CN113887741B

CN113887741B - Data generation method, device, equipment and storage medium based on federal learning

Info

Publication number: CN113887741B
Application number: CN202111304096.7A
Authority: CN
Inventors: 周才军; 王志辉; 杨振燕; 曾依峰; 雷庆璋; 罗燕武; 宁海亮; 樊鹏辉
Original assignee: Shenzhen Digital Certificate Authority Center Co ltd
Current assignee: Shenzhen Digital Certificate Authority Center Co ltd
Priority date: 2021-11-05
Filing date: 2021-11-05
Publication date: 2022-09-30
Anticipated expiration: 2041-11-05
Also published as: CN113887741A

Abstract

The invention relates to the field of artificial intelligence, and discloses a data generation method, a data generation device, data generation equipment and a data generation storage medium based on federal learning, which are used for improving the safety of vehicle service data of each organization. The data generation method based on the federal learning comprises the following steps: acquiring a plurality of organization service basic data and a plurality of corresponding organization service interest rate data; obtaining a model training instruction, and performing homomorphic encryption and calculation according to the model training instruction, a plurality of organization business basic data and a plurality of organization business interest rate data to generate a plurality of target homomorphic encryption business data; combining the model training instruction, performing model training and model convergence, and generating a target vehicle service federal learning model; and acquiring basic data of the vehicle to be identified, inputting the basic data of the vehicle to be identified into a target vehicle business federal learning model for calculation, and generating target vehicle business data. In addition, the invention also relates to a block chain technology, and a plurality of mechanism service basic data can be stored in the block chain.

Description

Data generation method, device, equipment and storage medium based on federal learning

Technical Field

The invention relates to the field of machine learning, in particular to a data generation method, a data generation device, data generation equipment and a storage medium based on federal learning.

Background

With the development of life, vehicles have become indispensable commuting tools in life, and with the popularization of vehicles, various problems have arisen. For example, traffic accidents or natural disasters cause damage to vehicles, and owners gradually attach importance to property protection conditions after vehicles are damaged, so that various mechanisms provide a lot of strategies for property protection against vehicle damage, the strategies need to be formulated by combining with relevant service data of vehicles to be protected, and the relevant service data of the vehicles need to be acquired during formulation.

In the prior art, vehicle-related service data and historical service data of each organization are called for model training under the condition of a block chain distributed account book, so that the safety of the vehicle service data of each organization is low, and property guarantee strategies matched with vehicles cannot be obtained.

Disclosure of Invention

The invention provides a data generation method, a data generation device, data generation equipment and a data generation storage medium based on federal learning, which are used for improving the safety of vehicle service data of each organization.

The invention provides a data generation method based on federal learning in a first aspect, which comprises the following steps: acquiring a plurality of organization service basic data and a plurality of corresponding organization service interest rate data, wherein the organization service basic data and the organization service interest rate data correspond to each other one by one; obtaining a model training instruction, and performing homomorphic encryption and calculation according to the model training instruction, the plurality of organization business basic data and the plurality of organization business interest rate data to generate a plurality of target homomorphic encrypted business data; calling the plurality of organization business basic encrypted data, the plurality of organization business interest rate encrypted data and the plurality of target homomorphic encrypted business data by combining the model training instruction to carry out model training and model convergence, and generating a target vehicle business federal learning model; and acquiring basic data of the vehicle to be identified, inputting the basic data of the vehicle to be identified into the target vehicle business federal learning model for calculation, and generating the target vehicle business data.

Optionally, in a first implementation manner of the first aspect of the present invention, the obtaining multiple pieces of organization service basic data and multiple corresponding pieces of organization service interest rate data, where the one-to-one correspondence between the organization service basic data and the organization service interest rate data includes: acquiring a plurality of user service basic data and a plurality of corresponding user service interest rate data, wherein the user service basic data correspond to the user service interest rate data one to one; and reading a corresponding mechanism identification from each user service basic data, transmitting the corresponding user service basic data and the corresponding user service interest rate data to the corresponding service mechanism according to each mechanism identification, generating mechanism service basic data corresponding to each service mechanism and corresponding mechanism service interest rate data, and obtaining a plurality of mechanism service basic data and a plurality of corresponding mechanism service interest rate data.

Optionally, in a second implementation manner of the first aspect of the present invention, the obtaining a model training instruction, and performing homomorphic encryption and calculation according to the model training instruction, the multiple organization business basic data, and the multiple organization business interest rate data, and generating multiple target homomorphic encrypted business data includes: obtaining a model training instruction, analyzing the model training instruction, and generating a target mechanism identifier; the plurality of mechanism business basic data and the plurality of mechanism business interest rate data are authenticated according to the target mechanism identification, a plurality of mechanism business basic data to be encrypted and a plurality of mechanism business interest rate data to be encrypted are generated, and the mechanism business basic data to be encrypted and the mechanism business interest rate data to be encrypted correspond one to one; and calling a homomorphic encryption algorithm to homomorphic encrypt each mechanism service basic data to be encrypted and the corresponding mechanism service interest rate data to be encrypted, generating target homomorphic encrypted service data corresponding to each initial mechanism service basic data, and obtaining a plurality of target homomorphic encrypted service data.

Optionally, in a third implementation manner of the first aspect of the present invention, the authenticating, according to the target organization identifier, the multiple organization service basic data and the multiple organization service interest rate data to generate multiple organization service basic data to be encrypted and multiple organization service interest rate data to be encrypted, where the one-to-one correspondence between the organization service basic data to be encrypted and the organization service interest rate data to be encrypted includes: acquiring a corresponding digital signature based on the target organization identification and the basic data of each organization business to obtain a plurality of organization business digital signatures; judging whether each organization business digital signature is a null value; and if the target organization service digital signature is not a null value, determining organization service basic data and corresponding organization service interest rate data corresponding to the target organization service digital signature as service basic data to be encrypted and organization service interest rate data to be encrypted respectively.

Optionally, in a fourth implementation manner of the first aspect of the present invention, the invoking a homomorphic encryption algorithm to homomorphic encrypt each mechanism service basic data to be encrypted and corresponding mechanism service interest rate data to be encrypted, to generate target homomorphic encrypted service data corresponding to each initial mechanism service basic data, and obtaining a plurality of target homomorphic encrypted service data includes: calling a preset homomorphic encryption algorithm and a data public key to homomorphic encrypt each mechanism service basic data to be encrypted and the corresponding mechanism service interest rate data to be encrypted, and generating mechanism service basic encrypted data corresponding to each mechanism service basic data to be encrypted and corresponding mechanism service interest rate encrypted data to obtain a plurality of mechanism service basic encrypted data and a plurality of corresponding mechanism service interest rate encrypted data; and calling a preset calculation function to perform service data calculation on the basic encrypted data of each mechanism service and the corresponding mechanism service interest rate encrypted data, generating target homomorphic encrypted service data corresponding to the basic encrypted data of each mechanism service, and obtaining a plurality of target homomorphic encrypted service data.

Optionally, in a fifth implementation manner of the first aspect of the present invention, the invoking, in combination with the model training instruction, the multiple organization business basic encrypted data, the multiple organization business interest rate encrypted data, and the multiple target homomorphic encrypted business data to perform model training and model convergence, and generating the target vehicle business federation learning model includes: performing model training on the plurality of organization business basic encrypted data, the plurality of organization business interest rate encrypted data and the plurality of homomorphic encrypted business data by combining the model training instruction to generate an initial vehicle business federal learning model; acquiring a homomorphic encrypted service test data set, and inputting the homomorphic encrypted service test data set into the initial vehicle service federal learning model to generate a homomorphic encrypted service test probability set; and carrying out model convergence on the initial vehicle service federation learning model based on a preset constraint function and the homomorphic encryption service test probability set to generate a target vehicle service federation learning model.

Optionally, in a sixth implementation manner of the first aspect of the present invention, after the obtaining of the vehicle basic data to be recognized, inputting the vehicle basic data to be recognized into the federal learning model of the target vehicle service for calculation, and generating the target vehicle service data, the data generating method based on federal learning includes: and carrying out homomorphic encryption on the target vehicle service data to obtain homomorphic encrypted service data of the target vehicle, and transmitting the homomorphic encrypted service data of the target vehicle to a third party terminal.

A second aspect of the present invention provides a data generation apparatus based on federal learning, including: the system comprises a data acquisition module, a data processing module and a data processing module, wherein the data acquisition module is used for acquiring a plurality of organization service basic data and a plurality of corresponding organization service interest rate data, and the organization service basic data and the organization service interest rate data correspond to each other one by one; the encrypted data generation module is used for acquiring a model training instruction, and performing homomorphic encryption and calculation according to the model training instruction, the plurality of institution business basic data and the plurality of institution business interest rate data to generate a plurality of target homomorphic encrypted business data; the learning model generation module is used for calling the plurality of organization business basic encrypted data, the plurality of organization business interest rate encrypted data and the plurality of target homomorphic encrypted business data in combination with the model training instruction to carry out model training and model convergence, and generating a target vehicle business federation learning model; and the business data generation module is used for acquiring basic data of the vehicle to be identified, inputting the basic data of the vehicle to be identified into the target vehicle business federal learning model for calculation, and generating the target vehicle business data.

Optionally, in a first implementation manner of the second aspect of the present invention, the data obtaining module may be further specifically configured to: acquiring a plurality of user service basic data and a plurality of corresponding user service interest rate data, wherein the user service basic data correspond to the user service interest rate data one to one; and reading a corresponding mechanism identification from each user service basic data, transmitting the corresponding user service basic data and the corresponding user service interest rate data to the corresponding service mechanism according to each mechanism identification, generating mechanism service basic data corresponding to each service mechanism and corresponding mechanism service interest rate data, and obtaining a plurality of mechanism service basic data and a plurality of corresponding mechanism service interest rate data.

Optionally, in a second implementation manner of the second aspect of the present invention, the encrypted data generating module includes: the instruction acquisition unit is used for acquiring a model training instruction, analyzing the model training instruction and generating a target mechanism identifier; the data authentication unit is used for authenticating the plurality of mechanism business basic data and the plurality of mechanism business interest rate data according to the target mechanism identification to generate a plurality of mechanism business basic data to be encrypted and a plurality of mechanism business interest rate data to be encrypted, wherein the mechanism business basic data to be encrypted corresponds to the mechanism business interest rate data to be encrypted one by one; and the homomorphic encryption unit is used for calling a homomorphic encryption algorithm to homomorphic encrypt each mechanism service basic data to be encrypted and the corresponding mechanism service interest rate data to be encrypted, generating target homomorphic encrypted service data corresponding to each initial mechanism service basic data and obtaining a plurality of target homomorphic encrypted service data.

Optionally, in a third implementation manner of the second aspect of the present invention, the data authentication unit may be further specifically configured to: acquiring a corresponding digital signature based on the target organization identification and the basic data of each organization business to obtain a plurality of organization business digital signatures; judging whether each organization business digital signature is a null value; and if the target organization service digital signature is not a null value, determining organization service basic data and corresponding organization service interest rate data corresponding to the target organization service digital signature as service basic data to be encrypted and organization service interest rate data to be encrypted respectively.

Optionally, in a fourth implementation manner of the second aspect of the present invention, the homomorphic encryption unit may be further specifically configured to: calling a preset homomorphic encryption algorithm and a data public key to homomorphic encrypt each mechanism service basic data to be encrypted and the corresponding mechanism service interest rate data to be encrypted, and generating mechanism service basic encrypted data corresponding to each mechanism service basic data to be encrypted and corresponding mechanism service interest rate encrypted data to obtain a plurality of mechanism service basic encrypted data and a plurality of corresponding mechanism service interest rate encrypted data; and calling a preset calculation function to perform service data calculation on the basic encrypted data of each mechanism service and the corresponding mechanism service interest rate encrypted data, generating target homomorphic encrypted service data corresponding to the basic encrypted data of each mechanism service, and obtaining a plurality of target homomorphic encrypted service data.

Optionally, in a fifth implementation manner of the second aspect of the present invention, the learning model generation module may be further specifically configured to: performing model training on the plurality of organization business basic encrypted data, the plurality of organization business interest rate encrypted data and the plurality of homomorphic encrypted business data by combining the model training instruction to generate an initial vehicle business federal learning model; acquiring a homomorphic encrypted service test data set, and inputting the homomorphic encrypted service test data set into the initial vehicle service federal learning model to generate a homomorphic encrypted service test probability set; and carrying out model convergence on the initial vehicle service federation learning model based on a preset constraint function and the homomorphic encryption service test probability set to generate a target vehicle service federation learning model.

Optionally, in a sixth implementation manner of the second aspect of the present invention, the data generating apparatus based on federal learning further includes: and the transmission module is used for homomorphic encryption of the target vehicle service data to obtain homomorphic encrypted service data of the target vehicle and transmitting the homomorphic encrypted service data of the target vehicle to a third party terminal.

A third aspect of the present invention provides a data generation device based on federal learning, including: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the federal learning based data generation device to execute the federal learning based data generation method described above.

A fourth aspect of the present invention provides a computer-readable storage medium having stored therein instructions, which, when run on a computer, cause the computer to execute the above-described federal learning based data generation method.

In the technical scheme provided by the invention, a plurality of organization business basic data and a plurality of corresponding organization business interest rate data are obtained, and the organization business basic data and the organization business interest rate data are in one-to-one correspondence; obtaining a model training instruction, and performing homomorphic encryption and calculation according to the model training instruction, the plurality of organization business basic data and the plurality of organization business interest rate data to generate a plurality of target homomorphic encrypted business data; calling the plurality of organization business basic encrypted data, the plurality of organization business interest rate encrypted data and the plurality of target homomorphic encrypted business data by combining the model training instruction to carry out model training and model convergence, and generating a target vehicle business federal learning model; and acquiring basic data of the vehicle to be identified, inputting the basic data of the vehicle to be identified into the target vehicle business federal learning model for calculation, and generating the target vehicle business data. In the embodiment of the invention, a plurality of organization business basic data and a plurality of corresponding organization business interest rate data are authenticated, homomorphic encrypted and calculated to generate a plurality of target homomorphic encrypted business data, model training and convergence are carried out based on the data to generate a target vehicle business federal learning model, and finally, the vehicle basic data to be identified are input to generate the target vehicle business data; the certification of each organization is obtained, and model training is carried out after homomorphic encryption is carried out on the basic service data and the interest rate data of each organization, so that the safety of the vehicle service data of each organization is improved.

Drawings

FIG. 1 is a schematic diagram of an embodiment of a data generation method based on federated learning according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of another embodiment of the data generation method based on federated learning according to the embodiment of the present invention;

FIG. 3 is a schematic diagram of an embodiment of a data generation apparatus based on federated learning according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of another embodiment of a data generation device based on federated learning in the embodiment of the present invention;

fig. 5 is a schematic diagram of an embodiment of a data generation device based on federal learning in an embodiment of the present invention.

Detailed Description

The embodiment of the invention provides a data generation method, a data generation device, data generation equipment and a storage medium based on federal learning, which are used for improving the safety of vehicle service data of each organization.

The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

For convenience of understanding, a specific flow of the embodiment of the present invention is described below, and referring to fig. 1, an embodiment of the data generation method based on federal learning in the embodiment of the present invention includes:

101. acquiring a plurality of organization service basic data and a plurality of corresponding organization service interest rate data, wherein the organization service basic data correspond to the organization service interest rate data one by one;

the server acquires a plurality of organization service basic data and a plurality of corresponding organization service interest rate data, wherein one organization service basic data corresponds to one organization service interest rate data. It is emphasized that, in order to further ensure the privacy and security of the enterprise business basic data, the enterprise business basic data can also be stored in a node of a block chain.

In this embodiment, the organization business basic data is organization vehicle insurance policy data, and the organization vehicle insurance policy data includes vehicle basic information, vehicle owner driving information, driving area environment information, and the like, where the vehicle basic information includes vehicle usage, vehicle type, vehicle historical accident information, and the like, the vehicle owner driving information includes vehicle owner identity information, vehicle owner driving age information, and the like, and the vehicle area environment information refers to environment information of a route with a high vehicle driving frequency. The agency vehicle warranty data is also used to indicate the agency to which the vehicle warranty data pertains, such as agency A vehicle warranty data, indicating that the vehicle warranty data belongs to agency A. The agency service interest rate data is agency vehicle insurance policy rate data, is used for displaying rate conditions of vehicle insurance policy and is also used for indicating an agency affiliated to the vehicle insurance policy rate data.

It is to be understood that the executing entity of the present invention may be a data generating apparatus based on federal learning, and may also be a terminal or a server, which is not limited herein. The embodiment of the present invention is described by taking a server as an execution subject. The server may be an independent server, or may be a cloud server that provides basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, a middleware service, a domain name service, a security service, a Content Delivery Network (CDN), and a big data and artificial intelligence platform.

102. Obtaining a model training instruction, and performing homomorphic encryption and calculation according to the model training instruction, a plurality of organization business basic data and a plurality of organization business interest rate data to generate a plurality of target homomorphic encryption business data;

the server obtains the model training instruction, and homomorphic encryption and calculation are carried out on the basis of the model training instruction, the plurality of organization business basic data and the corresponding plurality of organization business interest rate data to generate a plurality of target homomorphic encryption business data.

The model training instruction is used for indicating which mechanism business basic data and mechanism business interest rate data corresponding to the mechanism business basic data are used, therefore, the server determines a plurality of mechanism business basic data matched with the model training instruction and a plurality of mechanism business interest rate data matched with the model training instruction in a plurality of mechanism business basic data and a plurality of mechanism business interest rate data corresponding to the mechanism business basic data according to the model training instruction, then homomorphic encryption is carried out on the mechanism business basic data and business data calculation is carried out on the basis of the homomorphic encryption, and a plurality of target homomorphic encrypted business data are generated. In this embodiment, the target homomorphic encrypted service data is target vehicle homomorphic encrypted premium data.

103. Calling a plurality of organization business basic encrypted data, a plurality of organization business interest rate encrypted data and a plurality of target homomorphic encrypted business data by combining with a model training instruction to carry out model training and model convergence, and generating a target vehicle business federal learning model;

and the server performs model training and model convergence by combining the model training instruction to generate a target vehicle service federal learning model.

After a model training instruction, a plurality of organization business basic encrypted data, a plurality of organization business interest rate encrypted data and a plurality of target homomorphic encrypted business data are obtained, the server combines the plurality of organization business basic encrypted data, the plurality of organization business interest rate encrypted data and the plurality of target homomorphic encrypted business data according to the model training instruction to perform model preliminary training to obtain an initial learning model, and then model convergence is performed on the initial learning model to obtain a target vehicle business federation learning model.

104. And acquiring basic data of the vehicle to be identified, inputting the basic data of the vehicle to be identified into a target vehicle business federal learning model for calculation, and generating target vehicle business data.

After acquiring the basic data of the vehicle to be identified, the server inputs the basic data of the vehicle to be identified into a vehicle business federal learning model for convolution calculation, and generates target vehicle business data.

The server can input basic data of the vehicle to be identified corresponding to the vehicle service data needing to be calculated into a target vehicle service federal learning model for calculation, so that the target vehicle service data are generated, the basic data of the vehicle to be identified comprise vehicle use conditions, vehicle types, vehicle historical accident information and the like, the basic data of the vehicle to be identified are input into the vehicle service federal learning model for calculation, the target vehicle service data are generated, and the target vehicle service data comprise but are not limited to premium data of the target vehicle and interest rate data of the target vehicle.

In the embodiment of the invention, a plurality of organization business basic data and a plurality of corresponding organization business interest rate data are authenticated, homomorphic encrypted and calculated to generate a plurality of target homomorphic encrypted business data, model training and convergence are carried out based on the data to generate a target vehicle business federal learning model, and finally the vehicle basic data to be identified are input to generate the target vehicle business data; the certification of each organization is obtained, and model training is carried out after homomorphic encryption is carried out on the basic service data and the interest rate data of each organization, so that the safety of the vehicle service data of each organization is improved.

Referring to fig. 2, another embodiment of the data generation method based on federal learning according to the embodiment of the present invention includes:

201. acquiring a plurality of organization service basic data and a plurality of corresponding organization service interest rate data, wherein the organization service basic data and the organization service interest rate data correspond to each other one by one;

In this embodiment, the organization business basic data is organization vehicle insurance policy data, and the organization vehicle insurance policy data includes vehicle basic information, vehicle owner driving information, driving area environment information, and the like, where the vehicle basic information includes vehicle usage, vehicle type, vehicle historical accident information, and the like, the vehicle owner driving information includes vehicle owner identity information, vehicle owner driving age information, and the like, and the vehicle area environment information refers to environment information of a route with a high vehicle driving frequency. The agency vehicle warranty data is also used to indicate the agency to which the vehicle warranty data pertains, such as agency A vehicle warranty data, indicating that the vehicle warranty data belongs to agency A. The agency service interest rate data is agency vehicle policy rate data, is used for displaying the rate condition of the vehicle policy and is also used for indicating an agency affiliated to the vehicle policy rate data.

Specifically, the server acquires a plurality of user service basic data and a plurality of corresponding user service interest rate data, wherein the user service basic data and the user service interest rate data correspond to each other one by one; the server reads the corresponding mechanism identification from each user service basic data, transmits the corresponding user service basic data and the corresponding user service interest rate data to the corresponding service mechanism according to each mechanism identification, generates mechanism service basic data corresponding to each service mechanism and corresponding mechanism service interest rate data, and obtains a plurality of mechanism service basic data and a plurality of corresponding mechanism service interest rate data.

The organization business basic data and the corresponding organization business interest rate data are explained in an organization angle, before that, the organization business basic data is obtained through the user business basic data, and the corresponding organization business interest rate data is obtained through the user business interest rate data. The server firstly obtains a plurality of user service basic data and a plurality of user service interest rate data, wherein one user service basic data corresponds to one user service interest rate data. After acquiring a plurality of user service basic data and a plurality of corresponding user service interest rate data, the server reads a corresponding mechanism identifier from each user service basic data, then transmits the corresponding user service basic data and the corresponding user service interest rate data to a corresponding service mechanism according to each mechanism identifier, generates mechanism service basic data and corresponding mechanism service interest rate data corresponding to each service mechanism, and obtains a plurality of mechanism service basic data and a plurality of corresponding mechanism service interest rate data.

202. Obtaining a model training instruction, and performing homomorphic encryption and calculation according to the model training instruction, a plurality of organization business basic data and a plurality of organization business interest rate data to generate a plurality of target homomorphic encryption business data;

the server acquires the model training instruction, and then homomorphic encryption and calculation are carried out on the basis of the model training instruction, the plurality of organization business basic data and the plurality of corresponding organization business interest rate data to generate a plurality of target homomorphic encryption business data.

Specifically, the server acquires a model training instruction, analyzes the model training instruction and generates a target mechanism identifier; the server authenticates the plurality of organization service basic data and the plurality of organization service interest rate data according to the target organization identification to generate a plurality of organization service basic data to be encrypted and a plurality of organization service interest rate data to be encrypted, wherein the organization service basic data to be encrypted and the organization service interest rate data to be encrypted are in one-to-one correspondence; and the server calls a homomorphic encryption algorithm to homomorphic encrypt each mechanism service basic data to be encrypted and the corresponding mechanism service interest rate data to be encrypted, and generates target homomorphic encrypted service data corresponding to each initial mechanism service basic data to obtain a plurality of target homomorphic encrypted service data.

The method comprises the steps that a server firstly obtains a model training instruction, possibly obtains a target mechanism identification through analyzing the model training instruction, the target mechanism identification points to a mechanism to which data belongs, for example, data of a target mechanism identification A pointing to a mechanism B and data of a mechanism C, then the server authenticates in a plurality of mechanism service data and a plurality of mechanism service interest rate data based on the target mechanism identification A, and the mechanism service data authenticated by the target mechanism identification A and the corresponding mechanism service interest rate data are determined as mechanism service basic data to be encrypted and mechanism service interest rate data to be encrypted. And then, a homomorphic encryption algorithm is called to encrypt the mechanism service basic data to be encrypted and the mechanism service interest rate data to be encrypted to generate target homomorphic encrypted service data, and the same homomorphic encryption operation is also carried out on other mechanism service basic data to be encrypted and the corresponding mechanism service interest rate data to be encrypted, so that a plurality of target homomorphic encrypted service data are generated.

The server authenticates the plurality of mechanism service basic data and the plurality of mechanism service interest rate data according to the target mechanism identification to generate a plurality of mechanism service basic data to be encrypted and a plurality of mechanism service interest rate data to be encrypted, wherein the one-to-one correspondence between the mechanism service basic data to be encrypted and the mechanism service interest rate data to be encrypted comprises the following steps:

the server acquires a corresponding digital signature based on the target organization identification and the basic data of each organization business to obtain a plurality of organization business digital signatures; the server judges whether each organization service digital signature is a null value; and if the target organization service digital signature is not a null value, the server respectively determines the organization service basic data corresponding to the target organization service digital signature and the corresponding organization service interest rate data as the organization service basic data to be encrypted and the organization service interest rate data to be encrypted.

The server transmits the target mechanism identification to each service mechanism, each service mechanism receives the target mechanism identification and sends a digital signature corresponding to mechanism service basic data of the mechanism to the server, the server receives the digital signature corresponding to the mechanism service basic data to obtain a plurality of mechanism service basic data signatures, whether each mechanism service basic data signature is a null value or not is judged, if the mechanism service basic data signature is the null value, the service mechanism corresponding to the mechanism service digital signature does not open data authorization to the server, and at the moment, the corresponding mechanism service basic data and the corresponding mechanism service interest rate data cannot be used. If the organization business basic data is not null value and comprises specific data, the organization business digital signature opens data authorization to the server, and at the moment, the organization business digital signature is determined as the organization business basic data to be encrypted and the organization business interest rate data to be encrypted.

The server calls a homomorphic encryption algorithm to homomorphic encrypt each mechanism service basic data to be encrypted and the corresponding mechanism service interest rate data to be encrypted, target homomorphic encrypted service data corresponding to each initial mechanism service basic data are generated, and the step of obtaining a plurality of target homomorphic encrypted service data comprises the following steps:

calling a preset homomorphic encryption algorithm and a data public key to homomorphic encrypt each mechanism business basic data to be encrypted and the corresponding mechanism business interest rate data to be encrypted, and generating mechanism business basic encrypted data corresponding to each mechanism business basic data to be encrypted and corresponding mechanism business interest rate encrypted data to obtain a plurality of mechanism business basic encrypted data and a plurality of corresponding mechanism business interest rate encrypted data; and calling a preset calculation function to calculate the business data of each organization business basic encrypted data and the corresponding organization business interest rate encrypted data, generating target homomorphic encrypted business data corresponding to each organization business basic encrypted data, and obtaining a plurality of target homomorphic encrypted business data.

The server calls a preset homomorphic encryption algorithm and a data public key to homomorphically encrypt each mechanism business basic data to be encrypted and the corresponding mechanism business interest rate data to be encrypted, and generates mechanism business basic encrypted data corresponding to each mechanism business basic data to be encrypted and corresponding mechanism business interest rate encrypted data to obtain a plurality of mechanism business basic encrypted data and a plurality of corresponding mechanism business interest rate encrypted data; and then the server calls a preset calculation function to calculate the business data of each organization business basic encrypted data and the corresponding organization business interest rate encrypted data, and generates target homomorphic encrypted business data corresponding to each organization business basic encrypted data to obtain a plurality of target homomorphic encrypted business data.

It should be noted that the server may invoke the data private key to decrypt, and the data private key may only decrypt the plurality of target homomorphic encrypted service data, and may not decrypt the plurality of organization service basic encrypted data and the corresponding plurality of organization service interest rate encrypted data.

203. Calling a plurality of organization business basic encrypted data, a plurality of organization business interest rate encrypted data and a plurality of target homomorphic encrypted business data by combining with a model training instruction to carry out model training and model convergence, and generating a target vehicle business federal learning model;

Specifically, the server performs model training on a plurality of organization business basic encrypted data, a plurality of organization business interest rate encrypted data and a plurality of homomorphic encrypted business data by combining with a model training instruction to generate an initial vehicle business federal learning model; the method comprises the steps that a server obtains a homomorphic encryption service test data set, the homomorphic encryption service test data set is input into an initial vehicle service federal learning model, and a homomorphic encryption service test probability set is generated; and the server performs model convergence on the initial vehicle service federal learning model based on a preset constraint function and a homomorphic encryption service test probability set to generate a target vehicle service federal learning model.

The server combines the model training instruction to perform model training after obtaining a plurality of homomorphic encrypted service data to generate an initial vehicle service federal learning model, and in order to ensure the accuracy of the federal learning model, the server obtains homomorphic encrypted service test data sets to converge the initial vehicle service federal learning model. The server inputs the homomorphic encrypted service test data set into an initial vehicle service federal learning model to generate a homomorphic encrypted service test probability set, the homomorphic encrypted service test probability set comprises a plurality of homomorphic encrypted service test probabilities, and homomorphic encrypted service test data in the homomorphic encrypted service test data set correspond to the homomorphic encrypted service test probabilities in the homomorphic encrypted service test probability set one by one. The server calls a preset constraint function to perform model convergence based on the homomorphic encryption service test probability set, wherein the preset constraint function is as follows: and l V _ FED-V _ SUM | < delta, wherein V _ FED is homomorphic encryption service test probability, V _ SUM is preset service test probability, and delta is a constraint threshold value. When the homomorphic encryption service test probability cannot meet the preset constraint function, the accuracy of the model is not expected, a model adjusting instruction is obtained at the moment, the initial vehicle service federal learning model is adjusted according to the model adjusting instruction, and the target vehicle service federal learning model is generated.

204. Acquiring basic data of a vehicle to be identified, inputting the basic data of the vehicle to be identified into a target vehicle business federal learning model for calculation, and generating target vehicle business data;

205. And carrying out homomorphic encryption on the target vehicle service data to obtain homomorphic encrypted service data of the target vehicle, and transmitting the homomorphic encrypted service data of the target vehicle to a third party terminal.

After the target vehicle business data are generated, the server carries out homomorphic encryption on the target vehicle business data, and then transmits the homomorphic encrypted business data of the target vehicle to a third party terminal, wherein the third party terminal can be an enterprise terminal or a government terminal, and the homomorphic encrypted business data of the target vehicle provide reference for enterprise operation upgrading and also provide reference for government decision upgrading.

In the above description of the data generating method based on federal learning in the embodiment of the present invention, the following description of the data generating device based on federal learning in the embodiment of the present invention refers to fig. 3, and an embodiment of the data generating device based on federal learning in the embodiment of the present invention includes:

the data acquisition module 301 is configured to acquire a plurality of organization service basic data and a plurality of corresponding organization service interest rate data, where the organization service basic data and the organization service interest rate data correspond to each other one by one;

the encrypted data generating module 302 is configured to obtain a model training instruction, perform homomorphic encryption and calculation according to the model training instruction, the plurality of organization business basic data, and the plurality of organization business interest rate data, and generate a plurality of target homomorphic encrypted business data;

the learning model generation module 303 is configured to invoke the multiple organization business basic encrypted data, the multiple organization business interest rate encrypted data, and the multiple target homomorphic encrypted business data in combination with the model training instruction to perform model training and model convergence, so as to generate a target vehicle business federation learning model;

and the service data generation module 304 is configured to obtain basic data of the vehicle to be identified, input the basic data of the vehicle to be identified into the target vehicle service federal learning model, and calculate the basic data to generate service data of the target vehicle.

Referring to fig. 4, another embodiment of the data generating apparatus based on federal learning according to the embodiment of the present invention includes:

Optionally, the data obtaining module 301 may be further specifically configured to:

acquiring a plurality of user service basic data and a plurality of corresponding user service interest rate data, wherein the user service basic data and the user service interest rate data correspond to each other one by one;

and reading a corresponding mechanism identification from each user service basic data, transmitting the corresponding user service basic data and the corresponding user service interest rate data to the corresponding service mechanism according to each mechanism identification, generating mechanism service basic data corresponding to each service mechanism and corresponding mechanism service interest rate data, and obtaining a plurality of mechanism service basic data and a plurality of corresponding mechanism service interest rate data.

Optionally, the encrypted data generating module 302 includes:

the instruction obtaining unit 3021 is configured to obtain a model training instruction, analyze the model training instruction, and generate a target mechanism identifier;

a data authentication unit 3022, configured to authenticate the multiple organization service basic data and the multiple organization service interest rate data according to the target organization identifier, and generate multiple organization service basic data to be encrypted and multiple organization service interest rate data to be encrypted, where the organization service basic data to be encrypted and the organization service interest rate data to be encrypted correspond to each other one by one;

the homomorphic encryption unit 3023 is configured to invoke a homomorphic encryption algorithm to perform homomorphic encryption on each mechanism to be encrypted service basic data and corresponding mechanism to be encrypted service interest rate data, generate target homomorphic encrypted service data corresponding to each initial mechanism service basic data, and obtain a plurality of target homomorphic encrypted service data.

Optionally, the data authentication unit 3022 may be further specifically configured to:

acquiring a corresponding digital signature based on the target organization identification and the basic data of each organization business to obtain a plurality of organization business digital signatures;

judging whether each organization business digital signature is a null value;

and if the target organization service digital signature is not a null value, determining organization service basic data and corresponding organization service interest rate data corresponding to the target organization service digital signature as service basic data to be encrypted and organization service interest rate data to be encrypted respectively.

Optionally, the homomorphic encryption unit 3023 may be further specifically configured to:

calling a preset homomorphic encryption algorithm and a data public key to homomorphic encrypt each mechanism business basic data to be encrypted and the corresponding mechanism business interest rate data to be encrypted, and generating mechanism business basic encrypted data corresponding to each mechanism business basic data to be encrypted and corresponding mechanism business interest rate encrypted data to obtain a plurality of mechanism business basic encrypted data and a plurality of corresponding mechanism business interest rate encrypted data;

and calling a preset calculation function to calculate the business data of each organization business basic encrypted data and the corresponding organization business interest rate encrypted data, generating target homomorphic encrypted business data corresponding to each organization business basic encrypted data, and obtaining a plurality of target homomorphic encrypted business data.

Optionally, the learning model generating module 303 may be further specifically configured to:

performing model training on the plurality of organization business basic encrypted data, the plurality of organization business interest rate encrypted data and the plurality of homomorphic encrypted business data by combining the model training instruction to generate an initial vehicle business federal learning model;

acquiring a homomorphic encrypted service test data set, and inputting the homomorphic encrypted service test data set into the initial vehicle service federal learning model to generate a homomorphic encrypted service test probability set;

and carrying out model convergence on the initial vehicle service federal learning model based on a preset constraint function and the homomorphic encryption service test probability set to generate a target vehicle service federal learning model.

Optionally, the data generating apparatus based on federal learning further includes:

the transmission module 305 is configured to perform homomorphic encryption on the target vehicle service data to obtain target vehicle homomorphic encrypted service data, and transmit the target vehicle homomorphic encrypted service data to a third party terminal.

Fig. 3 and 4 describe the data generating device based on federal learning in the embodiment of the present invention in detail from the perspective of a modular functional entity, and the data generating device based on federal learning in the embodiment of the present invention is described in detail from the perspective of hardware processing.

Fig. 5 is a schematic structural diagram of a federal learning based data generation device 500 according to an embodiment of the present invention, where the federal learning based data generation device 500 may have relatively large differences due to different configurations or performances, and may include one or more processors (CPUs) 510 (e.g., one or more processors) and a memory 520, and one or more storage media 530 (e.g., one or more mass storage devices) for storing applications 533 or data 532. Memory 520 and storage media 530 may be, among other things, transient or persistent storage. The program stored on the storage medium 530 may include one or more modules (not shown), each of which may include a sequence of instructions operating on the federal learning based data generation apparatus 500. Still further, the processor 510 may be configured to communicate with the storage medium 530 to execute a series of instruction operations in the storage medium 530 on the federal learning based data generation apparatus 500.

The federal learning based data generation facility 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input-output interfaces 560, and/or one or more operating systems 531, such as Windows server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will appreciate that the federated learning-based data generation facility architecture depicted in FIG. 5 does not constitute a limitation on federated learning-based data generation facilities, and may include more or fewer components than those depicted, or some components in combination, or a different arrangement of components.

The invention also provides a data generation device based on federal learning, which comprises a memory and a processor, wherein computer readable instructions are stored in the memory, and when the computer readable instructions are executed by the processor, the processor executes the steps of the data generation method based on federal learning in the embodiments.

The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, which may also be a volatile computer-readable storage medium, having stored therein instructions, which, when run on a computer, cause the computer to perform the steps of the federated learning-based data generation method.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention, which is substantially or partly contributed by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A data generation method based on federal learning is characterized by comprising the following steps:

acquiring a plurality of organization service basic data and a plurality of corresponding organization service interest rate data, wherein the organization service basic data and the organization service interest rate data correspond to each other one by one; the system comprises a service management system, a service management system and a service management system, wherein the service basic data of an organization are vehicle insurance policy data of the organization, and the service interest rate data of the organization are vehicle insurance policy rate data of the organization;

obtaining a model training instruction, and performing homomorphic encryption and calculation according to the model training instruction, the plurality of organization business basic data and the plurality of organization business interest rate data to generate a plurality of target homomorphic encrypted business data;

calling a plurality of organization business basic encrypted data, a plurality of organization business interest rate encrypted data and a plurality of target homomorphic encrypted business data by combining the model training instruction to carry out model training and model convergence, and generating a target vehicle business federal learning model;

acquiring basic data of a vehicle to be identified, inputting the basic data of the vehicle to be identified into the target vehicle business federal learning model for calculation, and generating target vehicle business data;

the obtaining of the model training instruction, performing homomorphic encryption and calculation according to the model training instruction, the plurality of organization business basic data and the plurality of organization business interest rate data, and generating a plurality of target homomorphic encrypted business data includes: according to the model training instruction, determining a plurality of organization business basic data matched with the model training instruction and a plurality of organization business interest rate data matched with the model training instruction in the plurality of organization business basic data and the plurality of organization business interest rate data corresponding to the plurality of organization business basic data, carrying out homomorphic encryption on the organization business interest rate data and carrying out business data calculation on the basis of homomorphic encryption to generate a plurality of target homomorphic encrypted business data.

2. The method according to claim 1, wherein the obtaining a plurality of agency service basic data and a plurality of corresponding agency service interest rate data, the one-to-one correspondence between the agency service basic data and the agency service interest rate data includes:

3. The method of claim 1, wherein the obtaining of the model training command, homomorphic encryption and calculation according to the model training command, the plurality of agency business basic data and the plurality of agency business interest rate data, and generating the plurality of target homomorphic encrypted business data comprises:

acquiring a model training instruction, analyzing the model training instruction, and generating a target mechanism identifier;

the plurality of organization business basic data and the plurality of organization business interest rate data are authenticated according to the target organization identification, a plurality of organization business basic data to be encrypted and a plurality of organization business interest rate data to be encrypted are generated, and the organization business basic data to be encrypted and the organization business interest rate data to be encrypted are in one-to-one correspondence;

and calling a homomorphic encryption algorithm to homomorphic encrypt each mechanism service basic data to be encrypted and the corresponding mechanism service interest rate data to be encrypted, generating target homomorphic encrypted service data corresponding to each initial mechanism service basic data, and obtaining a plurality of target homomorphic encrypted service data.

4. The method according to claim 3, wherein the authenticating the plurality of agency service basic data and the plurality of agency service interest rate data according to the target agency identifier generates a plurality of agency service basic data to be encrypted and a plurality of agency service interest rate data to be encrypted, and the one-to-one correspondence between the agency service basic data to be encrypted and the agency service interest rate data to be encrypted includes:

judging whether each organization business digital signature is a null value;

5. The method for generating data based on federal learning as claimed in claim 4, wherein the step of invoking a homomorphic encryption algorithm to homomorphic encrypt each mechanism service basic data to be encrypted and the corresponding mechanism service interest rate data to be encrypted, and generating target homomorphic encrypted service data corresponding to each initial mechanism service basic data to obtain a plurality of target homomorphic encrypted service data comprises:

6. The method according to claim 1, wherein the invoking of the plurality of agency business basic encrypted data, the plurality of agency business interest rate encrypted data, and the plurality of target homomorphic encrypted business data in combination with the model training instruction for model training and model convergence, generating the target vehicle business federal learning model comprises:

performing model training on the plurality of organization business basic encrypted data, the plurality of organization business interest rate encrypted data and the plurality of target homomorphic encrypted business data by combining the model training instruction to generate an initial vehicle business federal learning model;

7. The federal learning-based data generation method as claimed in any one of claims 1 to 6, wherein after acquiring basic data of a vehicle to be recognized, inputting the basic data of the vehicle to be recognized into the federal learning model of a target vehicle service for calculation, and generating target vehicle service data, the federal learning-based data generation method comprises:

and carrying out homomorphic encryption on the target vehicle service data to obtain homomorphic encrypted service data of the target vehicle, and transmitting the homomorphic encrypted service data of the target vehicle to a third party terminal.

8. A federal learning-based data generation apparatus, characterized in that the federal learning-based data generation apparatus comprises:

the system comprises a data acquisition module, a data processing module and a data processing module, wherein the data acquisition module is used for acquiring a plurality of organization service basic data and a plurality of corresponding organization service interest rate data, and the organization service basic data and the organization service interest rate data correspond to each other one by one; the system comprises a service management system, a service management system and a service management system, wherein the service basic data of an organization are vehicle insurance policy data of the organization, and the service interest rate data of the organization are vehicle insurance policy rate data of the organization;

the encrypted data generation module is used for acquiring a model training instruction, and performing homomorphic encryption and calculation according to the model training instruction, the plurality of institution business basic data and the plurality of institution business interest rate data to generate a plurality of target homomorphic encrypted business data;

the learning model generation module is used for calling a plurality of organization business basic encrypted data, a plurality of organization business interest rate encrypted data and a plurality of target homomorphic encrypted business data by combining the model training instruction to carry out model training and model convergence, and generating a target vehicle business federation learning model;

the service data generation module is used for acquiring basic data of the vehicle to be identified, inputting the basic data of the vehicle to be identified into the target vehicle service federal learning model for calculation, and generating service data of the target vehicle;

the encrypted data generation module is further used for determining a plurality of organization business basic data matched with the model training instruction and a plurality of organization business interest rate data matched with the model training instruction in the plurality of organization business basic data and the plurality of organization business interest rate data corresponding to the plurality of organization business basic data according to the model training instruction, performing homomorphic encryption on the organization business interest rate data and performing business data calculation on the basis of homomorphic encryption to generate the plurality of target homomorphic encrypted business data.

9. A federal learning-based data generation apparatus, characterized in that the federal learning-based data generation apparatus includes: a memory and at least one processor, the memory having instructions stored therein;

the at least one processor invokes the instructions in the memory to cause the federated learning-based data generation facility to perform the federated learning-based data generation method of any one of claims 1-7.

10. A computer readable storage medium having instructions stored thereon, wherein the instructions, when executed by a processor, implement the federal learning based data generation method as claimed in any of claims 1-7.