Individual tax data-based pension insured system adaptation reasoning device, method and equipment and medium

Abstract

The invention discloses an individual tax data-based pension insurance system adaptation reasoning device, method and equipment and a medium, and relates to the technical field of information analysis, and the device comprises an employment information acquisition module which is used for processing tax information through a preset employment information recognition model to determine a target employer and a plurality of employees; the pre-judgment result acquisition module is used for processing the employment information through a preset retirement age judgment model, so that a preset insured qualification pre-judgment model processes the retirement age to obtain an insured qualification pre-judgment result; the ownership result acquisition module is used for processing the insured qualification pre-judgment result through a preset insured ownership judgment model to obtain an insured ownership result; and the insured state judgment module is used for comparing the insured attribution result with preset insured information and judging whether a preset condition is met or not, and if the preset condition is not met, it is judged that the state should be insured but is not insured. Therefore, the to-be-insured information of the pension insurance can be determined based on the employment information.

Description

Technical Field

[0001] The present invention relates to the field of information analysis technology, and in particular to a pension insurance system adaptation reasoning device, method, equipment and medium based on individual tax data. Background Art

[0002] Under the current social insurance fee management system in my country, the participation of employee pension insurance mainly depends on the subjective initiative of the unit or individual, which leads to the fact that employee pension insurance does not fully cover all employed people, that is, there is a problem that a considerable number of employed people are excluded from the employee pension insurance system. Among them, employee pension insurance can be participated in as an enterprise employee or a flexible employee.

[0003] For enterprise employees, the current collection and management methods and mechanisms of employee pension insurance are still based on the unit. In addition, the current data chain for insurance business is serial, and the data source relies only on the report of the employment unit, which easily leads to information asymmetry. Some units take negative actions such as not to insure their employees for the sake of reducing human resource costs, and individual employees may not protect their insurance rights and interests in time due to lack of social security professional knowledge. For flexible employment personnel, they participate in insurance and pay premiums on their own, and the participation in pension insurance is very dependent on personal willingness, because flexible employment information is relatively scattered and hidden. Therefore, to promote flexible employment personnel to participate in pension insurance and enhance their willingness to participate in insurance, the first condition is to find an effective source of information on flexible employment and lock in flexible employment people. Therefore, whether it is enterprise employees or flexible employment personnel, the insurance management cannot be separated from individual tax data.

[0004] From the above, it can be seen that how to determine the pension insurance information of employed persons based on personal tax data and identify the employed people who should participate but have not participated are issues that need to be urgently resolved. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a pension insurance system adaptation reasoning device, method, equipment and medium based on personal tax data, which can determine the pension insurance information of employed persons based on their employment information, thereby solving and identifying the problem of employed persons excluded from employee pension insurance, improving the efficiency of employee pension insurance collection and management and service quality, and promoting the sustainable development of the pension insurance system. The specific scheme is as follows:

[0006] In a first aspect, the present application provides a pension insurance system adaptation reasoning device based on individual tax data, including:

[0007] An employment information acquisition module is used to process the tax registration information and individual tax declaration data through a preset employment information recognition model to determine a target employer and a number of employees corresponding to the target employer;

[0008] A prediction result acquisition module is used to process the employment information corresponding to the plurality of employed persons through a preset retirement age judgment model, so that the preset insurance qualification prediction model processes the retirement age output by the preset retirement age judgment model to obtain insurance qualification prediction results corresponding to the plurality of employment information respectively;

[0009] The attribution result acquisition module is used to process the insurance qualification pre-judgment result and the corresponding employment information through a preset insurance attribution judgment model to obtain the insurance attribution result corresponding to the employed person;

[0010] The insurance status judgment module is used to compare the insurance attribution result with the preset insurance information, and judge whether the comparison result meets the preset conditions. If not, the insurance status of the employed person corresponding to the comparison result is judged as a state of being insured but not being insured.

[0011] Optionally, the prediction result acquisition module includes:

[0012] A retiree determination unit, used to determine whether the employed person is a former statutory retiree;

[0013] A first retirement age determination unit is used for, if the employed person is a former statutory retiree, processing the employment information corresponding to the employed person by using a preset former statutory retirement age judgment model to obtain a first retirement age corresponding to the employed person;

[0014] A first prediction result determination unit is used to process the first retirement age and the employment information corresponding to the employed person through a preset insurance qualification prediction model to obtain a first insurance qualification prediction result corresponding to the employment information;

[0015] A second retirement age determination unit is used for processing the employment information corresponding to the employed person through a preset model for determining the statutory retirement age after delayed retirement to obtain a second retirement age corresponding to the employed person if the employed person is not an original statutory retiree;

[0016] The second prediction result determination unit is used to process the second retirement age and the employment information corresponding to the employed person using a preset insurance qualification prediction model to obtain a second insurance qualification prediction result corresponding to the employment information; wherein the first insurance qualification prediction result and the second insurance qualification prediction result are employee insurance qualifications or flexible employment insurance qualifications.

[0017] Optionally, the attribution result acquisition module includes:

[0018] A qualification judgment unit is used to judge whether the qualification corresponding to the employed person is an employee qualification based on the pre-judgment result of the qualification;

[0019] A first insurance unit determination unit is used for processing the employment information corresponding to the employed person through a preset insurance unit judgment model to obtain an insurance unit result if the insurance qualification corresponding to the employed person is an employee insurance qualification;

[0020] The insurance location determination unit is used to process the employment information corresponding to the employed person through a preset insurance location judgment model to obtain an insurance location result if the insurance qualification corresponding to the employed person is not an employee insurance qualification.

[0021] Optionally, the attribution result acquisition module includes:

[0022] A multiple-income source determination unit, used to determine whether the employment information corresponding to the employed person meets the preset multiple-income source condition;

[0023] The unit score determination unit is used for, if the employed person meets the preset multiple-source income condition, using the preset insurance attribution judgment model and setting corresponding weights for the corresponding employment information based on the insurance qualification prediction result, and determining the unit scores corresponding to the several employment information based on the several weights and the corresponding employment information; wherein the insurance qualification prediction result is that the weight corresponding to the salary is greater than the weight corresponding to the owner of the self-employed business with employees, the weight corresponding to the owner of the self-employed business with employees is greater than the weight corresponding to the sole proprietorship investor with employees, and the weight corresponding to the sole proprietorship investor with employees is greater than the weight corresponding to the partner of a partnership with employees;

[0024] A second insurance-eligible unit determination unit is used to extract a target score with the highest score from the scores of the units, and set the unit corresponding to the target score as the insurance-eligible unit;

[0025] The third insurable unit determination unit is used to set the unit corresponding to the employee as an insurable unit if the employee does not meet the preset multiple income source conditions.

[0026] Optionally, the attribution result acquisition module includes:

[0027] A multiple salary determination unit, used to determine whether the employee has multiple salary incomes if the employee meets the preset multiple income source conditions;

[0028] The first target matrix determination unit is used to normalize the withholding and declared income amounts of several previous periods and several employment dates in the employment information respectively through a preset insurance attribution judgment model if the employed person has multiple salary incomes, so as to obtain the current normalization result, and determine the first target matrix based on the current naturalization result and the first preset weight matrix; wherein the first target matrix includes weighted scores corresponding to the withholding and declared income amounts of several previous periods and several employment dates in the employment information.

[0029] Optionally, the attribution result acquisition module includes:

[0030] A multiple self-employment judgment unit, used to judge whether the employee has multiple self-employments with the same registration type if the employee meets the preset multiple income source conditions;

[0031] The second target matrix determination unit is used to normalize the investment time, labor contribution mapping factors, average social wages of investment places, and consistency between the registered places of residence of several employees and the investment places in the employment information through a preset insurance participation judgment model if the employed person has multiple self-employed businesses with the same registration type, so as to obtain the current normalization result, and determine the second target matrix based on the current normalization result and a second preset weight matrix; wherein the second target matrix includes weighted scores corresponding to the investment time, labor contribution mapping factors, average social wages of investment places, and consistency between the registered places of residence of several employees and the investment places in the employment information.

[0032] Optionally, the attribution result acquisition module includes:

[0033] The weight optimization unit is used to set corresponding optimized weights for the employment information through a preset insurance attribution judgment model and a preset weight adjustment strategy, so that the preset insurance attribution judgment model uses the optimized weights and the corresponding employment information to re-determine the optimized unit scores corresponding to several pieces of employment information.

[0034] Secondly, this application provides a pension insurance system adaptation reasoning method based on individual tax data, including:

[0035] Processing tax registration information and individual tax declaration data through a preset employment information recognition model to determine a target employer and a number of employees corresponding to the target employer;

[0036] The employment information corresponding to the plurality of employed persons is processed by a preset retirement age judgment model, so that the preset insurance qualification prediction model processes the retirement age output by the preset retirement age judgment model to obtain insurance qualification prediction results corresponding to the plurality of employment information respectively;

[0037] Processing the insurance qualification pre-judgment result and the corresponding employment information through a preset insurance attribution judgment model to obtain an insurance attribution result corresponding to the employed person;

[0038] The insurance attribution result is compared with the preset insurance information, and it is determined whether the comparison result meets the preset conditions. If not, the insurance status of the employed person corresponding to the comparison result is determined as a state where he should be insured but is not insured.

[0039] In a third aspect, the present application provides an electronic device, including:

[0040] Memory, used to store computer programs;

[0041] A processor is used to execute the computer program to implement the aforementioned pension insurance system adaptation reasoning method based on individual tax data.

[0042] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the aforementioned pension insurance system adaptation reasoning method based on individual tax data.

[0043] As can be seen from the above, in the present application, the employment information acquisition module is used to process the tax registration information and the individual tax declaration data through a preset employment information recognition model to determine the target employer and a number of employees corresponding to the target employer; the prediction result acquisition module is used to process the employment information corresponding to the several employees through a preset retirement age judgment model, so that the preset insurance qualification prediction model processes the retirement age output by the preset retirement age judgment model to obtain the insurance qualification prediction results corresponding to the several employment information respectively; the attribution result acquisition module is used to process the insurance qualification prediction result and the corresponding employment information through a preset insurance attribution judgment model to obtain the insurance attribution result corresponding to the employees; the insurance status judgment module is used to compare the insurance attribution result with the preset insurance information, and judge whether the comparison result meets the preset conditions. If not, the insurance status of the employed person corresponding to the comparison result is judged as a state of being insured but not insured. It can be seen that, on the one hand, this application processes the tax registration information and individual tax declaration data through a preset employment information identification model to determine the target employer and a number of employees corresponding to the target employer; on the other hand, this application processes the employment information through a preset retirement age judgment model to obtain the retirement age, and processes the retirement age through a preset insurance qualification prediction model to obtain the insurance qualification prediction result; on the other hand, this application processes the employment information through a preset insurance attribution judgment model to obtain the insurance attribution result; and then compares the insurance attribution result with the preset insurance information to obtain the insurance status of the employed person. In this way, the insurance information of the employed person's pension insurance can be determined based on the individual tax data of the employed person, thereby solving and identifying the problem of employed persons excluded from the employee pension insurance, improving the efficiency of employee pension insurance collection and management and service quality, promoting the sustainable development of the pension insurance system, and promoting the high-quality development of my country's social security cause. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0045] Figure 1 This is a schematic diagram of the structure of a pension insurance system adaptation reasoning device based on individual tax data disclosed in this application;

[0046] Figure 2 It is a schematic diagram of identification rules of a specific employment information identification model;

[0047] Figure 3 This is a schematic diagram of a privacy computing cluster processing flow;

[0048] Figure 4 A schematic diagram of identification rules corresponding to a specific original statutory retirement age judgment model and a statutory retirement age judgment model after delayed retirement;

[0049] Figure 5 This is a schematic diagram of identification rules for a specific insurance eligibility prediction model;

[0050] Figure 6 It is a schematic diagram of identification rules of a specific insurance eligibility judgment model;

[0051] Figure 7 It is a schematic diagram of identification rules of a specific model for determining employment units that should participate in insurance;

[0052] Figure 8 A schematic diagram for extracting information for a specific data;

[0053] Fig. 9 A schematic diagram for extracting information for another specific data;

[0054] Fig.10 It is a schematic diagram of identification rules of a specific insurance-eligible location judgment model;

[0055] Fig.11 A flowchart of a pension insurance system adaptation reasoning method based on individual tax data disclosed in this application;

[0056] Fig.12 A flowchart for determining the pension insurance participation status of employed persons based on employment information;

[0057] Fig.13 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0058] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] Under the current social insurance fee management system in my country, the participation of employee pension insurance depends on the subjective initiative of the unit or individual, which leads to the fact that the employee pension insurance does not fully cover all employed people, that is, there is a problem that some employed persons are excluded from the employee pension insurance. To this end, this application provides a pension insurance system adaptation reasoning device based on individual tax data, which can determine the pension insurance information of employed persons based on the employment information of employed persons, thereby solving and identifying the problem of employed persons excluded from the employee pension insurance, and improving the collection and management efficiency and service quality of employee pension insurance.

[0060] See also Figure 1 As shown, the embodiment of the present invention discloses a pension insurance system adaptation reasoning device based on individual tax data, including:

[0061] The employment information acquisition module 11 is used to process the tax registration information and the individual tax declaration data through a preset employment information recognition model to determine a target employer and a number of employees corresponding to the target employer.

[0062] In this embodiment, the tax information of employed persons covers a wide range of employment data, has good timeliness, and has certain legal effect. Therefore, the embodiment of the present application uses the tax information of employed persons as a reflection of the employment information of employed persons, and thus uses the tax information of employed persons as a basis for judging whether the employed persons are covered by pension insurance. In addition, the tax registration information records the investment relationship between individuals and individual business entities; the individual tax declaration data includes individual tax comprehensive income declaration data and individual tax business income declaration data. Among them, the wage and salary or labor remuneration withholding declaration data in the individual tax comprehensive income declaration data records the employment or employment relationship between the employer and the individual and the specific amount of labor or labor remuneration obtained by the individual; the individual tax business income declaration data records the business fact information of the individual business entity.

[0063] In this embodiment, the employment information must first be identified; it is worth mentioning that in the process of identifying the employment information, all current employment information of the individual must be collected in advance so that it can be used to predict the insurance eligibility of the employed persons in the future. The tax registration information and the individual tax declaration data are processed by the preset employment information identification model to obtain the employment information corresponding to the employed population across the country. It is worth mentioning that the preset employment information identification model first determines the target employer based on the tax registration information; then, the embodiment of the present application can determine the employed persons who have a salary or labor remuneration withholding relationship with the target employer and the relevant employment information corresponding to the employed persons based on the individual tax comprehensive income declaration data in the individual tax declaration data; in addition, the embodiment of the present application can determine the effective self-employment information of the individual business owner, the natural person investor of the sole proprietorship, and the natural person partner of the partnership based on the individual tax business income declaration data.

[0064] It is understandable that the model processing rules of the preset employment information recognition model are as follows: Figure 2 As shown:

[0065] In this embodiment, the model processing rules of the preset employment information identification model include the identification rules of the employers who should be insured and the identification rules of the employed population. Among them, the rules for identifying the corresponding insured employers, that is, the target employers, are: first, screen the valid unit tax registration and individual business tax registration, among which individual business includes individual industrial and commercial households, sole proprietorships, and partnerships; second, exclude rural collective-owned enterprises. It is worth mentioning that if such entities of the farmer's professional cooperatives in the province are not in the scope of mandatory insurance, the farmer's professional cooperatives should also be excluded; finally, exclude the individual industrial and commercial households, sole proprietorships, and partnerships without employees; among which, no employees means that there is no individual income tax salary withholding declaration. The identification rules for employed people are to meet one of the following three conditions: the first condition is that there are persons who should be insured and whose employers should withhold taxes on wages and salaries; the second condition is that there are persons who should be insured and whose employers should withhold taxes on labor remuneration; the third condition is a valid individual business owner, a natural person investor of a sole proprietorship, or a natural person partner of a partnership (including those with or without employees).

[0066] It is worth mentioning that for some data involving personal privacy, such as income data, privacy computing and other technical methods can be used to achieve the data utilization requirements of "original data does not leave the domain, data is available but not visible". To this end, a preset privacy computing cluster is added in the embodiment of the present application, wherein the preset privacy computing cluster includes a privacy computing scheduling platform and computing nodes deployed on several data provider cloud platforms respectively, so as to use the privacy computing scheduling platform and privacy computing nodes to process tax registration information and personal tax declaration data, so as to realize the sharing of privacy data calculation results without leaving the computing node. Among them, the relationship and process between the privacy computing scheduling platform and each privacy computing node are as follows: Figure 3 shown.

[0067] In this embodiment, privacy computing needs to be completed in the computing node, and it is impossible to use the aggregated multi-source data for computing. Therefore, the technologies that need to be used in the process of computing tax registration information and individual tax declaration data include: privacy data protection technology: using common data privacy protection technologies such as homomorphic encryption and differential privacy; privacy computing technology: based on encrypted or obfuscated data, the implementation technology of multi-party secure privacy computing without plaintext data leaving the computing node, including privacy intersection, hidden query, joint statistical analysis, and support for plaintext and ciphertext mixed programming, where privacy intersection includes but is not limited to malicious models and semi-honest models. Security networking technology: quickly create a privacy computing network or cluster on the existing hardware infrastructure of the privacy computing scheduling platform, data provider, etc. to achieve secure data transmission within the network. Parallel computing logic: because privacy computing requires that the original data does not leave the domain, for the data mixed computing scenario of multiple data sources, it is necessary to reduce the parallel computing logic of cross-departmental data, use single-node computing as much as possible, and merge the results to achieve the final parallel computing result requirements.

[0068] The prediction result acquisition module 12 is used to process the employment information corresponding to the plurality of employed persons through the preset retirement age judgment model, so that the preset insurance qualification prediction model processes the retirement age output by the preset retirement age judgment model to obtain the insurance qualification prediction results corresponding to the plurality of employment information respectively.

[0069] In this embodiment, after obtaining the target employer and a number of employees corresponding to the target employer, it is necessary to process the employment information corresponding to the several employees through a preset retirement age judgment model to obtain the retirement age output by the preset retirement age judgment model. The prediction result acquisition module includes: a retiree judgment unit, used to judge whether the employee is a former statutory retiree; a first retirement age determination unit, used to process the employment information corresponding to the employee through a preset original statutory retirement age judgment model if the employee is a former statutory retiree, to obtain the first retirement age corresponding to the employee; a first prediction result determination unit, used to process the first retirement age and the employment information corresponding to the employee through a preset insurance qualification prediction model to obtain a first insurance qualification prediction result corresponding to the employment information;

[0070] The second retirement age determination unit is used to process the employment information corresponding to the employed person through the preset model for judging the statutory retirement age after delayed retirement if the employed person is not the original statutory retiree, so as to obtain the second retirement age corresponding to the employed person; the second prediction result determination unit is used to process the second retirement age and the employment information corresponding to the employed person using the preset insurance qualification prediction model, so as to obtain the second insurance qualification prediction result corresponding to the employment information. Among them, the first insurance qualification prediction result and the second insurance qualification prediction result are employee insurance qualifications or flexible employment insurance qualifications. It is worth mentioning that if a person has multiple employment information, the insurance qualification prediction results corresponding to each employment information will be output respectively.

[0071] In a specific implementation manner, the model rules corresponding to the preset original statutory retirement age judgment model and the preset statutory retirement age judgment model after delayed retirement are as follows: Figure 4As shown in the following figure: In the preset model for judging the original statutory retirement age, for the eligible workers, the retirement age is 60 for men, 55 for female cadres, and 50 for female workers; for the eligible flexible employment, the retirement age is 60 for men and 55 for women. In the preset model for judging the statutory retirement age after delayed retirement, for the eligible workers, the retirement age is: if the date of birth is December 1976 or earlier: retirement age = 60 + (year of birth - 1965) * 3 / 12 + if (month of birth < 5, 1 / 12, if (month of birth < 9, 2 / 12, 3 / 12)); if the date of birth is January 1977 or later: retirement age = 63. The retirement age for female workers is: if the date of birth is December 1981 or earlier: retirement age = 55 + (birth year - 1970) * 3 / 12 + if (birth month < 5, 1 / 12, if (birth month < 9, 2 / 12, 3 / 12)); if the date of birth is January 1982 or later: retirement age = 58. The retirement age for female workers is: if the date of birth is December 1984 or earlier: retirement age = 50 + (birth year - 1975) * 6 / 12 + if (birth month < 3, 1 / 12, if (birth month < 5, 2 / 12, if (birth month < 7, 3 / 12, if (birth month < 9, 4 / 12, if (birth month < 11, 5 / 12, 6 / 12))))); if the date of birth is January 1985 or later: retirement age = 55.

[0072] In another specific implementation, the model rules corresponding to the preset insurance qualification prediction model are as follows: Figure 5As shown: For the judgment of whether the eligible group of employees participating in the insurance is a salaried worker, the embodiment of the present application needs to meet the following four conditions at the same time: The first condition: the employed person is a person who has not received benefits; The second condition: the nationality (region) of the employed person is the target nationality; The third condition: if the employed person is a male, the employed person's age is between 16 years old and the corresponding retirement age; if the employed person is a female, the employed person's age is between 16 years old and the corresponding retirement age; The fourth condition: there is a withholding declaration of salary income in the previous period, and as of now, the employed person's status in the withholding unit is "employed". Regarding the judgment of whether the eligible group of employees participating in the insurance is an individual, the embodiments of the present application need to meet the following four conditions at the same time: the first condition: the employed person is a person who has not received benefits; the second condition: the nationality (region) of the employed person is the target nationality; the third condition: if the employed person is a male, the employed person's age is between 16 and the corresponding retirement age; if the employed person is a female, the employed person's age is between 16 and the corresponding retirement age; the fourth condition: there is currently valid registration information of the natural person investor of a sole proprietorship / natural person partner of a partnership / individual business owner, and the unit has employment or salary withholding declaration information for the December period of the previous year.

[0073] For the judgment of whether the eligible people for gig employment insurance are laborers, the embodiments of the present application need to meet the following four conditions at the same time: the first condition: the employed person is a person who has not received benefits; the second condition: the nationality (region) of the employed person is the target nationality; the third condition: if the employed person is male, the employed person's age is between 16 and the corresponding retirement age; if the employed person is female, the employed person's age is between 16 and the corresponding retirement age; the fourth condition: there is a withholding declaration of labor remuneration income in the previous period, and the following two conditions are met at the same time: 1. The employed person has a cumulative withholding declaration record of comprehensive income for n months or more in the local area, or the employed person has declared salary income in the local area within 12 months; it is worth mentioning that the above conditions can be set according to the local flexible employment qualification judgment standards, or not; 2. The employed person's cumulative average monthly labor remuneration income in the local area reaches m yuan in the current year, where m is the payment amount based on the local minimum wage and the lowest level of flexible employment, and is determined by increasing a certain margin to determine whether the employed person has the ability to pay.

[0074] Regarding the judgment of whether the people eligible for insurance for gig workers are individuals, the embodiments of the present application need to meet the following four conditions at the same time: the first condition: the employed person is a person who has not received benefits; the second condition: the nationality (region) of the employed person is the target nationality; the third condition: if the employed person is male, the employed person's age is between 16 and the corresponding retirement age; if the employed person is female, the employed person's age is between 16 and the corresponding retirement age; the fourth condition: there is currently valid registration information of the natural person investor of a sole proprietorship / natural person partner of a partnership / individual business owner, and there is no employee information (wage withholding declaration) for the December period of the previous year for the unit.

[0075] The attribution result acquisition module 13 is used to process the insurance qualification pre-judgment result and the corresponding employment information through a preset insurance attribution judgment model to obtain the insurance attribution result corresponding to the employed person.

[0076] In this embodiment, after obtaining the insurance qualification prediction result corresponding to the employed person, the insurance qualification prediction result is processed by the preset insurance qualification judgment model to determine the insurance qualification judgment result. Among them, the model rules of the preset insurance qualification judgment model are as follows: Figure 6 As shown: When an individual has one or more predicted insurance qualification information, all the predicted insurance qualification information of the individual is collected, and the following is judged based on the collection results: when an individual has only one insurance qualification information, that insurance qualification is taken; when an individual has both employee and flexible employment insurance qualifications, the employee insurance qualification is taken.

[0077] It is worth mentioning that after obtaining the pre-judgment result of the insurance qualification corresponding to the employed person, it is necessary to judge whether the insurance qualification corresponding to the employed person is the employee insurance qualification based on the pre-judgment result of the insurance qualification, that is, the types of insurance qualifications corresponding to different employed persons correspond to different preset insurance attribution judgment models. The attribution result acquisition module includes: a qualification judgment unit, which is used to judge whether the insurance qualification corresponding to the employed person is the employee insurance qualification based on the pre-judgment result of the insurance qualification; a first insured unit determination unit, which is used to process the employment information corresponding to the employed person through the preset insured employment unit judgment model if the insurance qualification corresponding to the employed person is the employee insurance qualification, so as to obtain the insured unit result; and an insured place determination unit, which is used to process the employment information corresponding to the employed person through the preset insured place judgment model if the insurance qualification corresponding to the employed person is not the employee insurance qualification, so as to obtain the insured place result.

[0078] Among them, the model rules corresponding to the preset insured employment unit judgment model are as follows: Figure 7As shown: When an individual has only one withholding declaration for wage and salary income nationwide, the withholding unit will be identified as the insured employment unit; when an individual has only one self-employed investment registration with employees nationwide, the self-employed entity will be identified as the insured employment unit; if an individual has multiple wage and salary income, or self-employed investment registrations with employees nationwide at the same time, the judgment model will be used to determine the only insured employment unit. The attribution result acquisition module includes: a multiple-income source judgment unit, which is used to judge whether the employment information corresponding to the employed person meets the preset multiple-income source condition; a unit score determination unit, which is used to, if the employed person meets the preset multiple-income source condition, use the preset insurance attribution judgment model and set corresponding weights for the corresponding employment information based on the insurance qualification prediction result, and determine the unit scores corresponding to the several employment information based on the several weights and the corresponding employment information; wherein the insurance qualification prediction result is that the weight corresponding to the salary is greater than the weight corresponding to the owner of the individual industrial and commercial household with employees, the weight corresponding to the owner of the individual industrial and commercial household with employees is greater than the weight corresponding to the sole proprietorship investor with employees, and the weight corresponding to the sole proprietorship investor with employees is greater than the weight corresponding to the partner of a partnership with employees; a second insured unit determination unit, which is used to extract the target score with the highest score from the several unit scores, and set the unit corresponding to the target score as the insured unit;

[0079] The third insurable unit determination unit is used to set the unit corresponding to the employee as an insurable unit if the employee does not meet the preset multiple income source conditions.

[0080] In the first specific implementation method, compared with self-employed owners with employees, the insurance participation of employers who have formed a labor relationship with individuals is more mandatory. Therefore, the wage and salary withholding units that have a stronger orientation towards labor relations are preferred as the insured units; self-employment includes three categories: self-employed businesses and sole proprietorships with only one owner, which take precedence over partnerships with at least 2 partners; and self-employed businesses are superior to sole proprietorships.

[0081] In the second specific implementation method, for those with multiple sources of wage and salary income, the weights of each factor are initially assigned based on expert experience through the factors of "income amount" and "time of employment", and the employer with the highest score is calculated as the employer that should participate in the insurance. The model is subsequently optimized based on actual feedback results to make the model output results more in line with effective full-time labor relations.

[0082] In the third specific implementation method, for individual businesses with multiple registration types with the same type, the weights of the factors are initially assigned based on expert experience through the factors of "investment time", "total number of employees per year", "local treatment (social average) level", and "consistency between the owner's registered residence and the investment location", and the individual business entity with the highest score is calculated as the employment unit that should participate in the insurance. The model is subsequently optimized based on actual feedback results to make the model output results more in line with the individual's willingness to participate in insurance.

[0083] Among them, the size relationship between the various types of priorities is as follows: wage and salary income > self-employed business owners with employees > sole proprietors with employees > partners of partnerships with employees;

[0084] It is worth mentioning that if there are multiple sources of salary income, the employer with the highest weighted score among the weighted scores corresponding to the multiple sources shall be the insured unit. The formula for determining the weighted score is as follows:

[0085] Weighted score = w1*income amount + w2*time of employment; w1 and w2 are the weights initially assigned to each factor based on expert experience.

[0086] If there are multiple individual business investment registrations with the same registration type, the individual business entity with the highest weighted score among the individual businesses with the same registration type shall be the insured entity. The formula for determining the weighted score is as follows:

[0087] Weighted score = w3*investment time + w4*total number of employees in an year + w5*local treatment (social average) level + w6*consistency between the owner's registered residence and the investment location; among them, w3, w4, w5 and w6 are the weights initially assigned to each factor based on expert experience.

[0088] It is worth mentioning that if an employed person has only one source of salary income, or is a self-employed business owner with employees, or is a sole proprietor with employees, or is a partner in a partnership with employees, a deterministic judgment can be made based on deterministic rules. However, when an employed person has multiple sources of salary income, or there are multiple sources of self-employed income with the same registration type, how to determine the insured employment unit of the employed person becomes a process of uncertain reasoning. Therefore, the preset insured unit judgment model is a hybrid model that performs deterministic reasoning based on rules for general situations and uses a variant of the greedy algorithm for uncertain reasoning based on the characteristics of the factors involved in special situations.

[0089] Among them, there are two special situations, namely, situation one: employed persons have multiple salary incomes, and situation two: there are multiple individual business incomes with the same registration type.

[0090] In this embodiment, when processing the first situation, the data needs to be imported and processed first, that is, the required rows and columns stored in the database are extracted through data slicing and connection operations. In a specific implementation, the field type for extracting the employment information of the employed person is as follows: Figure 8 As shown: Taking Zhang San as an example, all records with the attribute value of "Zhang San" under the "Name" field and the salary income under the "Employment Type" field are extracted. In the final table, the company name is listed from A to X, the withholding income amount of the previous period is listed from A1 to X1, and the employment date is listed from A2 to X2.

[0091] It is worth mentioning that the attribution result acquisition module includes: a multiple salary judgment unit, which is used to judge whether the employed person has multiple salary incomes if the employed person meets the preset multiple income source conditions; a first target matrix determination unit, which is used to normalize the several previous period withholding and declared income amounts and several employment dates in the employment information through a preset insurance attribution judgment model if the employed person has multiple salary incomes, so as to obtain the current naturalization result, and determine the first target matrix based on the current naturalization result and the first preset weight matrix; wherein the first target matrix includes weighted scores corresponding to the several previous period withholding and declared income amounts and several employment dates in the employment information.

[0092] In a specific implementation, the withholding declared income amount (A1 to X1) of the previous period is normalized, and the processing result is represented by a1 to x1. It is worth mentioning that the normalization method used in the embodiment of the present application is the Min-Max normalization method, and the process is as follows:

[0093] First, calculate the maximum and minimum values ​​from A1 to X1, and the maximum value x max =K1, minimum value x min =M1. It is worth mentioning that the normalized value of each eigenvalue is represented by its corresponding lowercase letter. For example, the normalized value of A1 is a1, and the normalization formula is as follows: a1 = (A1-M1) / (K1-M1), and the normalized eigenvector is [a1, b1, c1, d1, ..., x1].

[0094] Secondly, for the employment date (A2 to X2), the embodiment of the present application needs to solve the number of days relative to the current date, take the reciprocal, and then perform normalization. Here, Min-Max normalization is also used, the process is the same as above, and it will not be repeated here. The normalized eigenvalue vector is [a2, b2, c2, d2, ..., x2].

[0095] Finally, the embodiment of the present application mixes the feature information matrix of different types of variables in the original data Transformed into the normalized eigenvalue matrix That is, a matrix expression of the model input space.

[0096] It is worth mentioning that the embodiment of the present application uses a weighted greedy algorithm (i.e., Weighted GreedyAlgorithm) to construct a preset insured unit judgment model. For example, a certain weight is assigned to the employee's withholding income and other characteristics, such as working hours, in the previous period. For example, let the rating index (Ranking) be R, the withholding income (Salary) in the previous period be s, and the corresponding weight (weight) be ω s , working time t, working time weight is ω t The formula for determining the weighted score is as follows:

[0097] R=ω s *s+ω t *t;

[0098] After obtaining several weighted scores according to the above formula, the embodiment of the present application selects the unit with the highest score from the several weighted scores as the insured employment unit of the employed person. Therefore, based on the above steps, the expression for constructing the preset insured unit judgment model is as follows:

[0099]

[0100] It is worth mentioning that if R c R1 to R x The maximum value among them, then R c The corresponding C unit is Zhang San's insured employment unit. If, after subsequent comparison with the collected Zhang San information, the insured employment unit selected by Zhang San is indeed C unit, it means that the preset insured employment unit judgment model prediction is correct. For other employees who also have multiple salary situations, the above operation should also be performed, and the data with correct result predictions should be counted to use the accuracy determination formula to calculate the accuracy of the model output. In this embodiment, the present application embodiment selects accuracy as an indicator for model optimization. Among them, the accuracy determination formula is as follows:

[0101] Accuracy = Number of Correct Prediction (number of correctly predicted samples) / Total number of Prediction (total number of predicted samples).

[0102] It is worth mentioning that when setting initialization parameters for the weight vector, the higher the importance of the feature corresponding to the initialization parameter, the higher the corresponding weight. In a specific implementation, the initialization parameters may be ω1=1, ω2=1.

[0103] It is worth mentioning that the weights corresponding to each factor can be optimized. The attribution result acquisition module includes: a weight optimization unit, which is used to set corresponding optimized weights for the employment information by using a preset insurance attribution judgment model and a preset weight adjustment strategy, so that the preset insurance attribution judgment model uses the optimized weights and the corresponding employment information to re-determine the optimized unit scores corresponding to the employment information. In a specific implementation, the preset weight adjustment strategy can be an index adjustment strategy, and the optimized weight determination method is as follows:

[0104] ω1=2ω1;

[0105] ω2=ω2(1-5%);

[0106] Subsequently, the embodiment of the present application can use the adjusted weights to retrain the model and calculate new performance indicators. It is worth mentioning that as the weights are gradually optimized, the accuracy of the model will gradually increase, but if the weights are optimized too many times, the accuracy of the model will be reduced, that is, overfitting will occur. Therefore, the accuracy of the output of the optimized model must be determined each time, and the optimization of the weights will be stopped when the accuracy of the model drops for the first time, and the weights obtained from the last training will be used as the optimal weights. It is worth mentioning that in order to ensure the generalization ability of the model in the future, k-fold cross-validation (i.e., k-fold cross-validation) is also required.

[0107] In this embodiment, when processing the second situation, the data needs to be imported and processed first, that is, the required rows and columns stored in the database are extracted through data slicing and connection operations. In a specific implementation, taking Li Si as an example, all records of individual businesses with the attribute value of "Li Si" under the "Name" field and employees under the "Employment Type" field are extracted. The extracted results are as follows: Fig. 9As shown. The attribution result acquisition module includes: a multiple self-employment judgment unit, which is used to judge whether the employed person has multiple self-employments of the same registration type if the employed person meets the preset multiple income source conditions; a second target matrix determination unit, which is used to normalize the investment time, the labor contribution mapping factors of the employees, and the average social wages of the investment locations in the employment information through the preset insurance attribution judgment model if the employed person has multiple self-employments of the same registration type, so as to obtain the current normalized result, and determine the second target matrix based on the current normalized result and the second preset weight matrix; wherein the second target matrix includes weighted scores corresponding to the investment time, the labor contribution mapping factors of the employees, the average social wages of the investment locations, and the consistency between the registered residence of the employees and the investment locations in the employment information.

[0108] It is worth mentioning that the normalization method used for the feature processing of investment time in the embodiment of the present application is the Min-Max normalization method, and the feature vector obtained after processing is [a1, b1, ..., x1] T .

[0109] When the mapping factor of the employee labor contribution corresponding to Li Si is processed, the factor analysis method is used to sum up the number of employees of the individual business in each month of the most recent year. It is worth mentioning that the above method can also be used for sole proprietorships and partnerships, which will not be described one by one here. Through calculation, the value of Li Si's employee labor contribution mapping factor in each unit can be obtained, and the vector obtained after Min-Max normalization of each unit's employee labor contribution mapping factor is [a2, b2, ..., x2] T When processing the characteristics of the average social wage of the investment location, the vector [a3, b3, ..., x3] formed by Min-Max normalization of the average social wage of the investment location T When processing the feature of the consistency between the owner's registered residence and the investment location, it is necessary to take 1 if it is consistent and 0 if it is inconsistent, and obtain the feature vector [a4, b4, ..., x4] T .

[0110] Therefore, based on the above steps, the expression of the preset insured unit judgment model is as follows:

[0111]

[0112] It is understandable that the higher the average social wage level of the investment location, the higher the benefits of the insured. To some extent, the greater the mapping factor of the employee labor contribution, the stronger the operational stability. The importance of these two features is higher than the importance of the two features of investment time and consistency between the investment location and the registered permanent residence, so the weight adjustment method is as follows:

[0113] ω1=ω1(1-5%);

[0114] ω2=2ω2;

[0115] ω3=2ω3;

[0116] ω4=ω4*(1-5%);

[0117] For people who are eligible for insurance for flexible employment, the embodiment of the present application processes personal employment information through a preset insurance location judgment model to obtain the insurance location result. It is worth mentioning that the model rules corresponding to the preset insurance location judgment model are as follows: Fig.10 As shown:

[0118] First, when an employed person has only one registered self-employed investment without employees in the country, the registered place of the self-employed entity will be identified as the place where insurance should be participated.

[0119] Secondly, when an employed person has only one place in the country where he / she withholds and declares income from labor remuneration that meets the requirements, the place where he / she withholds and declares income shall be the place where he / she should participate in the insurance.

[0120] Furthermore, if an employed person has multiple self-employed investment registrations without employees or eligible labor remuneration income across the country, his / her only place to be insured will be determined through the preset insured place judgment model.

[0121] It is worth mentioning that compared with labor remuneration, self-employment is relatively more stable and more sustainable, so the embodiment of this application prioritizes self-employment over labor remuneration. Among them, self-employment includes three categories: self-employed businesses and sole proprietorships with only one owner, which have priority over partnerships with at least 2 partners; and self-employed businesses are superior to sole proprietorships.

[0122] In the first specific implementation, for the same employee who has multiple self-employed businesses with the same registration type, the embodiment of the present application assigns corresponding weights to the elements of "investment time", "location treatment (social average) level", and "consistency between the owner's registered place and investment place" based on expert experience, and selects the self-employed investment place with the highest score from the weight score as the place to be insured. In addition, the embodiment of the present application can optimize the model according to the actual feedback results, so that the model output results are more in line with the individual's willingness to participate in insurance.

[0123] In the second specific implementation, for the same employed person who has multiple eligible labor remuneration incomes but not in the same place, the factors of "cumulative income amount over a certain period of time", "income duration", "local treatment (social average) level", and "consistency between personal household registration and labor remuneration source" are initially weighted based on expert experience, and the labor remuneration source with the highest score is calculated as the place to be insured; in addition, the embodiment of the present application can optimize the model according to actual feedback results, so that the model output results are more in line with the individual's willingness to participate in insurance.

[0124] The priority order of each type is as follows:

[0125] The income of a self-employed business owner without employees is greater than that of a sole proprietor without employees, the income of a sole proprietor without employees is greater than that of a partner in a partnership without employees, and the income of a partner in a partnership without employees is greater than the qualified labor remuneration income.

[0126] It is worth mentioning that if the same employed person has multiple self-employed investment registrations of the same registration type, the location of the self-employed entity with the higher score shall be the corresponding insurance location of the employed person; the formula for determining the weighted score is as follows:

[0127] Weighted score = w1*investment time + w2*location treatment (social average) level + w3*consistency between the owner’s registered residence and the investment location;

[0128] If the same employed person has multiple qualified labor remuneration incomes and they are not in the same place, the labor remuneration source with the highest score shall be the corresponding insurance location of the employed person; the formula for determining the weighted score is as follows:

[0129] Weighted score = w1*accumulated income amount in a certain period + w2*duration of income + w3*level of treatment (social average) at the location + w4*consistency between individual household registration place and source of labor remuneration.

[0130] It is worth mentioning that the theoretical basis of the preset employment information identification model, the preset original statutory retirement age judgment model, the preset statutory retirement age judgment model after delayed retirement, the preset insurance qualification prediction model, and the preset insurance qualification judgment model are all rule-based methods. Among them, the rule-based method relies on predefined rules and logic for reasoning and decision-making, and all rules and logic are predefined. In addition, the behavior of the system is predictable and deterministic.

[0131] In this embodiment, for the preset insured employment unit judgment model and the preset insured location judgment model, after the model initially outputs the results, it is necessary to confirm the relevant subjects and receive feedback to form sample data, and then input the sample data into the model to train the model.

[0132] The insurance status judgment module 14 is used to compare the insurance attribution result with the preset insurance information, and judge whether the comparison result meets the preset conditions. If not, the insurance status of the employed person corresponding to the comparison result is judged as a state of being insured but not insured.

[0133] In this embodiment, the groups with employee insurance qualifications and flexible employment insurance qualifications need to undergo insurance comparison processes respectively, and the comparison process is as follows: all the insurance information collected in advance, that is, the individual's current insurance system and insurance unit are compared with the insurance qualifications and insurance employment units corresponding to the employed population output by the model, and the comparison results are divided into several types using a classification model, that is, whether the insurance system is correctly compared and whether the insurance unit is correctly compared, so as to screen out the groups with abnormal comparisons. It is worth mentioning that the embodiment of the present application needs to give priority to responding to key groups with abnormal comparisons, among which key groups include employed persons who are currently only insured for urban and rural residents' pension insurance and those who are not insured at all in the pension insurance system.

[0134] In the present application, an employment information acquisition module is used to process tax registration information and individual tax declaration data through a preset employment information recognition model to determine a target employer and a number of employees corresponding to the target employer; a prediction result acquisition module is used to process the employment information corresponding to the number of employees through a preset retirement age judgment model, so that a preset insurance qualification prediction model processes the retirement age output by the preset retirement age judgment model to obtain insurance qualification prediction results corresponding to the number of employment information respectively; an attribution result acquisition module is used to process the insurance qualification prediction result and the corresponding employment information through a preset insurance attribution judgment model to obtain the insurance attribution result corresponding to the employee; an insurance status judgment module is used to compare the insurance attribution result with the preset insurance information, and judge whether the comparison result meets the preset conditions. If not, the insurance status of the employee corresponding to the comparison result is judged as a state of being insured but not insured. It can be seen that this application processes the tax registration information and individual tax declaration data through a preset employment information identification model to determine the target employer and a number of employees corresponding to the target employer; and processes the employment information through a preset retirement age judgment model to obtain the retirement age, and processes the retirement age through a preset insurance qualification prediction model to obtain the insurance qualification prediction result; then, the employment information is processed through a preset insurance attribution judgment model to obtain the insurance attribution result; and then the insurance attribution result is compared with the preset insurance information to obtain the insurance status of the employed person. In this way, the insurance information of the employed person's pension insurance can be determined based on the employment information of the employed person, thereby solving and identifying the problem of employed persons excluded from the employee pension insurance, improving the efficiency of employee pension insurance collection and management and service quality, promoting the sustainable development of the pension insurance system, and promoting the high-quality development of my country's social security cause.

[0135] For further information, see Fig.11 As shown, the embodiment of the present invention discloses a pension insurance system adaptation reasoning method based on individual tax data, including:

[0136] Step S11: Process the tax registration information and individual tax declaration data through a preset employment information recognition model to determine a target employer and a number of employees corresponding to the target employer.

[0137] Step S12: The employment information corresponding to the plurality of employed persons is processed by a preset retirement age judgment model, so that the preset insurance qualification prediction model processes the retirement age output by the preset retirement age judgment model to obtain insurance qualification prediction results corresponding to the plurality of employment information respectively.

[0138] Step S13: Process the insurance qualification prediction result and the corresponding employment information through a preset insurance attribution judgment model to obtain an insurance attribution result corresponding to the employed person.

[0139] Step S14, compare the insurance attribution result with the preset insurance information, and determine whether the comparison result meets the preset conditions. If not, the insurance status of the employed person corresponding to the comparison result is determined as a state of being insured but not insured.

[0140] In this embodiment, the flowchart of determining the pension insurance participation status of employed persons based on employment information is as follows Fig.12 As shown:

[0141] First, the tax registration information is identified through the employment information identification model to obtain the target employers and employees. The tax information includes tax registration information, individual tax comprehensive income declaration data, and individual tax business income declaration data.

[0142] Secondly, after obtaining the target employers and employees, the embodiment of the present application needs to process the target employer and employee information through the retirement age judgment model and the insurance qualification prediction model to obtain the insurance qualification prediction result.

[0143] Then, when the insurance qualification prediction result corresponding to the employed person is obtained, the embodiment of the present application needs to process the insurance qualification prediction result through the insurance qualification judgment model to obtain the insurance qualification judgment result.

[0144] Subsequently, the embodiment of the present application needs to determine whether the insured qualification judgment result is the employee insurance qualification. If the insured qualification judgment result is the employee insurance qualification, the embodiment of the present application needs to process the employment information through the insured employment unit judgment model to obtain the insured unit result. If the insured qualification judgment result is not the employee insurance qualification, the insured qualification judgment result is the flexible employment insurance qualification, and the employment information is processed through the insured location judgment model to obtain the insured location result.

[0145] Furthermore, the embodiment of the present application compares the insured unit result, the insured eligibility judgment result and the collected insurance information to obtain a first comparison result corresponding to the employee insurance eligibility; and compares the insured location result, the insured eligibility judgment result and the collected insurance information to obtain a second comparison result corresponding to the flexible employment insurance eligibility.

[0146] Finally, the first comparison result and the second comparison result are classified respectively through the classification model, and the employed persons corresponding to the inconsistent comparison results are determined as target persons; among them, the target persons are employed persons who should be insured but whose current insurance status is not insured.

[0147] As can be seen from the above, before the application conducts the adaptation reasoning of the pension insurance system based on personal tax data, it is necessary to process the tax registration information and the personal tax declaration data through the preset employment information recognition model to determine the target employer and the several employees corresponding to the target employer; the employment information corresponding to the several employees is processed through the preset retirement age judgment model, so that the preset insurance qualification prediction model processes the retirement age output by the preset retirement age judgment model to obtain the insurance qualification prediction results corresponding to the several employment information respectively; the insurance qualification prediction results and the corresponding employment information are processed through the preset insurance attribution judgment model to obtain the insurance attribution results corresponding to the employees; the insurance attribution results are compared with the preset insurance information, and it is determined whether the comparison results meet the preset conditions. If not, the insurance status of the employees corresponding to the comparison results is determined as the state of being insured but not insured. It can be seen from this that the application processes the tax registration information and the personal tax declaration data through the preset employment information recognition model to determine the target employer and employees. Subsequently, the employment information is processed by a preset retirement age judgment model, so that the preset insurance qualification pre-judgment model processes the retirement age and obtains the insurance qualification pre-judgment result; then, the preset insurance attribution judgment model is used to process the information to obtain the insurance attribution result; finally, the insurance attribution result is compared with the preset insurance information, and it is determined whether the comparison result meets the preset conditions. If not, the insurance status of the employed person corresponding to the comparison result is determined as a state of being insured but not insured. In this way, the insurance information of the employed person's pension insurance can be determined based on the employment information of the employed person, thereby solving and identifying the problem of employed persons who are excluded from the employee pension insurance, and improving the efficiency of employee pension insurance collection and management and service quality.

[0148] Furthermore, the present application also discloses an electronic device. Fig.13 It is a structural diagram of an electronic device 20 shown according to an exemplary embodiment, and the content in the figure cannot be regarded as any limitation on the scope of use of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input and output interface 25 and a communication bus 26. Among them, the memory 22 is used to store a computer program, and the computer program is loaded and executed by the processor 21 to implement the relevant steps in the pension insurance system adaptation reasoning method based on personal tax data disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment can specifically be an electronic computer.

[0149] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0150] In addition, the memory 22, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0151] The operating system 221 is used to manage and control the hardware devices and computer program 222 on the electronic device 20, which can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program that can be used to complete the pension insurance system adaptation reasoning method based on individual tax data executed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include a computer program that can be used to complete other specific tasks.

[0152] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned disclosed method for reasoning about the adaptation of the pension insurance system based on individual tax data is implemented. The specific steps of the method can be referred to the corresponding contents disclosed in the aforementioned embodiments, and will not be repeated here.

[0153] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0154] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0155] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0156] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0157] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technicians in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A pension insurance system adaptation reasoning device based on individual tax data, characterized in that: include: An employment information acquisition module is used to process the tax registration information and individual tax declaration data through a preset employment information recognition model to determine a target employer and a number of employees corresponding to the target employer; A prediction result acquisition module is used to process the employment information corresponding to the plurality of employed persons through a preset retirement age judgment model, so that the preset insurance qualification prediction model processes the retirement age output by the preset retirement age judgment model to obtain insurance qualification prediction results corresponding to the plurality of employment information respectively; The attribution result acquisition module is used to process the insurance qualification pre-judgment result and the corresponding employment information through a preset insurance attribution judgment model to obtain the insurance attribution result corresponding to the employed person; The insurance status judgment module is used to compare the insurance attribution result with the preset insurance information, and judge whether the comparison result meets the preset conditions. If not, the insurance status of the employed person corresponding to the comparison result is judged as a state of being insured but not being insured.

2. The pension insurance system adaptation reasoning device based on individual tax data according to claim 1 is characterized in that: The prediction result acquisition module includes: A retiree determination unit, used to determine whether the employed person is a former statutory retiree; A first retirement age determination unit is used for, if the employed person is a former statutory retiree, processing the employment information corresponding to the employed person by using a preset former statutory retirement age judgment model to obtain a first retirement age corresponding to the employed person; A first prediction result determination unit is used to process the first retirement age and the employment information corresponding to the employed person through a preset insurance qualification prediction model to obtain a first insurance qualification prediction result corresponding to the employment information; A second retirement age determination unit is used for processing the employment information corresponding to the employed person through a preset model for determining the statutory retirement age after delayed retirement to obtain a second retirement age corresponding to the employed person if the employed person is not an original statutory retiree; The second prediction result determination unit is used to process the second retirement age and the employment information corresponding to the employed person using a preset insurance qualification prediction model to obtain a second insurance qualification prediction result corresponding to the employment information; wherein the first insurance qualification prediction result and the second insurance qualification prediction result are employee insurance qualifications or flexible employment insurance qualifications.

3. The pension insurance system adaptation reasoning device based on individual tax data according to claim 1 is characterized in that: The attribution result acquisition module includes: A qualification judgment unit is used to judge whether the qualification corresponding to the employed person is an employee qualification based on the pre-judgment result of the qualification; A first insurance unit determination unit is used for processing the employment information corresponding to the employed person through a preset insurance unit judgment model to obtain an insurance unit result if the insurance qualification corresponding to the employed person is an employee insurance qualification; The insurance location determination unit is used to process the employment information corresponding to the employed person through a preset insurance location judgment model to obtain an insurance location result if the insurance qualification corresponding to the employed person is not an employee insurance qualification.

4. The pension insurance system adaptation reasoning device based on individual tax data according to claim 1 is characterized in that: The attribution result acquisition module includes: A multiple-income source determination unit, used to determine whether the employment information corresponding to the employed person meets the preset multiple-income source condition; The unit score determination unit is used for, if the employed person meets the preset multiple-source income condition, using the preset insurance attribution judgment model and setting corresponding weights for the corresponding employment information based on the insurance qualification prediction result, and determining the unit scores corresponding to the several employment information based on the several weights and the corresponding employment information; wherein the insurance qualification prediction result is that the weight corresponding to the salary is greater than the weight corresponding to the owner of the self-employed business with employees, the weight corresponding to the owner of the self-employed business with employees is greater than the weight corresponding to the sole proprietorship investor with employees, and the weight corresponding to the sole proprietorship investor with employees is greater than the weight corresponding to the partner of a partnership with employees; A second insurance-eligible unit determination unit is used to extract a target score with the highest score from the scores of the units, and set the unit corresponding to the target score as the insurance-eligible unit; The third insurable unit determination unit is used to set the unit corresponding to the employee as an insurable unit if the employee does not meet the preset multiple income source conditions.

5. The pension insurance system adaptation reasoning device based on individual tax data according to claim 4 is characterized in that: The attribution result acquisition module includes: A multiple salary determination unit, used to determine whether the employee has multiple salary incomes if the employee meets the preset multiple income source conditions; The first target matrix determination unit is used to normalize the withholding and declared income amounts of several previous periods and several employment dates in the employment information respectively through a preset insurance attribution judgment model if the employed person has multiple salary incomes, so as to obtain the current normalization result, and determine the first target matrix based on the current naturalization result and the first preset weight matrix; wherein the first target matrix includes weighted scores corresponding to the withholding and declared income amounts of several previous periods and several employment dates in the employment information.

6. The pension insurance system adaptation reasoning device based on individual tax data according to claim 4 is characterized in that: The attribution result acquisition module includes: A multiple self-employment judgment unit, used to judge whether the employee has multiple self-employments with the same registration type if the employee meets the preset multiple income source conditions; The second target matrix determination unit is used to normalize the investment time, labor contribution mapping factors, average social wages of investment places, and consistency between the registered places of residence of several employees and the investment places in the employment information through a preset insurance participation judgment model if the employed person has multiple self-employed businesses with the same registration type, so as to obtain the current normalization result, and determine the second target matrix based on the current normalization result and a second preset weight matrix; wherein the second target matrix includes weighted scores corresponding to the investment time, labor contribution mapping factors, average social wages of investment places, and consistency between the registered places of residence of several employees and the investment places in the employment information.

7. The pension insurance system adaptation reasoning device based on individual tax data according to any one of claims 1 to 6, characterized in that: The attribution result acquisition module includes: The weight optimization unit is used to set corresponding optimized weights for the employment information through a preset insurance attribution judgment model and a preset weight adjustment strategy, so that the preset insurance attribution judgment model uses the optimized weights and the corresponding employment information to re-determine the optimized unit scores corresponding to several pieces of employment information.

8. A pension insurance system adaptation reasoning method based on individual tax data, characterized in that: include: Processing tax registration information and individual tax declaration data through a preset employment information recognition model to determine a target employer and a number of employees corresponding to the target employer; The employment information corresponding to the plurality of employed persons is processed by a preset retirement age judgment model, so that the preset insurance qualification prediction model processes the retirement age output by the preset retirement age judgment model to obtain insurance qualification prediction results corresponding to the plurality of employment information respectively; Processing the insurance qualification pre-judgment result and the corresponding employment information through a preset insurance attribution judgment model to obtain an insurance attribution result corresponding to the employed person; The insurance attribution result is compared with the preset insurance information, and it is determined whether the comparison result meets the preset conditions. If not, the insurance status of the employed person corresponding to the comparison result is determined as a state where he should be insured but is not insured.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor is used to execute the computer program to implement the steps of the pension insurance system adaptation reasoning method based on personal tax data as described in claim 8.

10. A computer-readable storage medium, characterized in that: Used to store a computer program, wherein when the computer program is executed by a processor, the steps of the pension insurance system adaptation reasoning method based on individual tax data as described in claim 8 are implemented.

Images