CN116049648B - Multiparty projection method and multiparty data analysis method based on data security - Google Patents
Multiparty projection method and multiparty data analysis method based on data security Download PDFInfo
- Publication number
- CN116049648B CN116049648B CN202211461454.XA CN202211461454A CN116049648B CN 116049648 B CN116049648 B CN 116049648B CN 202211461454 A CN202211461454 A CN 202211461454A CN 116049648 B CN116049648 B CN 116049648B
- Authority
- CN
- China
- Prior art keywords
- projection
- result
- dimension
- projection result
- multiparty
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/60—Protecting data
- G06F21/62—Protecting access to data via a platform, e.g. using keys or access control rules
- G06F21/6218—Protecting access to data via a platform, e.g. using keys or access control rules to a system of files or objects, e.g. local or distributed file system or database
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Bioethics (AREA)
- General Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a multiparty projection method and multiparty data analysis method based on data security, comprising the following steps: the multiparty client projects the high-dimensional data to obtain an initial projection result; the method comprises the steps that a server receives initial projection results from multi-party clients and adjusts the initial projection results by combining relationship features among the multi-party clients to obtain second projection results; the multiparty client verifies and adjusts the second projection effect based on the high-dimensional data to obtain a target projection result; the server receives and fuses the target projection results to obtain multiparty projection results, and the invention realizes accurate multiparty data projection and ensures projection effects under the condition of ensuring data security.
Description
Technical Field
The invention relates to the field of data processing, in particular to a multiparty projection method based on data security and a multiparty data analysis method.
Background
The software development is a process of building a software system or a software part in the system according to the requirements of a user, so that the software production data visualization technology is developed, auxiliary decision making is provided, production efficiency analysis and intelligent analysis on key indexes such as efficiency cost in the software development process are performed, and the effective rate of production data analysis is often influenced by the fact that the data dimension is too high. According to research, the dimension of the data is reduced before the production data is analyzed, and then the dimension reduced data is used for data classification or clustering, so that the problem of dimension disasters can be well avoided.
The high-dimensional data projection method is a commonly used data analysis method. It projects high-dimensional data into a low-dimensional space, thereby supporting the data analyst to analyze high-dimensional data features from the low-dimensional projection results.
The existing multiparty data projection method is characterized in that multiparty data are gathered in a server to carry out projection analysis of the data, all the data are required to be shared, a certain data safety problem exists, the other mode is that after the data are projected on a client, the data are transmitted to the server to carry out analysis of the data, and the method cannot carry out intercommunication among all the data, so that the data projection effect is poor and the characteristic of the data cannot be accurately reflected.
Disclosure of Invention
The invention provides a multiparty projection method and a multiparty data analysis method based on data security, which can realize accurate multiparty data projection, ensure projection effect and finally realize effective analysis of data under the condition of ensuring data security.
A multiparty projection method based on data security, comprising:
s1: the multiparty client projects the high-dimensional data to obtain an initial projection result;
s2: the method comprises the steps that a server receives initial projection results from multi-party clients and adjusts the initial projection results by combining relationship features among the multi-party clients to obtain second projection results;
s3: the multiparty client verifies and adjusts the second projection result based on the high-dimensional data to obtain a target projection result;
s4: and the server receives and fuses the target projection results to obtain multiparty projection results.
Preferably, in S1, the projecting the high-dimensional data by the multiparty client to obtain an initial projection result includes:
performing dimension analysis on the high-dimension data, determining the highest dimension, establishing a projection plane corresponding to each dimension based on the characteristics of each dimension, and normalizing the projection plane to obtain a standard projection plane;
and respectively projecting the high-dimension data onto the standard projection plane, and displaying the standard projection plane after the high-dimension data projection in the same coordinate system to obtain an initial projection result.
Preferably, in S2, the adjusting the initial projection result to obtain a second projection result in combination with the relationship feature between the multiparty clients includes:
based on the business characteristics of the multiparty client, a plurality of relation rules are established, and attribute information of the multiparty client is obtained; matching the multiparty clients with the relationship rules according to the attribute information and combining the relationship rules to obtain a plurality of groups of clients;
determining the relation characteristics of each group of clients according to the relation rule, and sequencing the plurality of groups of clients from less to more based on the number of each group of clients to obtain a combined sequence;
and based on the combined sequence and the relation characteristic, adjusting the initial projection result of each group of clients to obtain a second projection result.
Preferably, based on the combined sequence and the relational feature, the initial projection result of each group of clients is adjusted to obtain a second projection result, including:
extracting an initial projection result of each client in a first group of clients in the combined sequence, and acquiring a dimension principal component of high dimension from the initial projection result; judging whether the matching degree between the dimension principal component and the unit feature of the relation feature of the first group of clients meets the preset matching degree;
if yes, not adjusting the initial projection result;
otherwise, adjusting the dimension principal component based on the unit feature to obtain a new dimension principal component, and adjusting the initial projection result based on the new dimension principal component to obtain a first group of projection results;
and sequentially adjusting the initial projection results of the combined sequence according to the projection result adjustment mode of the first combined client to obtain a second projection result.
Preferably, according to a projection result adjustment manner of the first combined client, initial projection results of the combined sequence are sequentially adjusted to obtain a second projection result, including:
according to the projection result adjustment mode of the first combined client, adjusting the initial projection effect of the second group of clients in the combined sequence to obtain a second group of projection effects, and updating the projection result of the common client in the first group of clients and the second group of clients according to the second group of projection results;
according to the adjustment and update modes of the first group of projection results, the initial projection results of each group of clients in the combined sequence are adjusted and updated in sequence to obtain updated projection results, and the updated projection results are standardized to obtain second projection results.
Preferably, in S3, the multi-party client verifies and adjusts the second projection result based on the high-dimensional data to obtain a target projection result, including:
acquiring a first projection difference between the second projection result and the initial projection result, and judging whether the first projection difference is within a preset difference range;
if yes, not verifying the second projection result, and taking the second projection result as a target projection result;
otherwise, an integral projection model is built based on the second projection result, a local projection model is built based on the initial projection result, the integral projection model and the local projection model are respectively trained based on preset setting conditions to obtain a corresponding first projection model and a corresponding second projection model, and the second projection result is verified and adjusted based on a second projection difference between the first projection model and the second projection model to obtain a target projection result.
Preferably, based on a second projection difference between the first projection model and the second projection model, verifying and adjusting the second projection result to obtain a target projection result, including:
acquiring a second projection difference between the first projection model and a second projection model, and judging whether the second projection difference is within the preset difference range;
if yes, determining that the second projection result passes verification, and taking the second projection result as a target projection result;
otherwise, determining that the second projection result is not verified, and adjusting the second projection result based on the second projection difference to obtain a target projection result.
Preferably, adjusting the second projection result based on the second projection difference to obtain a target projection result includes:
performing difference analysis on the second projection difference, determining a projection dimension difference characteristic, determining a target projection dimension of the high-dimensional data based on the projection dimension difference characteristic, determining an optimal projection dimension range based on the data characteristic of the high-dimensional data, and judging whether the initial projection dimension is in the optimal projection dimension range;
if so, adjusting the projection dimension in the second projection result based on the target projection dimension to obtain a target projection result;
otherwise, based on the optimal projection dimension range, the target projection dimension is adjusted, and the projection dimension in the second projection result is adjusted by the adjusted target projection dimension, so that a target projection result is obtained.
Preferably, in S4, the server receives and fuses the target projection result to obtain a multiparty projection result, including:
determining the sorting weight of the target projection result based on the data characteristics of the target projection result;
determining the position of the target projection result in the multiparty projection result based on the sequencing weight;
determining a credibility coefficient of the target projection result based on the dimension characteristics of the target projection result;
and if the position overlapping part exists in the target projection result, selecting the target projection result with a larger credibility coefficient as a multi-direction projection result of the position overlapping part.
Preferably, the user in the server analyzes the received multiparty data based on the multiparty projection result.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.
The technical scheme of the invention is further described in detail through the drawings and the embodiments.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:
FIG. 1 is a flow chart of a multiparty projection method based on data security in an embodiment of the invention;
FIG. 2 is a flowchart of acquiring a second projection result according to an embodiment of the present invention;
fig. 3 is a flowchart of obtaining a target projection result in an embodiment of the present invention.
Detailed Description
The preferred embodiments of the present invention will be described below with reference to the accompanying drawings, it being understood that the preferred embodiments described herein are for illustration and explanation of the present invention only, and are not intended to limit the present invention.
Example 1
The embodiment of the invention provides a multiparty projection method based on data security, as shown in fig. 1, comprising the following steps:
s1: the multiparty client projects the high-dimensional data to obtain an initial projection result;
s2: the method comprises the steps that a server receives initial projection results from multi-party clients and adjusts the initial projection results by combining relationship features among the multi-party clients to obtain second projection results;
s3: the multiparty client verifies and adjusts the second projection result based on the high-dimensional data to obtain a target projection result;
s4: and the server receives and fuses the target projection results to obtain multiparty projection results.
In this embodiment, the relationship features are, for example, user relationship features, sub-business relationship features, time relationship features, and the like.
The beneficial effects of above-mentioned design scheme are: the method has the advantages that the high-dimensional data is not sent to the server, the projection result is directly regulated at the server, the privacy and safety of the data in the data projection process are guaranteed, the projection result is repeatedly regulated, the projection effect of the multi-party projection result is guaranteed, and effective analysis of the data is finally achieved.
Example 2
Based on embodiment 1, the embodiment of the invention provides a multiparty projection method based on data security, in S1, a multiparty client projects high-dimensional data to obtain an initial projection result, which comprises the following steps:
performing dimension analysis on the high-dimension data, determining the highest dimension, establishing a projection plane corresponding to each dimension based on the characteristics of each dimension, and normalizing the projection plane to obtain a standard projection plane;
and respectively projecting the high-dimension data onto the standard projection plane, and displaying the standard projection plane after the high-dimension data projection in the same coordinate system to obtain an initial projection result.
In this embodiment, the number of projection planes is equal to the highest number of dimensions.
In this embodiment, the unit vectors of the standard projection planes are the same, so that clear display of the data projection results is facilitated.
The beneficial effects of above-mentioned design scheme are: the initial projection result is obtained by establishing a projection plane corresponding to the high-dimensional data dimension and displaying the standard projection plane after the high-dimensional data is projected under the same coordinate system, so that the obtained spatial position relationship of the high-dimensional data in the dimension supply space can be ensured to provide a basis for multiparty projection.
Example 3
Based on embodiment 1, the embodiment of the present invention provides a multiparty projection method based on data security, as shown in fig. 2, in S2, the initial projection result is adjusted by combining relationship features between multiparty clients, so as to obtain a second projection result, which includes:
s201: based on the business characteristics of the multiparty client, a plurality of relation rules are established, and attribute information of the multiparty client is obtained; matching the multiparty clients with the relationship rules according to the attribute information and combining the relationship rules to obtain a plurality of groups of clients;
s202: determining the relation characteristics of each group of clients according to the relation rule, and sequencing the plurality of groups of clients from less to more based on the number of each group of clients to obtain a combined sequence;
s203: and based on the combined sequence and the relation characteristic, adjusting the initial projection result of each group of clients to obtain a second projection result.
In this embodiment, the business features of the multiparty client are, for example, financial business features, medical business features, sales business features, etc.
In this embodiment, the relationship rules are, for example, user relationship rules, sub-business relationship rules, temporal relationship rules, and the like.
In this embodiment, the attribute information of the multiparty client is, for example, location information, information interaction range, identity ID.
In this embodiment, each relationship rule corresponds to a set of clients, and there may be the same client in different relationship rules.
In this embodiment, adjusting the initial projection result of each group of clients to obtain a second projection result includes:
extracting an initial projection result of each client in a first group of clients in the combined sequence, and acquiring a dimension principal component of high dimension from the initial projection result; judging whether the matching degree between the dimension principal component and the unit feature of the relation feature of the first group of clients meets the preset matching degree;
if yes, not adjusting the initial projection result;
otherwise, adjusting the dimension principal component based on the unit feature to obtain a new dimension principal component, and adjusting the initial projection result based on the new dimension principal component to obtain a first group of projection results;
according to the projection result adjustment mode of the first combined client, adjusting the initial projection effect of the second group of clients in the combined sequence to obtain a second group of projection effects, and updating the projection result of the common client in the first group of clients and the second group of clients according to the second group of projection results;
according to the adjustment and update modes of the first group of projection results, the initial projection results of each group of clients in the combined sequence are adjusted and updated in sequence to obtain updated projection results, and the updated projection results are standardized to obtain second projection results.
The beneficial effects of above-mentioned design scheme are: the relation characteristics of the multiparty clients under each relation rule are determined according to the business characteristics and the attribute information of the multiparty clients, then the relation characteristics of the multiparty clients are clarified, the initial projection result is adjusted by utilizing the relation characteristics, the second projection result is ensured to be obtained, the relation characteristics of the multiparty clients can be represented while the high-dimensional data characteristics are represented, the satisfactory projection effect is brought, the server only receives the initial projection result, the initial projection result is adjusted, the high-dimensional data is directly obtained, and the safety of the high-dimensional data is ensured.
Example 4
Based on embodiment 3, the embodiment of the present invention provides a multiparty projection method based on data security, based on the combined sequence and the relationship feature, the initial projection result of each group of clients is adjusted to obtain a second projection result, which includes:
extracting an initial projection result of each client in a first group of clients in the combined sequence, and acquiring a dimension principal component of high dimension from the initial projection result; judging whether the matching degree between the dimension principal component and the unit feature of the relation feature of the first group of clients meets the preset matching degree;
if yes, not adjusting the initial projection result;
otherwise, adjusting the dimension principal component based on the unit feature to obtain a new dimension principal component, and adjusting the initial projection result based on the new dimension principal component to obtain a first group of projection results;
and sequentially adjusting the initial projection results of the combined sequence according to the projection result adjustment mode of the first combined client to obtain a second projection result.
In this embodiment, the dimension principal components may be, for example, user location, business name, time characteristics, and the like.
In this embodiment, the unit feature of the relationship feature may be, for example, a positional relationship feature, a business relationship feature, a time relationship feature, or the like.
In this embodiment, for example, the user position of the dimension principal component is based on the actual distance, and the unit feature is based on the transmission distance, and the initial projection result based on the actual distance needs to be adjusted to the projection result based on the transmission distance.
The beneficial effects of above-mentioned design scheme are: the method has the advantages that the method is characterized in that the matching degree between the main dimension component of the high-dimensional data obtained from the initial projection result and the unit feature of the relation feature is judged as a basis, the initial projection result is adjusted, the second projection result is ensured to be obtained, the relation feature among the multi-party clients can be represented while the high-dimensional data feature is represented, and a satisfactory projection effect is achieved.
Example 5
Based on embodiment 4, the embodiment of the invention provides a multiparty projection method based on data security, which sequentially adjusts initial projection results of a combined sequence according to a projection result adjustment mode of a first combined client to obtain a second projection result, comprising:
according to the projection result adjustment mode of the first combined client, adjusting the initial projection effect of the second group of clients in the combined sequence to obtain a second group of projection effects, and updating the projection result of the common client in the first group of clients and the second group of clients according to the second group of projection results;
according to the adjustment and update modes of the first group of projection results, the initial projection results of each group of clients in the combined sequence are adjusted and updated in sequence to obtain updated projection results, and the updated projection results are standardized to obtain second projection results.
In this embodiment, when the first set of projection results is updated according to the second set of projection results to include both the first set of projection results and the second set of projection results in the client, the first set of projection results is covered by the second set of projection results, and the second set of projection results is used as the reference.
In this embodiment, the combination sequence is arranged from less to more according to the number of clients, and the updated projection result obtained by updating uses the client group with the largest number of clients as the final projection result, so that the final second projection result is ensured to meet the requirements of most clients.
The beneficial effects of above-mentioned design scheme are: in the process of adjusting and updating the initial projection result, the updated projection result obtained by sequential updating of the combined sequence takes the client group with the largest number of clients as the final projection result, so that the finally obtained second projection result is ensured to meet the requirements of most clients, and the projection effect is ensured.
Example 6
Based on embodiment 1, the embodiment of the present invention provides a multiparty projection method based on data security, as shown in fig. 3, in S3, a multiparty client verifies and adjusts the second projection result based on the high-dimensional data to obtain a target projection result, which includes:
s301: acquiring a first projection difference between the second projection result and the initial projection result, and judging whether the first projection difference is within a preset difference range;
s302: if yes, not verifying the second projection result, and taking the second projection result as a target projection result;
s303: otherwise, an integral projection model is built based on the second projection result, a local projection model is built based on the initial projection result, the integral projection model and the local projection model are respectively trained based on preset setting conditions to obtain a corresponding first projection model and a corresponding second projection model, and the second projection result is verified and adjusted based on a second projection difference between the first projection model and the second projection model to obtain a target projection result.
In this embodiment, the overall projection model takes into account the characteristics of the multiparty client to project, and the local projection model projects only for the characteristics of the high-dimensional data of the client.
In this embodiment, the preset setting conditions include, for example, projection dimension number determination, projection data unit setting, and the like.
In this embodiment, the first projection model is a trained global projection model and the second projection model is a trained local projection model.
In this embodiment, the same preset setting conditions are used to train the integral projection model and the local projection model, so as to ensure that parameters of the first projection model and the second projection model are consistent, and provide a basis for verification and adjustment of the second projection result.
In this embodiment, verifying and adjusting the second projection result to obtain the target projection result includes:
acquiring a second projection difference between the first projection model and a second projection model, and judging whether the second projection difference is within the preset difference range;
if yes, determining that the second projection result passes verification, and taking the second projection result as a target projection result;
otherwise, determining that the second projection result is not verified, performing difference analysis on the second projection difference, determining a projection dimension difference feature, determining a target projection dimension of the high-dimensional data based on the projection dimension difference feature, determining an optimal projection dimension range based on the data feature of the high-dimensional data, and judging whether the initial projection dimension is in the optimal projection dimension range;
if so, adjusting the projection dimension in the second projection result based on the target projection dimension to obtain a target projection result;
otherwise, based on the optimal projection dimension range, the target projection dimension is adjusted, and the projection dimension in the second projection result is adjusted by the adjusted target projection dimension, so that a target projection result is obtained.
The beneficial effects of above-mentioned design scheme are: when the multiparty client acquires a second projection result regulated from the server, the original high-dimensional data is utilized to verify and regulate the second projection result, so that the obtained target projection result meets the requirements of the local client and the multiparty client, the projection effect is ensured, and the target projection result is obtained under the condition that the data security is ensured by transmitting and regulating the initial projection result between the server and the client.
Example 7
Based on embodiment 6, the embodiment of the invention provides a multiparty projection method based on data security, which verifies and adjusts a second projection result based on a second projection difference between the first projection model and the second projection model to obtain a target projection result, and comprises the following steps:
acquiring a second projection difference between the first projection model and a second projection model, and judging whether the second projection difference is within the preset difference range;
if yes, determining that the second projection result passes verification, and taking the second projection result as a target projection result;
otherwise, determining that the second projection result is not verified, and adjusting the second projection result based on the second projection difference to obtain a target projection result.
In this embodiment, when the second projection difference is within a preset difference range, it is further verified that the second projection result and the initial projection result are determined to be within the preset difference range.
The beneficial effects of above-mentioned design scheme are: and performing difference comparison after training the projection model, further verifying and determining that the second projection result and the initial projection result are in a preset difference range, ensuring the accuracy of difference acquisition between the determined second projection result and the initial projection result, and providing a basis for adjustment of the second projection result or not and an adjustment mode.
Example 8
Based on embodiment 7, the embodiment of the present invention provides a multiparty projection method based on data security, based on the second projection difference, the second projection result is adjusted to obtain a target projection result, which includes:
performing difference analysis on the second projection difference, determining a projection dimension difference characteristic, determining a target projection dimension of the high-dimensional data based on the projection dimension difference characteristic, determining an optimal projection dimension range based on the data characteristic of the high-dimensional data, and judging whether the initial projection dimension is in the optimal projection dimension range;
if so, adjusting the projection dimension in the second projection result based on the target projection dimension to obtain a target projection result;
otherwise, based on the optimal projection dimension range, the target projection dimension is adjusted, and the projection dimension in the second projection result is adjusted by the adjusted target projection dimension, so that a target projection result is obtained.
The beneficial effects of above-mentioned design scheme are: and performing difference analysis on the second projection difference to determine projection dimension difference characteristics, and adjusting the projection dimension of the target by combining the optimal projection dimension range determined by the high-dimensional data to ensure that the obtained projection result of the target meets the requirement of performing optimal projection display on the data characteristics of the high-dimensional data.
Example 9
Based on embodiment 1, the embodiment of the invention provides a multiparty projection method based on data security, in S4, a server receives and fuses the target projection result to obtain a multiparty projection result, which comprises the following steps:
determining the sorting weight of the target projection result based on the data characteristics of the target projection result;
the calculation formula of the sequencing weight is as follows:
wherein K is 0 Ranking weight representing the target projection result, N being the data set of the target projection result, N A Data set representing target projection results of all multiparty clients, M representing a data feature sum of the target projection results, T representing standard data feature values, T i A feature value representing the ith data feature, the value being (0, 1);
determining the position of the target projection result in the multiparty projection result based on the sequencing weight;
determining a credibility coefficient of the target projection result based on the dimension characteristics of the target projection result;
the calculation formula of the credibility coefficient of the target projection result is as follows:
wherein U is 0 A credibility coefficient representing the target projection result, R represents the dimension number of the target projection result, D j The dimension characteristic value of the j-th dimension in the target projection result is represented as (0, 1),representing average dimension characteristic values of target projection results of all multiparty clients in a j-th dimension;
and if the position overlapping part exists in the target projection result, selecting the target projection result with a larger credibility coefficient as a multi-direction projection result of the position overlapping part.
In this embodiment, the calculation of the ranking weight and the feasible coefficient is the effect of adding the data set, and the calculation of the corresponding ranking weight and the feasible coefficient has a larger effect on the target projection result with more data sets.
In this embodiment, the target projection results with higher ranking weights have a preferential position selection weight, and the position selection is better among the multiparty projection results.
The beneficial effects of above-mentioned design scheme are: by analyzing the data characteristics and the dimension characteristics of the target projection result, a reference is provided for determining the position of the target projection result in the multiparty projection result, the obtained target projection result is ensured to be displayed more reasonably, and the analysis of the data is facilitated.
Example 10
Based on embodiments 1 to 9, the present invention provides a multiparty data analysis method based on a multiparty projection method of data security, and a user in the server analyzes received multiparty data based on the multiparty projection result.
The beneficial effects of above-mentioned design scheme are: and analyzing the received multiparty data through the multiparty projection result, so that the data analysis efficiency is improved.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (6)
1. A data security-based multi-party projection method, comprising:
s1: the multiparty client projects the high-dimensional data to obtain an initial projection result;
s2: the method comprises the steps that a server receives initial projection results from multi-party clients and adjusts the initial projection results by combining relationship features among the multi-party clients to obtain second projection results;
s3: the multiparty client verifies and adjusts the second projection result based on the high-dimensional data to obtain a target projection result;
s4: the server receives and fuses the target projection results to obtain multiparty projection results;
s2, combining relation features among the multiparty clients, adjusting the initial projection result to obtain a second projection result, wherein the method comprises the following steps:
based on the business characteristics of the multiparty client, a plurality of relation rules are established, and attribute information of the multiparty client is obtained; matching the multiparty clients with the relationship rules according to the attribute information and combining the relationship rules to obtain a plurality of groups of clients;
determining the relation characteristics of each group of clients according to the relation rule, and sequencing the plurality of groups of clients from less to more based on the number of each group of clients to obtain a combined sequence;
based on the combined sequence and the relation characteristic, adjusting the initial projection result of each group of clients to obtain a second projection result;
based on the combined sequence and the relation feature, the initial projection result of each group of clients is adjusted to obtain a second projection result, which comprises the following steps:
extracting an initial projection result of each client in a first group of clients in the combined sequence, and acquiring a dimension principal component of high dimension from the initial projection result; judging whether the matching degree between the dimension principal component and the unit feature of the relation feature of the first group of clients meets the preset matching degree;
if yes, not adjusting the initial projection result;
otherwise, adjusting the dimension principal component based on the unit feature to obtain a new dimension principal component, and adjusting the initial projection result based on the new dimension principal component to obtain a first group of projection results;
sequentially adjusting initial projection results of the combined sequence according to a projection result adjustment mode of the first combined client to obtain a second projection result;
according to the projection result adjustment mode of the first combined client, sequentially adjusting the initial projection results of the combined sequence to obtain a second projection result, wherein the method comprises the following steps:
according to a projection result adjustment mode of the first combined client, adjusting initial projection results of a second group of clients in the combined sequence to obtain a second group of projection results, and updating the projection results of the first group of clients and the common clients in the second group of clients according to the second group of projection results;
according to the adjustment and update modes of the first group of projection results, sequentially adjusting and updating the initial projection results of each group of clients in the combined sequence to obtain updated projection results, and normalizing the updated projection results to obtain second projection results;
in S3, the multiparty client verifies and adjusts the second projection result based on the high-dimensional data to obtain a target projection result, including:
acquiring a first projection difference between the second projection result and the initial projection result, and judging whether the first projection difference is within a preset difference range;
if yes, not verifying the second projection result, and taking the second projection result as a target projection result;
otherwise, an integral projection model is built based on the second projection result, a local projection model is built based on the initial projection result, the integral projection model and the local projection model are respectively trained based on preset setting conditions to obtain a corresponding first projection model and a corresponding second projection model, and the second projection result is verified and adjusted based on a second projection difference between the first projection model and the second projection model to obtain a target projection result.
2. The multi-party projection method based on data security as claimed in claim 1, wherein in S1, the multi-party client projects the high-dimensional data to obtain an initial projection result, comprising:
performing dimension analysis on the high-dimension data, determining the highest dimension, establishing a projection plane corresponding to each dimension based on the characteristics of each dimension, and normalizing the projection plane to obtain a standard projection plane;
and respectively projecting the high-dimension data onto the standard projection plane, and displaying the standard projection plane after the high-dimension data projection in the same coordinate system to obtain an initial projection result.
3. The data security-based multi-party projection method according to claim 1, wherein verifying and adjusting the second projection result based on the second projection difference between the first projection model and the second projection model to obtain the target projection result comprises:
acquiring a second projection difference between the first projection model and a second projection model, and judging whether the second projection difference is within the preset difference range;
if yes, determining that the second projection result passes verification, and taking the second projection result as a target projection result;
otherwise, determining that the second projection result is not verified, and adjusting the second projection result based on the second projection difference to obtain a target projection result.
4. A data security-based multi-sided projection method according to claim 3, wherein adjusting the second projection result based on the second projection difference to obtain a target projection result comprises:
performing difference analysis on the second projection difference, determining a projection dimension difference characteristic, determining a target projection dimension of the high-dimensional data based on the projection dimension difference characteristic, determining an optimal projection dimension range based on the data characteristic of the high-dimensional data, and judging whether an initial projection dimension is within the optimal projection dimension range;
if so, adjusting the projection dimension in the second projection result based on the target projection dimension to obtain a target projection result;
otherwise, based on the optimal projection dimension range, the target projection dimension is adjusted, and the projection dimension in the second projection result is adjusted by the adjusted target projection dimension, so that a target projection result is obtained.
5. The data security-based multi-party projection method according to claim 1, wherein in S4, the server receives and fuses the target projection result to obtain a multi-party projection result, and the method comprises the following steps:
determining the sorting weight of the target projection result based on the data characteristics of the target projection result;
determining the position of the target projection result in the multiparty projection result based on the sequencing weight;
determining a credibility coefficient of the target projection result based on the dimension characteristics of the target projection result;
and if the position overlapping part exists in the target projection result, selecting the target projection result with a larger credibility coefficient as a multi-direction projection result of the position overlapping part.
6. A multiparty data analysis method comprising a multiparty data-security-based projection method according to any one of claims 1-5, wherein a user in said server analyzes received multiparty data based on said multiparty projection result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211461454.XA CN116049648B (en) | 2022-11-17 | 2022-11-17 | Multiparty projection method and multiparty data analysis method based on data security |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211461454.XA CN116049648B (en) | 2022-11-17 | 2022-11-17 | Multiparty projection method and multiparty data analysis method based on data security |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116049648A CN116049648A (en) | 2023-05-02 |
| CN116049648B true CN116049648B (en) | 2023-08-04 |
Family
ID=86124409
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211461454.XA Active CN116049648B (en) | 2022-11-17 | 2022-11-17 | Multiparty projection method and multiparty data analysis method based on data security |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116049648B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102707917A (en) * | 2012-05-23 | 2012-10-03 | 中国科学院对地观测与数字地球科学中心 | Method and device for visualizing high-dimensional data |
| CN105160352A (en) * | 2015-08-18 | 2015-12-16 | 中南大学 | High-dimensional data subspace clustering projection effect optimization method based on dimension reconstitution |
| CN113158022A (en) * | 2021-01-29 | 2021-07-23 | 北京达佳互联信息技术有限公司 | Service recommendation method, device, server and storage medium |
| CN114329300A (en) * | 2022-03-14 | 2022-04-12 | 中南大学 | Multi-party projection method based on data security and multi-party production data analysis method |
| CN115080801A (en) * | 2022-07-22 | 2022-09-20 | 山东大学 | Cross-modal retrieval method and system based on federal learning and data binary representation |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8812274B2 (en) * | 2009-04-24 | 2014-08-19 | Hermant Virkar | Methods for mapping data into lower dimensions |
| WO2015167526A1 (en) * | 2014-04-30 | 2015-11-05 | Hewlett-Packard Development Company, L.P | Facilitating interpretation of high-dimensional data clusters |
-
2022
- 2022-11-17 CN CN202211461454.XA patent/CN116049648B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102707917A (en) * | 2012-05-23 | 2012-10-03 | 中国科学院对地观测与数字地球科学中心 | Method and device for visualizing high-dimensional data |
| CN105160352A (en) * | 2015-08-18 | 2015-12-16 | 中南大学 | High-dimensional data subspace clustering projection effect optimization method based on dimension reconstitution |
| CN113158022A (en) * | 2021-01-29 | 2021-07-23 | 北京达佳互联信息技术有限公司 | Service recommendation method, device, server and storage medium |
| CN114329300A (en) * | 2022-03-14 | 2022-04-12 | 中南大学 | Multi-party projection method based on data security and multi-party production data analysis method |
| CN115080801A (en) * | 2022-07-22 | 2022-09-20 | 山东大学 | Cross-modal retrieval method and system based on federal learning and data binary representation |
Non-Patent Citations (1)
| Title |
|---|
| 基于局部差分隐私的高维关联数据发布研究;刘芸菲;中国优秀硕士学位论文全文数据库 信息科技辑;第2-4章 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116049648A (en) | 2023-05-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11487831B2 (en) | Compatibility scoring of users | |
| US20100185580A1 (en) | Compatibility scoring of users in a social network | |
| US20190051390A1 (en) | Team-based tele-diagnostics blockchain-enabled system | |
| US20200034164A1 (en) | System for utilizing one or more data sources to generate a customized interface | |
| CN110020022B (en) | Data processing method, device, equipment and readable storage medium | |
| US20210157842A1 (en) | Information providing system and information providing method | |
| CN113190562A (en) | Report generation method and device and electronic equipment | |
| CN116049648B (en) | Multiparty projection method and multiparty data analysis method based on data security | |
| CN112527897B (en) | Data processing method and system | |
| CN113139191B (en) | Statistical method for bug disposal repair priority | |
| CN118052529A (en) | Future job seeker occupation planning system, method, equipment and storage medium | |
| US20020032690A1 (en) | Method for processing data and generating test report of test system | |
| CN117575706A (en) | Information pushing method and device, nonvolatile storage medium and electronic equipment | |
| CN116186150A (en) | Mobile user data visualization system and method | |
| CN111694884B (en) | Intelligent government affair request processing method based on big data | |
| CN112237015B (en) | Methods, systems, and computer readable media for communicating information about an entity | |
| WO2019089056A1 (en) | Evaluating ip location on a client device | |
| US20200167326A1 (en) | System and method for acting on potentially incomplete data | |
| US8069079B1 (en) | Co-location opportunity evaluation | |
| CN113220758A (en) | Project maturity analysis demonstration method | |
| Liu et al. | Two‐tuple linguistic utility aggregation operator and its applications to group decision‐making | |
| CN112861900A (en) | Image system and detection method | |
| CN116627784B (en) | Image monitoring analysis system and method for personalized data analysis | |
| CN114401427B (en) | Streaming media data transmission system and method for cloud game | |
| CN114817216B (en) | Perceived data credibility assessment method and system based on equipment fingerprint identification |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |