Privacy protection collaborative analysis method for commodity supply chain platform based on multi-party computation

Abstract

The application relates to the technical field of data processing protection, in particular to a commodity supply chain platform privacy protection collaborative analysis method based on multi-party calculation, which comprises the following steps: acquiring information data uploaded by each participant in each data update period, acquiring feature coefficients of dynamic comprehensive changes of each participant in each period, obtaining feature values of dynamic changes of sensitivities of various information data by combining differences of the information data in each period compared with other periods, calculating first adjustment coefficients and second adjustment coefficients of differential privacy protection adjustment of the various information data, then acquiring proportion coefficients of various information data of each participant for privacy budget allocation, and acquiring privacy budgets of various information data of each participant in combination with a preset upper limit of the privacy budget, so as to perform differential privacy protection on the information data uploaded by each participant. The application can improve the accuracy of multi-party collaborative calculation and analysis.

Description

Technical Field

[0001] This application relates to the field of data processing and protection technology, specifically to a collaborative analysis method for privacy protection of commodity supply chain platforms based on multi-party computation. Background Technology

[0002] In the commodity supply chain system, multiple stakeholders, including suppliers, manufacturers, and retailers, possess core business data such as transaction data, inventory information, and logistics tracking. Privacy-preserving collaborative analysis methods based on multi-party computation can achieve data value sharing and mining while ensuring data privacy is not compromised. This provides support for intelligent decision-making in the supply chain and is of great significance for promoting the digital transformation of the supply chain and fostering collaborative industrial development.

[0003] However, in existing privacy protection schemes based on differential privacy, traditional methods often adopt static budget allocation strategies, which do not fully consider the dynamic changes in data and the differences in the sensitivity of data among different participants. This results in some highly sensitive data not being adequately protected, while low-sensitivity data consumes too much budget, leading to a waste of resources. Furthermore, due to the lack of comprehensive analysis of data distribution characteristics, data volume, and update frequency, the availability of data after adding noise is reduced, affecting the accuracy and efficiency of collaborative analysis, making it difficult to achieve an optimal balance between privacy protection and data value utilization. Summary of the Invention

[0004] To address the aforementioned technical issues, this application provides a collaborative analysis method for privacy protection in a commodity supply chain platform based on multi-party computation, thereby resolving the existing problems.

[0005] The privacy-preserving collaborative analysis method for commodity supply chain platforms based on multi-party computation in this application adopts the following technical solution:

[0006] One embodiment of this application provides a privacy-preserving collaborative analysis method for a commodity supply chain platform based on multi-party computation, including the following steps:

[0007] Obtain the information data uploaded by each participant in each data update cycle and convert it into structured data format;

[0008] For each participant's uploaded information data, the characteristic coefficient of each participant's dynamic comprehensive change in each period is obtained by observing the differences in the information data of each participant in different periods. Combined with the differences in information data in each period compared with other periods, the characteristic value of the dynamic change of sensitivity of various information data is obtained.

[0009] By using the feature values ​​of the dynamic changes in the sensitivity of all information data of each participant, the participants are clustered. Based on the average level of the feature values ​​of the dynamic changes in the sensitivity of all participants with respect to various information data in each cluster, and the amount of each type of information data of each participant in the cluster, the first adjustment coefficient and the second adjustment coefficient of differential privacy protection adjustment of various information data are obtained respectively, and then the proportion coefficient of privacy budget allocation of various information data of each participant is obtained.

[0010] Based on the proportion coefficient of privacy budget allocation according to the various information data of each participant, and combined with the preset privacy budget upper limit, the privacy budget of various information data of each participant is obtained in order to carry out differential privacy protection on the information data uploaded by each participant.

[0011] Preferably, the method for obtaining the characteristic coefficients of each participant's dynamic comprehensive change in each period is as follows:

[0012] By analyzing the differences in the same information data of each participant in each period and other periods, the first feature value of the dynamic update and change of various information data of each participant in each period is obtained. This is used to extract the feature sequence of each participant in each period. The mean of the Euclidean distance of each participant in each period and other periods with respect to the feature sequence is used as the feature coefficient of each participant in the dynamic comprehensive change in each period.

[0013] Preferably, the method for obtaining the first feature value further includes: calculating the average Hamming distance between the participants in each period and other periods regarding the same information data, and using this average as the first feature value of the dynamic update and change of the same information data in each period.

[0014] Preferably, the method for extracting the feature sequence of each participant for each period further includes: using the sequence of first feature values ​​that dynamically update and change all information data of each participant within each period as the feature sequence of each participant for each period.

[0015] Preferably, the method for calculating the characteristic values ​​of the dynamic changes in the sensitivity of the various information data is as follows:

[0016] ;in, Indicates the first The first type of participant The characteristic values ​​of the sensitivity of information data change dynamically; Indicates the first The first type of participant This type of information data in the first The first characteristic value is dynamically updated and changes within each cycle; Indicates the first Various participants in the first The characteristic coefficients of the dynamic comprehensive change of each cycle; Indicates the first The sum of the characteristic coefficients of the dynamic comprehensive changes of all participants across all cycles; Indicates the first The number of cycles for each participating party.

[0017] Preferably, the sequence of feature values ​​of the dynamic changes in sensitivity of all information data uploaded by each participant is used as the information feature sequence of each participant. In the process of clustering the participants, the DTW distance between the information feature sequences of different participants is used as the metric distance between different participants.

[0018] Preferably, the mean of the normalized feature values ​​of the same information data of all participants in the statistical cluster is used as the first adjustment coefficient for differential privacy protection adjustment of the same information data.

[0019] Preferably, the normalized result of the quantity of each type of information data of each participant in the statistical cluster is used as the second adjustment coefficient for differential privacy protection adjustment of each type of information data of each participant.

[0020] Preferably, the method for calculating the proportion coefficients for allocating the privacy budget to the various information data of each participating party is as follows:

[0021] ;in, Indicates the first The first type of participant The proportion of information data allocated to the privacy budget; and They represent the first species and first The first adjustment coefficient for differential privacy protection adjustment of information data; and They represent the first The first type of participant species and first The second adjustment coefficient for differential privacy protection of information data; Indicates the first The number of types of information data of various participants.

[0022] Preferably, the method for calculating the privacy budget of various information data of the participating parties is as follows: ,in, Indicates the first The first type of participant Privacy budget for information data; A privacy budget cap; Indicates the first The first type of participant The proportion of information data allocated to the privacy budget.

[0023] This application has at least the following beneficial effects:

[0024] This application addresses the significant limitations of traditional budget allocation methods in privacy-preserving collaborative analysis within a commodity supply chain platform. These methods neglect the dynamic changes in information data over time, leading to a substantial decrease in data usability after the addition of differential privacy noise, thus severely impacting the accuracy and efficiency of multi-party collaborative analysis. Therefore, this application first conducts in-depth analysis of the information data uploaded by each participant in the platform at different time periods. By quantifying the differences in dynamic changes within different update cycles, it accurately extracts the dynamic update characteristics of each participant's data sensitivity.

[0025] Simultaneously, clustering analysis is used to classify data based on the differences in the dynamic changes of multidimensional data from participating parties. By comprehensively comparing and analyzing the sensitivity characteristics of different clustered categories, a refined allocation of the differential privacy protection budget for each participant's data is achieved. The beneficial effect lies in that, by closely combining the dynamic updates and differences in the characteristics of each participant's data, and through a dynamic and precise privacy budget control mechanism, it can significantly improve the level of data privacy and security protection, and effectively ensure the accuracy of multi-party collaborative computation and analysis, providing solid technical support for the efficient and secure operation of the commodity supply chain platform. Attached Figure Description

[0026] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 A flowchart illustrating the steps of the collaborative analysis method for privacy protection in a commodity supply chain platform based on multi-party computation provided in this application. Detailed Implementation

[0028] To further illustrate the technical means and effects adopted by this application to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the multi-party computation-based collaborative analysis method for privacy protection in a commodity supply chain platform proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0029] Unless otherwise defined, terms such as “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a circuit structure, article, or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the article or device that includes said element. Furthermore, the term “and / or” as used herein includes any and all combinations of one or more of the associated listed items. All technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0030] The following, in conjunction with the accompanying drawings, details the specific scheme of the collaborative analysis method for privacy protection of commodity supply chain platforms based on multi-party computation provided in this application.

[0031] This application provides an embodiment of a privacy-preserving collaborative analysis method for a commodity supply chain platform based on multi-party computation. For details, please refer to [link to specific documentation]. Figure 1 This includes the following steps:

[0032] Step 1: Obtain the information data uploaded by each participant in each data update cycle and convert it into structured data format.

[0033] Each participant uploads data to the commodity supply chain platform through different methods, including but not limited to suppliers, manufacturers, wholesalers, and retailers. Specifically, they directly connect to the platform through internal information system interfaces (such as ERP and CRM systems) and transmit information data in structured data form according to the corresponding data update cycle. The update cycle can be daily, monthly, or quarterly, depending on the actual application scenario. In this embodiment, the information data includes a sequence of transaction commodity categories, a sequence of transaction amounts, and inventory data.

[0034] Taking suppliers as an example, the transmitted information data consists of raw transaction data, qualification documents, supply records, etc., stored in the enterprise's ERP system. Through data interface programs, using API calls or database connection technology, the information data is extracted and converted into a platform-recognizable format, resulting in structured data that conforms to the platform's specifications. Furthermore, to improve the quality of the acquired data, this embodiment employs data cleaning methods to preprocess the raw information data uploaded by the participants. This includes removing duplicate data, filling in missing values, and identifying and removing abnormal data. The specific data cleaning process is well-known to those skilled in the art, and detailed procedures will not be elaborated upon here.

[0035] Step 2: For the various information data uploaded by each participant, obtain the characteristic coefficient of dynamic comprehensive change of each participant in each period by observing the differences in the various information data of each participant in different periods, and obtain the characteristic value of the dynamic change of sensitivity of each information data by combining the differences of each period with the information data in other periods.

[0036] After the above processing, the preprocessed information data of each participant in the commodity supply chain platform is obtained. Since there are significant differences in the sensitivity and distribution characteristics of the data during actual transmission, privacy protection is implemented for the transmitted data in the commodity supply chain platform. In the process of collaborative analysis and calculation between different participants, dynamic analysis of the changes in sensitivity and distribution characteristics of the data uploaded by different participants is required. In this way, the privacy budget for differential privacy protection during collaborative analysis and calculation is optimized and adjusted.

[0037] Specifically, considering the changes in the commodity supply characteristics of different participants at different times, leading to differences in the sensitivity characteristics of information data among participants at different times, a comparative analysis of the information data changes uploaded by different participants to the commodity supply chain platform is conducted. For each type of information data uploaded by each participant, the sensitivity characteristics of the information data are analyzed based on the following: The first type of participant uploaded Taking the first type of information data as an example, the first... The information data is divided according to its corresponding data update cycle, and the first data in each cycle is calculated separately. This type of information data and other information within each cycle The Hamming distance of the information data is defined as follows: the larger the Hamming distance, the greater the value of the first Hamming distance within different data update cycles. The greater the difference in the characteristics of the updated information data, the more significant the variation; furthermore, the average value of all the Hamming distances is taken as the first... The first type of participant The first characteristic value of the information data is dynamically updated and changed in each period; the larger the first characteristic value, the more significant the dynamic update and change of the information data uploaded by the current participant is due to changes in commodity supply characteristics, and the greater the fluctuation of information change in different data update periods.

[0038] To further highlight the comprehensive differences in the dynamic changes of different information data characteristics during the data update process, the first... The sequence of first characteristic values, which dynamically update and change the information data of all participants in each period, serves as the first... The characteristic sequences of each participant in each period are compared based on the consistency of information data feature changes within different periods of the characteristic sequences. Specifically, the first period is calculated. The mean of the Euclidean distances of each participant's period to all other periods with respect to the feature sequence is used as the first... The characteristic coefficients of the dynamic changes of each participant in each cycle are calculated. The larger the characteristic coefficient, the greater the difference in the dynamic changes of information data during the data update process, due to the different commodity supply characteristics of different participants in the actual transaction process at different times, and the more significant the sensitivity characteristics of the corresponding data. Based on the above calculation results, the sensitivity of the dynamic changes of each type of information data in the uploaded information data is comprehensively analyzed to obtain the characteristic value of the dynamic changes of the sensitivity of each type of information data in the data uploaded by each participant. The specific calculation relationship is as follows:

[0039] ;

[0040] in, Indicates the first The first type of participant The characteristic values ​​of the sensitivity of information data change dynamically; Indicates the first The first type of participant This type of information data in the first The first characteristic value corresponding to each period; Indicates the first Various participants in the first The characteristic coefficients of the dynamic comprehensive change of each cycle; Indicates the first The sum of the characteristic coefficients of the dynamic comprehensive changes of all participants across all cycles; Indicates the first The number of cycles for each participating party.

[0041] The larger the calculated characteristic value of the dynamic change in sensitivity, the more significant the sensitivity change characteristics of the corresponding information data in the information data uploaded by the participants. Compared with the traditional analysis method based on the overall data fluctuation, this application fully considers the differences in the changes of information data updates by participants at different times and the consistency differences in the dynamic updates of different information data, thereby conducting a precise sensitivity change characteristic analysis of the information data uploaded by different participants.

[0042] Step 3: Cluster the participants using the feature values ​​of the dynamic changes in the sensitivity of all information data of each participant. Based on the average level of the feature values ​​of the dynamic changes in the sensitivity of all participants regarding various information data in each cluster, and the amount of various information data in each cluster, obtain the first adjustment coefficient and the second adjustment coefficient of differential privacy protection adjustment for various information data, and then obtain the proportional coefficient of privacy budget allocation for various information data.

[0043] Before collaborative analysis of data uploaded by different participants using secure multi-party computation in a commodity supply chain platform, differential privacy protection must be applied to the data provided by each participant to prevent information leakage. Therefore, based on the characteristic values ​​of dynamic changes in the sensitivity of information data uploaded by different participants, all participants are clustered. The purpose is to accurately identify participants with similar characteristics of dynamic changes in the sensitivity of information data, thereby achieving precise allocation of privacy budget.

[0044] Furthermore, the information feature sequence of each participant is formed by the sequence of feature values ​​of the dynamic changes in sensitivity calculated from all the information data uploaded by each participant. In this embodiment, for ease of description, the information feature sequence corresponding to each participant is used as input, and the DBSCAN clustering algorithm is used to obtain the clustering results of each participant. Preferably, in this embodiment, the DTW distance between different information feature sequences is used as the metric distance between different participants during the clustering process. The specific implementation process of the DBSCAN clustering algorithm is well known to those skilled in the art and will not be described in detail here.

[0045] For each cluster in the clustering results, the mean of the normalized result of the characteristic value of the dynamic change of the sensitivity of the same information data of all participants in the cluster is calculated, and the mean is used as the first adjustment coefficient for differential privacy protection adjustment of the corresponding information data; and the normalized result of the quantity of each type of information data of each participant in the cluster is calculated, and the normalized result is used as the second adjustment coefficient for differential privacy protection adjustment of each type of information data of each participant. The larger the second adjustment coefficient is, the greater the sensitivity of the corresponding data if there is a large dynamic change feature of information when the data volume is large.

[0046] Therefore, preferably, in this embodiment, based on the proportion of the first adjustment coefficient and the second adjustment coefficient of differential privacy protection adjustment for each type of information data among the first adjustment coefficient and the second adjustment coefficient of differential privacy protection adjustment for all types of information data, the privacy budget for differential privacy protection of each type of information data uploaded by each participant is proportionally divided to obtain the proportional coefficient of privacy budget allocation for each type of information data of each participant. The specific calculation relationship is as follows:

[0047] ;

[0048] in, Indicates the first The first type of participant The proportion of information data allocated to the privacy budget; and They represent the first species and first The first adjustment coefficient for differential privacy protection adjustment of information data; and They represent the first The first type of participant species and first The second adjustment coefficient for differential privacy protection of information data; Indicates the first The number of types of information data of various participants.

[0049] Understandably, the larger the calculated proportional coefficient, the more dynamic the changes in the information data uploaded by each participant are, indicating that the current participant's [number]th [participant's] [data] is more dynamic. The greater the sensitivity of dynamically changing information data, the smaller the privacy budget needs to be to improve the security protection of information data.

[0050] Step 4: Based on the proportion coefficient of privacy budget allocation according to the various information data of each participant, and combined with the preset privacy budget upper limit, obtain the privacy budget for various information data of each participant, so as to carry out differential privacy protection for the information data uploaded by each participant.

[0051] Based on the comprehensive analysis of the characteristics of each type of information data uploaded by the participants, a proportional coefficient for the allocation of the privacy budget for differential privacy protection of each type of information data uploaded by each participant is obtained. Therefore, the size of the privacy budget for each type of information data of the participants is calculated based on the proportional coefficient. The specific calculation formula is as follows: ,in Indicates the first The first type of participant Privacy budget for information data; The privacy budget cap is used to control the privacy budget of various information data during differential privacy processing. In order to avoid excessive noise causing a decline in information data quality, the privacy budget cap is set to 5 in this embodiment. Indicates the first The first type of participant The proportion of information data allocated to the privacy budget.

[0052] Furthermore, as the supply chain continues to operate, data changes from each participant are monitored in real time, and the level of privacy protection is dynamically adjusted. Based on the adjusted privacy budget allocation, differential privacy processing is applied to the information uploaded by different participants. Specifically, for transaction amount data, Gaussian noise is added, and an appropriate noise standard deviation is calculated based on the allocated privacy budget to preserve the statistical characteristics of the data as much as possible while ensuring privacy. For product category sequences, Laplace's algorithm is used to ensure that changes in a single product category do not affect the overall analysis results. Data processed with differential privacy can be safely used for collaborative analysis while meeting differential privacy constraints.

[0053] After differential privacy processing, homomorphic encryption is applied to the data to further ensure its security during computation and storage. Each participant encrypts the differentially privacy-processed data using a homomorphic encryption key, and the encrypted data is stored in ciphertext form in a decentralized storage network. For example, for transaction amount data, other participants cannot directly obtain the specific value after encryption, but they can perform calculations such as summation and averaging in the ciphertext state, such as calculating the total transaction amount of the entire supply chain. The decrypted result is consistent with the result calculated on the plaintext, thus achieving collaborative data computation under privacy protection.

[0054] The encrypted data is stored in a distributed storage system built on blockchain technology. Each participant, as a node in the blockchain network, possesses a complete copy of the data. Consensus mechanisms, such as the Practical Byzantine Fault Tolerance (PBFT) algorithm, ensure data consistency and immutability. During collaborative analysis, participants do not need to decrypt the data; they directly send the encrypted data to the computing nodes. The computing nodes utilize the properties of homomorphic encryption to perform predetermined computational tasks on the encrypted data, such as analyzing sales trends in different regions and predicting inventory demand. After computation, the encrypted results are returned to each participant, who uses their respective decryption keys to obtain the final result. This approach protects data privacy while enabling collaborative analysis of multi-party data across the supply chain, providing data-driven decision support for optimizing supply chain operations.

[0055] It is understood that references to "one embodiment" or "some embodiments" in this specification mean that one or more embodiments of this application include the specific features, structures, or characteristics described in connection with that embodiment. Therefore, the appearance of phrases such as "in one embodiment," "in some embodiments," "in other embodiments," or "in still other embodiments" in different parts of this specification does not necessarily refer to the same embodiment, but rather means "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0056] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous. Moreover, the sequence numbers of the steps in the embodiments do not imply a specific order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments in this specification.

[0057] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A privacy protection collaborative analysis method for a commodity supply chain platform based on multi-party computation, characterized in that, The method comprises the following steps: obtaining information data uploaded by each participant in each data update period and converting the information data into a structured data form; obtaining a characteristic coefficient of dynamic comprehensive changes of each participant in each period by means of the difference and change of various information data uploaded by each participant in different periods, and obtaining a characteristic value of dynamic sensitivity changes of various information data by means of the difference of each information data in each period compared with other periods; carrying out clustering and division of the participants by means of the characteristic value of dynamic sensitivity changes of all information data of each participant, obtaining a first adjustment coefficient and a second adjustment coefficient of differential privacy protection adjustment of various information data respectively according to the average level of the characteristic value of dynamic sensitivity changes of various information data of all participants in each cluster and the data volume of each information data of each participant in the cluster, and further obtaining a proportional coefficient of privacy budget allocation of various information data of each participant; obtaining the privacy budget of various information data of each participant according to the proportional coefficient of privacy budget allocation of various information data of each participant and combining a preset upper limit of the privacy budget, so as to carry out differential privacy protection on the information data uploaded by each participant; taking the average of the normalized results of the characteristic value of dynamic sensitivity changes of the same information data of all participants in each cluster as the first adjustment coefficient of differential privacy protection adjustment of the same information data; taking the normalized results of the number of each information data of each participant in each cluster as the second adjustment coefficient of differential privacy protection adjustment of each information data of each participant; the calculation method of the proportional coefficient of privacy budget allocation of various information data of each participant is as follows: ； wherein, represents the number of the first information data of the first party; and respectively represent the first adjustment coefficient of the first and the second information data difference privacy protection adjustment; and respectively represent the second adjustment coefficient of the first and the second information data difference privacy protection adjustment of the first party; represents the number of the first party information data.

2. The privacy-preserving collaborative analysis method based on multi-party computation for commodity supply chain platform according to claim 1, wherein, the method for obtaining the characteristic coefficient of dynamic comprehensive changes of each participant in each period is as follows: obtaining a first characteristic value of dynamic update changes of various information data of each participant in each period by means of the difference of the same information data of each participant in each period and other periods, so as to extract a characteristic sequence of each period of each participant, and taking the average of the Euclidean distances of the characteristic sequences of each period and other periods of each participant as the characteristic coefficient of dynamic comprehensive changes of each participant in each period.

3. The privacy-preserving collaborative analysis method based on multi-party computation for commodity supply chain platform according to claim 2, wherein, The method for obtaining the first characteristic value further comprises the following steps: taking the average of the Hamming distances of the same information data between each period and other periods of each participant as the first characteristic value of dynamic update changes of the same information data in each period.

4. The privacy-preserving collaborative analysis method based on multi-party computation for commodity supply chain platform of claim 2, wherein, The method for extracting the characteristic sequence of each period of each participant further comprises the following steps: taking a sequence formed by the first characteristic values of dynamic update changes of all information data of each participant in each period as the characteristic sequence of each period of each participant.

5. The privacy-preserving collaborative analysis method based on multi-party computation for commodity supply chain platform of claim 2, wherein, the calculation method of the characteristic value of dynamic sensitivity changes of various information data is as follows: ； in, Indicates the first The first of the participating parties The characteristic values ​​of the sensitivity of information data change dynamically; Indicates the first The first type of participant This type of information data in the first The first characteristic value is dynamically updated and changes within each cycle; Indicates the first Various participants in the first The characteristic coefficients of the dynamic comprehensive change of each cycle; Indicates the first The sum of the characteristic coefficients of the dynamic comprehensive changes of all participants across all cycles; Indicates the first The number of cycles for each participating party.

6. The privacy-preserving collaborative analysis method based on multi-party computation for commodity supply chain platform of claim 1, wherein, taking a sequence formed by the characteristic values of dynamic sensitivity changes of all information data uploaded by each participant as an information characteristic sequence of each participant, and taking the DTW distance between the information characteristic sequences of different participants as a measurement distance between different participants in the process of clustering and division of the participants.

7. The privacy-preserving collaborative analysis method based on multi-party computation for commodity supply chain platform of claim 1, wherein, The method for calculating the privacy budget of various information data of each participant is: ; wherein, represents the first information data of the first party; is a privacy budget upper limit; represents the first information data of the first party; and a proportionality coefficient for privacy budget allocation of the first information data of the first party.

Images