when creating a support service update strategy, considering the full-scale simulation of user data used in the data processing phase, and the geographically specific characteristics of the support service, different levels of early warning schemes can be implemented to improve the accuracy of strategy creation. This can be achieved in the following way:

[0112] The user protection information of simulated users in the simulated user set is integrated to obtain initial global protection information; the initial global protection information is divided according to the protection area corresponding to the protection service to obtain global protection information; the protection service information associated with the protection area is determined; service balance information is calculated based on the global protection information and the protection service information, and the protection service update strategy is created according to the service balance information.

[0113] Specifically, the initial global protection information refers to the protection information obtained after integrating the user protection information of all simulated users; correspondingly, the protection area refers to the geographical area corresponding to the protection business that needs to be updated; and correspondingly, the business balance information refers to the income and expenditure balance information of the protection business, which is used to balance the income and expenditure of the protection business to maintain balance.

[0114] Based on this, after obtaining the user protection information for each simulated user, and considering the need to adjust the revenue and expenditure balance of the protection business, the user protection information of simulated users in the simulated user set can be integrated to obtain initial global protection information for the associated protection business. Furthermore, considering the geographically regional nature of the protection business, the initial global protection information can be divided according to the protection region corresponding to the protection business. This initial global protection information is used to select the user protection information of simulated users associated with the protection region to form the global protection information. Next, the protection business information for the protection region associated with the protection business is determined. Then, the business balance information is calculated by combining the global protection information and the protection business information. Finally, a protection business update strategy for the associated protection business can be created based on the business balance information, so that the revenue and expenditure of the protection business can be balanced according to this strategy.

[0115] In summary, by creating a protection business update strategy based on the protection area, the protection business can take local decisions into account in subsequent updates, thereby accurately balancing the revenue and expenditure of the protection business.

[0116] Furthermore, when creating a coverage service update strategy, considering that coverage services may correspond to different service levels, such as in the medical insurance scenario, the payment standard can be divided into multiple levels, and correspondingly, the revenue and expenditure of different levels also differ. Therefore, the coverage service update strategy can be created according to the level. In this embodiment, the specific implementation method is as follows:

[0117] Determine at least one service level corresponding to the protection service; create a sub-protection service update strategy corresponding to each service level according to the service balance information; and obtain the protection service update strategy associated with the protection service by integrating the sub-protection service update strategies corresponding to each service level.

[0118] Specifically, the business level refers to the level at which the protection business provides protection services to users, with different levels corresponding to different payment and expenditure standards; correspondingly, the sub-protection business update strategy refers to the protection business update strategy corresponding to each business level.

[0119] Based on this, when creating a protection service update strategy corresponding to a protection service, you can first determine at least one business level corresponding to the protection service, then determine the sub-business balance information corresponding to each level based on the business balance information, and then create a sub-protection service update strategy corresponding to each business level according to the sub-business balance information. Finally, by integrating the sub-protection service update strategies corresponding to all business levels, you can obtain the protection service update strategy associated with the protection service.

[0120] In practical applications, after predicting the medical insurance expenditure information of each simulated user through the model, the predicted medical insurance expenditure information of each user can be integrated to obtain the total expenditure of the medical insurance pooling area. At the same time, the actual or planned fundraising information of the medical insurance business fund will also be determined. By combining the total expenditure and fundraising information of the medical insurance pooling area, the income and expenditure balance of the medical insurance fund can be calculated. Then, different levels of early warning schemes can be made for the medical insurance income and expenditure balance, so as to achieve the expenditure balance of the medical insurance fund.

[0121] The information processing method provided in this manual, in order to improve the accuracy of adjustments to protection services, first obtains the attribute distribution information associated with protection services in the business attribute dimension. Then, based on the attribute distribution information, a set of simulated users is created to simulate real users by associating them with relevant information of protection services. The protection service simulation information of the simulated users in the set of simulated users is determined, thereby identifying more standardized and realistic protection service simulation information based on the simulated users. The protection service simulation information is then input into the protection service prediction model for processing to obtain the user protection information of the simulated users in the set of simulated users. Global protection information is generated based on the user protection information. Finally, a protection service update strategy is created according to the protection service information corresponding to the protection service and the global protection information. This achieves the prediction of protection service information through simulated users, and on this basis, the creation of protection service update strategies is completed, effectively improving the prediction accuracy and ensuring the accuracy of the created update strategies. Adjustments to protection services based on this can enable protection services to provide protection services to users while maintaining a balance between revenue and expenditure.

[0122] The following is in conjunction with the appendix Figure 4 Taking the application of the information processing method provided in this specification in a medical insurance business scenario as an example, the information processing method will be further explained. Among them, Figure 4 A flowchart illustrating the processing steps of an information processing method provided in one embodiment of this specification is shown, specifically including the following steps.

[0123] Step S402: Determine the initial set of business users associated with medical insurance services.

[0124] Step S404: Filter the initial set of business users according to preset filtering conditions to obtain the set of business users.

[0125] Step S406: Obtain the initial business association data of business users in the business user set.

[0126] Step S408: According to the medical knowledge graph of the associated medical insurance business, the initial business association data of the business users in the business user set is standardized to obtain the intermediate business association data of the business users in the business user set.

[0127] Step S410: According to the natural language algorithm for related medical insurance business, repair the intermediate business association data of business users in the business user set to obtain the business association data of business users in the business user set.

[0128] In practical applications, in order to build higher-quality training and validation sets in the future, the business users can be sorted and standardized based on the initial business association data of medical insurance business in the real world. This allows for the selection of natural persons who meet the modeling standards for medical insurance business according to the needs of medical insurance business, and who also meet the standards for data integrity. Certain inclusion and exclusion criteria, i.e., filtering conditions, are set for the selected business users according to the modeling requirements, so that the data associated with the business users in the determined business user set has accuracy, completeness, and usability.

[0129] After determining a more standardized set of business users, knowledge graphs and natural language algorithms related to medical insurance business can be used to perform necessary processing on the initial business association data of business users in the medical insurance business scenario, such as alignment, unification, standardization, missing values, and extreme values, for modeling preparation, thereby obtaining the standard business association data corresponding to each business user.

[0130] Step S412: Obtain the business object information related to the medical insurance business, and the associated object information corresponding to the business object information.

[0131] Step S414: Construct a relationship graph of medical insurance business associations based on business object information and associated object information.

[0132] Step S416: Adjust the parameters of the initial medical insurance business prediction model based on the relationship graph; and execute step 418 based on the parameter adjustment results.

[0133] Step S418: Construct a training set and a validation set based on the business association data of the business users in the business user set.

[0134] Step S420: Train the initial medical insurance business prediction model using the training set to obtain the intermediate medical insurance business prediction model.

[0135] Step S422: Validate the intermediate medical insurance business prediction model using the validation set to obtain model prediction information.

[0136] Step S424: If the intermediate medical insurance business prediction model meets the training stopping condition based on the model prediction information, then the intermediate medical insurance business prediction model shall be used as the medical insurance business prediction model.

[0137] In the context of medical insurance business, drug prices vary across different scenarios (hospitals, pharmacies, communities, etc.) and patients' time-series medication use also varies, leading to changes in medical insurance expenditures. Therefore, to more accurately reflect drug prices each time, a dynamic relationship graph can be constructed based on the relationship between specific drugs and diseases in the medical insurance business scenario, as well as drug prices according to different hospital levels and consumption locations. This graph can be embedded into the initial medical insurance business prediction model during modeling. The initial medical insurance business prediction model can use a Transformer architecture model. This allows subsequent training of the initial medical insurance business prediction model with the above information to obtain a medical insurance business prediction model with higher prediction accuracy.

[0138] Step S426: Obtain the attribute distribution information associated with medical insurance business in the business attribute dimension.

[0139] Step S428: Obtain the business simulation parameters and decision simulation parameters of the related medical insurance business.

[0140] Step S430: Update the attribute distribution information based on the business simulation parameters and decision simulation parameters to obtain the target attribute distribution information, and create a set of simulated users based on the target attribute distribution information.

[0141] Step S432: Create simulated medical insurance business information for simulated users in the simulated user set according to the target attribute distribution information.

[0142] Specifically, to make simulated users more closely resemble real users and to create more realistic simulated medical insurance business information, an ABM (Agent-Based Model) can be used to create simulated users and their associated medical insurance business model information. In other words, based on the key simulation variables and important variables in actual medical insurance business decisions when modeling the ABM agent-based model associated with medical insurance business, a dynamic model is derived from the business relationships between the variables in the modeling and the medical insurance business. This dynamic model, or ABM agent-based model associated with medical insurance business, is used to simulate and generate a set of simulated users associated with medical insurance business. The composition of the simulated user set and the simulated medical insurance business information are based on real-world distributions, and are simulated using a dynamic model combined with business simulation parameters and decision simulation parameters to ensure the realism of the simulated users.

[0143] Step S434: Input the medical insurance business simulation information into the medical insurance business prediction model for processing to obtain the user medical insurance information of the simulated users in the simulated user set.

[0144] Step S436: Integrate the medical insurance information of the simulated users in the simulated user set to obtain global medical insurance information.

[0145] Step S438: Determine the medical insurance business information associated with the medical insurance business, calculate the business balance information based on the global medical insurance information and the medical insurance business information, and create a medical insurance business update strategy according to the business balance information.

[0146] Specifically, after predicting the medical insurance expenditure information of each simulated user through the model, the predicted medical insurance expenditure information of each user can be integrated to obtain the total expenditure of the medical insurance pooling area. At the same time, the actual or planned financing information of the medical insurance business fund will also be determined. By combining the total expenditure and financing information of the medical insurance pooling area, the income and expenditure balance of the medical insurance fund can be calculated. Then, different levels of early warning schemes can be made for the medical insurance income and expenditure balance, so as to achieve the expenditure balance of the medical insurance fund.

[0147] In summary, to improve the accuracy of adjustments to protection services, we can first obtain the attribute distribution information associated with protection services across the business attribute dimension. Then, based on this attribute distribution information, we create a set of simulated users to simulate real users by associating them with relevant information of the protection services. We then determine the simulated protection service information for these simulated users, enabling the identification of more standardized and realistic simulated protection service information. This simulated information is then input into a protection service prediction model for processing, yielding the user protection information for each simulated user in the set. Global protection information is generated based on this user protection information. Finally, we create protection service update strategies according to the corresponding protection service information and the global protection information. This approach achieves the prediction of protection service information through simulated users and the creation of protection service update strategies based on this prediction, effectively improving prediction accuracy and ensuring the accuracy of the created update strategies. Adjustments to protection services based on this approach allow the protection service to provide protection services to users while maintaining a balance between revenue and expenditure.

[0148] Corresponding to the above method embodiments, this specification also provides embodiments of an information processing apparatus. Figure 5 A schematic diagram of the structure of an information processing apparatus according to one embodiment of this specification is shown. Figure 5 As shown, the device includes:

[0149] The acquisition module 502 is configured to acquire attribute distribution information associated with the business attribute dimension of the guarantee business;

[0150] The determination module 504 is configured to create a set of simulated users based on the attribute distribution information, and determine the service guarantee simulation information of the simulated users in the set of simulated users;

[0151] Processing module 506 is configured to input the security service simulation information into the security service prediction model for processing, and obtain the user security information of the simulated users in the simulated user set;

[0152] The creation module 508 is configured to generate global protection information based on the user protection information, and to create a protection service update strategy according to the protection service information corresponding to the protection service and the global protection information.

[0153] In an optional embodiment, the apparatus further includes:

[0154] The model training module is configured to determine the set of business users associated with the guarantee service and obtain the business association data of the business users in the set of business users; construct a training set and a validation set based on the business association data of the business users in the set of business users; and train the initial guarantee service prediction model using the training set and the validation set until the guarantee service prediction model that meets the training stopping condition is obtained.

[0155] In an optional embodiment, the model training module is further configured to:

[0156] Determine the initial set of service users associated with the guaranteed service; filter the initial set of service users according to preset filtering conditions to obtain the service user set; obtain the initial service association data of the service users in the service user set; obtain the service association data of the service users in the service user set by preprocessing the initial service association data of the service users in the service user set.

[0157] In an optional embodiment, the model training module is further configured to:

[0158] Based on the knowledge graph associated with the protection services, the initial business association data of the business users in the business user set is standardized to obtain the intermediate business association data of the business users in the business user set; based on the natural language algorithm associated with the protection services, the intermediate business association data of the business users in the business user set is repaired to obtain the business association data of the business users in the business user set.

[0159] In an optional embodiment, the model training module is further configured to:

[0160] The initial support service prediction model is trained using the training set to obtain an intermediate support service prediction model; the intermediate support service prediction model is validated using the validation set to obtain model prediction information; if the intermediate support service prediction model meets the training stopping condition based on the model prediction information, the intermediate support service prediction model is adopted as the support service prediction model.

[0161] In an optional embodiment, the model training module is further configured to:

[0162] Obtain business object information associated with the guarantee service, and related object information corresponding to the business object information; construct a relationship graph of the guarantee service based on the business object information and the related object information; adjust the parameters of the initial guarantee service prediction model based on the relationship graph, and perform the step of training the initial guarantee service prediction model using the training set to obtain an intermediate guarantee service prediction model based on the parameter adjustment results.

[0163] In an optional embodiment, the determining module 504 is further configured to:

[0164] Obtain business simulation parameters and decision simulation parameters associated with the protection service; update the attribute distribution information based on the business simulation parameters and the decision simulation parameters to obtain target attribute distribution information, and create the simulated user set according to the target attribute distribution information; create protection service simulation information for the simulated users in the simulated user set according to the target attribute distribution information.

[0165] In an optional embodiment, the creation module 508 is further configured to:

[0166] The user protection information of simulated users in the simulated user set is integrated to obtain initial global protection information; the initial global protection information is divided according to the protection area corresponding to the protection service to obtain global protection information; the protection service information associated with the protection area is determined; service balance information is calculated based on the global protection information and the protection service information, and the protection service update strategy is created according to the service balance information.

[0167] In an optional embodiment, the creation module 508 is further configured to:

[0168] Determine at least one service level corresponding to the protection service; create a sub-protection service update strategy corresponding to each service level according to the service balance information; and obtain the protection service update strategy associated with the protection service by integrating the sub-protection service update strategies corresponding to each service level.

[0169] In an optional embodiment, the apparatus further includes:

[0170] The adjustment module is configured to obtain the initial business attribute dimension associated with the guaranteed service; and to process the initial business attribute dimension in response to the standard setting instruction to obtain the business attribute dimension.

[0171] The information processing device provided in this manual, in order to improve the accuracy of adjustments to protection services, can first obtain attribute distribution information associated with protection services in the business attribute dimension. Then, based on the attribute distribution information, a set of simulated users is created to simulate real users by associating them with relevant information of the protection services. The protection service simulation information of the simulated users in the simulated user set is determined, enabling the identification of more standardized and realistic protection service simulation information based on the simulated users. This simulation information is then input into the protection service prediction model for processing to obtain the user protection information of the simulated users in the simulated user set. Global protection information is generated based on the user protection information. Finally, a protection service update strategy is created according to the protection service information corresponding to the protection service and the global protection information. This achieves the prediction of protection service information through simulated users, and on this basis, the creation of a protection service update strategy is completed, effectively improving the prediction accuracy and ensuring the accuracy of the created update strategy. Adjustments to the protection service based on this strategy allow the protection service to provide protection services to users while maintaining a balance between revenue and expenditure.

[0172] The above is an illustrative scheme of an information processing device according to this embodiment. It should be noted that the technical solution of this information processing device and the technical solution of the information processing method described above belong to the same concept. For details not described in detail in the technical solution of the information processing device, please refer to the description of the technical solution of the information processing method described above.

[0173] Figure 6 A structural block diagram of a computing device 600 according to one embodiment of this specification is shown. The components of the computing device 600 include, but are not limited to, a memory 610 and a processor 620. The processor 620 is connected to the memory 610 via a bus 630, and a database 650 is used to store data.

[0174] The computing device 600 also includes an access device 640, which enables the computing device 600 to communicate via one or more networks 460. Examples of these networks include a Public Switched Telephone Network (PSTN), a Local Area Network (LAN), a Wide Area Network (WAN), a Personal Area Network (PAN), or a combination of communication networks such as the Internet. The access device 640 may include one or more of any type of wired or wireless network interface (e.g., a Network Interface Card (NIC)), such as an IEEE 802.11 Wireless Local Area Network (WLAN) interface, a Wi-MAX interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, a Near Field Communication (NFC) interface, and so on.

[0175] In one embodiment of this specification, the above-described components of the computing device 600 and Figure 6 Other components, not shown, can also be connected to each other, for example, via a bus. It should be understood that... Figure 6 The block diagram of the computing device shown is for illustrative purposes only and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0176] The computing device 600 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smartphones), wearable computing devices (e.g., smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or PCs. The computing device 600 can also be a mobile or stationary server.

[0177] The processor 620 is configured to execute the following computer-executable instructions, which, when executed by the processor, implement the steps of the above-described information processing method.

[0178] The above is an illustrative scheme of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the information processing method described above belong to the same concept. For details not described in detail in the technical solution of the computing device, please refer to the description of the technical solution of the information processing method described above.

[0179] An embodiment of this specification also provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the above-described information processing method.

[0180] The above is an illustrative scheme of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium and the technical solution of the information processing method described above belong to the same concept. For details not described in detail in the technical solution of the storage medium, please refer to the description of the technical solution of the information processing method described above.

[0181] An embodiment of this specification also provides a computer program, wherein when the computer program is executed in a computer, it causes the computer to perform the steps of the above-described information processing method.

[0182] The above is an illustrative example of a computer program according to this embodiment. It should be noted that the technical solution of this computer program and the technical solution of the information processing method described above belong to the same concept. Details not described in detail in the technical solution of the computer program can be found in the description of the technical solution of the information processing method described above.

[0183] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0184] The computer instructions include computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium may be appropriately added to or subtracted according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media may not include electrical carrier signals and telecommunication signals.

[0185] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments in this specification are not limited to the described order of actions, because according to the embodiments in this specification, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments in this specification.

[0186] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0187] The preferred embodiments disclosed above are merely illustrative of this specification. The optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the embodiments described herein. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the embodiments, thereby enabling those skilled in the art to better understand and utilize this specification. This specification is limited only by the claims and their full scope and equivalents.

Claims

1. An information processing method for medical insurance business, comprising: Obtain attribute distribution information related to medical insurance business in the business attribute dimension; Based on the attribute distribution information, a set of simulated users is created based on the ABM agent base model, and the simulated medical insurance business information of the simulated users in the set of simulated users is determined. The simulated medical insurance business information of the simulated users includes the enrollment time, periodic payment amount and medical insurance expenditure information. The simulated medical insurance business information is input into the medical insurance business prediction model for processing to obtain the user insurance information of the simulated users in the simulated user set. The training method of the medical insurance business prediction model includes: obtaining the business object information associated with the medical insurance business, and the associated object information corresponding to the business object information; constructing a relationship graph of the medical insurance business association based on the business object information and the associated object information; and adjusting the parameters of the initial medical insurance business prediction model based on the relationship graph. Global protection information is generated based on the user protection information, and a medical insurance service update strategy is created according to the medical insurance service information corresponding to the medical insurance service and the global protection information.

2. An information processing method, comprising: Obtain the attribute distribution information associated with the protection business in the business attribute dimension; Based on the attribute distribution information, a set of simulated users is created based on the ABM agent base model, and the insurance business simulation information of the simulated users in the set of simulated users is determined. The insurance business simulation information of the simulated users includes the insurance period, periodic payment amount and insurance expense information. The simulated information of the guarantee service is input into the guarantee service prediction model for processing to obtain the user guarantee information of the simulated users in the simulated user set. The training method of the guarantee service prediction model includes: obtaining the business object information associated with the guarantee service, and the associated object information corresponding to the business object information; constructing a relationship graph of the guarantee service association based on the business object information and the associated object information; and tuning the parameters of the initial guarantee service prediction model based on the relationship graph. Global protection information is generated based on the user protection information, and a protection service update strategy is created according to the protection service information corresponding to the protection service and the global protection information.

3. The method according to claim 2, wherein training the business prediction model includes: Determine the set of business users associated with the guaranteed service, and obtain the business association data of the business users in the set of business users; Based on the business association data of the business users in the business user set, construct a training set and a validation set; The initial service prediction model is trained using the training set and the validation set until a service prediction model that meets the training stopping condition is obtained.

4. The method according to claim 3, wherein determining the set of business users associated with the guaranteed service and obtaining the business association data of the business users in the set of business users includes: Determine the initial set of service users associated with the aforementioned protection services; The initial set of business users is filtered according to preset filtering conditions to obtain the set of business users; Obtain the initial business association data of the business users in the aforementioned business user set; By preprocessing the initial business association data of the business users in the business user set, the business association data of the business users in the business user set is obtained.

5. The method according to claim 4, wherein obtaining the business association data of the business users in the business user set by preprocessing the initial business association data of the business users in the business user set includes: Based on the knowledge graph associated with the security services, the initial business association data of the business users in the business user set is standardized to obtain the intermediate business association data of the business users in the business user set. According to the natural language algorithm associated with the security service, the intermediate business association data of the business users in the business user set is repaired to obtain the business association data of the business users in the business user set.

6. The method according to claim 3, wherein training the initial guarantee service prediction model using the training set and the validation set until obtaining the guarantee service prediction model that meets the training stopping condition comprises: The initial support service prediction model is trained using the training set to obtain an intermediate support service prediction model; The intermediate support service prediction model is validated using the validation set to obtain model prediction information; If the intermediate support service prediction model meets the training stopping condition based on the model prediction information, then the intermediate support service prediction model shall be used as the support service prediction model.

7. The method according to claim 6, further comprising, before the step of training the initial guarantee business prediction model using the training set to obtain the intermediate guarantee business prediction model: Obtain the business object information associated with the protection service, and the associated object information corresponding to the business object information; Construct the relationship graph of the guaranteed business association based on the business object information and the associated object information; The initial support business prediction model is tuned based on the relationship graph, and the intermediate support business prediction model is obtained by training the initial support business prediction model using the training set according to the tuning results.

8. The method according to claim 2, wherein creating a set of simulated users based on the attribute distribution information and determining the service guarantee simulation information of the simulated users in the set of simulated users includes: Obtain the business simulation parameters and decision simulation parameters associated with the protection services; The attribute distribution information is updated based on the business simulation parameters and the decision simulation parameters to obtain the target attribute distribution information, and the simulated user set is created based on the target attribute distribution information. Based on the target attribute distribution information, create the guarantee service simulation information for the simulated users in the simulated user set.

9. The method according to claim 2, wherein generating global protection information based on the user protection information and creating a protection service update strategy according to the protection service information corresponding to the protection service and the global protection information includes: The user protection information of the simulated users in the simulated user set is integrated to obtain initial global protection information; The initial global protection information is divided according to the protection area corresponding to the protection service to obtain the global protection information; Determine the protection service information associated with the protection region; Based on the global protection information and the protection service information, the service balance information is calculated, and the protection service update strategy is created according to the service balance information.

10. The method according to claim 9, wherein creating the service assurance update strategy according to the service balancing information comprises: Determine at least one service level corresponding to the aforementioned protection service; Create a sub-assistance service update strategy for each service level based on the aforementioned service balance information; By integrating the sub-assistance service update strategies corresponding to each service level, the assurance service update strategy associated with the assurance service is obtained.

11. The method according to any one of claims 2-10, wherein before the step of obtaining the attribute distribution information associated with the service in the service attribute dimension is executed, it further includes: Obtain the initial business attribute dimensions associated with the aforementioned security services; The initial business attribute dimension is processed in response to the standard setting command to obtain the business attribute dimension.

12. An information processing apparatus, comprising: The acquisition module is configured to acquire attribute distribution information associated with the business attribute dimension for the protection business; The determination module is configured to create a set of simulated users based on the attribute distribution information and the ABM agent base model, and determine the insurance business simulation information of the simulated users in the set of simulated users. The insurance business simulation information of the simulated users includes the insurance period, periodic payment amount and insurance expense information. The processing module is configured to input the simulated information of the guarantee service into the guarantee service prediction model for processing, and obtain the user guarantee information of the simulated users in the simulated user set. The training method of the guarantee service prediction model includes: obtaining the business object information associated with the guarantee service, and the associated object information corresponding to the business object information; constructing a relationship graph of the guarantee service association based on the business object information and the associated object information; and tuning the parameters of the initial guarantee service prediction model based on the relationship graph. The creation module is configured to generate global protection information based on the user protection information, and to create a protection service update strategy according to the protection service information corresponding to the protection service and the global protection information.

13. A computing device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method according to any one of claims 1 or 2 to 11.

14. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 11.

Images