A secure storage method for private business data in a business process management system

Abstract

The present application relates to the field of privacy data storage technology, and specifically to a method for securely storing privacy business data for a business process management system, the method comprising: constructing each privacy data encoding sequence; obtaining each clustering cluster of each privacy data encoding sequence, and then obtaining the key index of each clustering cluster; obtaining the correlation index between each privacy data encoding sequence; constructing a sliding window; obtaining the similarity index of the sliding window of each privacy data encoding sequence; and then obtaining the degree of difference between the sliding window encoding data sequence and other privacy data encoding sequences; calculating the loneliness index of the sliding window in each privacy data encoding sequence at each iteration, constructing a plaintext matrix, and securely storing the privacy business data. The present application reduces the similarity between plaintext matrices, and improves the security of the AES algorithm when processing highly correlated and highly similar privacy business data.

Description

Technical Field

[0001] The present application relates to the technical field of privacy data storage, and in particular to a method for securely storing privacy business data for a business process management system. Background Art

[0002] At present, business process management systems have become one of the core support platforms for modern enterprise management. Business process management systems can automatically integrate complex business operations within an enterprise, achieving all-round management from process to design, execution to optimization. However, in the process of massive business data being continuously generated, when it comes to sensitive and private business data, how to store it securely becomes an important challenge. Data security and privacy issues always run through the data life cycle, and business data security issues are often reflected in data loss, data leakage, and data attacks. When processing private business data, although the business process management system has a certain degree of data encryption and access control mechanisms, there are still problems such as plain text storage of sensitive information, lack of effective data desensitization measures, and low static data encryption strength. These problems will lead to the risk of information leakage of private business data.

[0003] The AES algorithm is a classic encryption algorithm with the characteristics of fast speed, easy standardization, nonlinearity, and high security. However, the plaintext matrix of the AES algorithm is constructed in sequence. If the plaintext matrix is ​​used to construct the plaintext matrix when processing high-correlation and high-similarity data, the plaintext matrix at the corresponding position in the data will have a high correlation, thereby reducing the security of the privacy data. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides a method for securely storing private business data for a business process management system to solve the existing problems.

[0005] The method for securely storing private business data in a business process management system of this application adopts the following technical solutions:

[0006] An embodiment of the present application provides a method for securely storing private business data for a business process management system, the method comprising the following steps:

[0007] Collect various privacy data coding sequences in business process management systems;

[0008] Using a clustering algorithm to obtain clusters of each private data encoding sequence, obtaining a key index of each cluster according to the distance between the encoding data in each cluster; obtaining a correlation index between private data encoding sequences based on the key index and the similarity between the clusters of each private data encoding sequence;

[0009] Constructing a window of a preset length as a sliding window of each private data encoding sequence, with a sliding step length being a preset step length; obtaining a similarity index of the sliding window of each private data encoding sequence based on the key index and the encoded data in the sliding window of the private data encoding sequence;

[0010] The sequence composed of all the coded data in the sliding window in each private data coding sequence is used as the sliding window coded data sequence, and the difference between the sliding window coded data sequence and other private data coding sequences is obtained based on the correlation index between the private data coding sequences and the distance between the sliding window coded data sequence and other private data coding sequences;

[0011] According to the similarity index and the degree of difference, the loneliness index of the sliding window in each privacy data encoding sequence is obtained in each iteration; a plaintext matrix is ​​constructed according to the loneliness index, and all plaintext matrices are encrypted separately using the key, and the plaintext is converted into ciphertext to encrypt and store the privacy business data.

[0012] Furthermore, the various privacy data coding sequences collected in the business process management system include:

[0013] Customer data, order data, address data and total consumption data are collected and encoded to construct various privacy data encoding sequences. The privacy data encoding sequences include customer information encoding data sequence, order information encoding data sequence, address information encoding data sequence and total consumption encoding data sequence.

[0014] Further, the step of obtaining the key index of each cluster according to the distance between the encoded data in each cluster includes:

[0015] For each cluster of each private data encoding sequence, calculate the edit distance between each encoding data in the cluster, and use the average of all the edit distances as the internal similarity index of each cluster;

[0016] The mean of the number of times the encoded data of all characters in the cluster appear in the encoded sequence of the private data is calculated as the first mean, and the ratio of the internal similarity index to the first mean is used as the key index of each cluster.

[0017] Further, obtaining the correlation index between the private data encoding sequences includes:

[0018] Obtaining a cluster with the largest key index in each private data encoding sequence as the first cluster of each private data encoding sequence;

[0019] For each private data encoding sequence, the number of occurrences of each encoding data in the first cluster in the private data sequence to which it belongs is sorted in ascending order to obtain a number change sequence of the first cluster in each private data encoding sequence;

[0020] For any two private data encoding sequences, the absolute value of the difference between the key indexes of the first clustering clusters of the two private data encoding sequences is calculated; the similarity between the frequency change sequences of the first clustering clusters of the two private data encoding sequences is calculated, and the product of the absolute value of the difference and the similarity is used as the correlation index between any two private data encoding sequences.

[0021] Furthermore, the obtaining of the similarity index of the sliding window of each privacy data encoding sequence includes:

[0022] Use the LCS algorithm to calculate the longest common subsequence length between each sliding window in each private data encoding sequence;

[0023] For each sliding window in each private data encoding sequence, if the characters corresponding to the encoded data in the sliding window are all located in the same cluster, the key index of the cluster is used as the criticality ratio of each sliding window; if the characters corresponding to the encoded data in the sliding window are located in multiple clusters, the average of the key indexes of the multiple clusters is used as the criticality ratio of each sliding window;

[0024] The similarity index is formulated as follows:

[0025]

[0026] Where P u is the similarity index of sliding window u in the address information encoding data sequence K3; N is the length of the address information encoding data sequence K3; L(u,i) is the longest common subsequence length between sliding window u and sliding window i, kd u , kd i are the criticality proportions of sliding window u and sliding window i respectively; n u is the number of characters included in the coded data in the sliding window u, n i is the number of characters included in the coded data in the sliding window i; τ is the preset parameter adjustment coefficient;

[0027] The similarity index of the sliding window in each other privacy data encoding sequence is obtained by the same method as that of the address information encoding data sequence.

[0028] Furthermore, obtaining the difference between the sliding window encoded data sequence and other privacy data encoded sequences includes:

[0029] For each private data encoding sequence, calculate the inverse of the correlation index between the private data encoding sequence and other private data encoding sequences;

[0030] Calculate the edit distance between each sliding window encoded data sequence in each private data encoding sequence and the encoded data with the same number of bits as the sliding encoded data sequence in other private data encoding sequences, and take the product of the mean of all the edit distances in other private data encoding sequences and the reciprocal as the degree of difference between the sliding window in each private data encoding sequence and the other private data encoding sequences.

[0031] Furthermore, the step of obtaining the loneliness index of the sliding window in each privacy data encoding sequence at each iteration includes:

[0032] According to the difference between the sliding coding data sequence of each private data coding sequence and other private data coding sequences, the overall difference coefficient of the sliding window in each private data coding sequence in each iteration is obtained;

[0033] According to the difference degree and similarity index, the global difference evaluation coefficient of the sliding window in each privacy data encoding sequence is obtained at each iteration;

[0034] The sum of the similarity index of the sliding window in each privacy data encoding sequence and the preset parameter adjustment factor is calculated as the first sum value, and the ratio of the global difference evaluation coefficient of the sliding window in each privacy data encoding sequence in each iteration to the first sum value is used as the loneliness index of the sliding window in each privacy data encoding sequence in each iteration.

[0035] Furthermore, the step of obtaining the overall difference coefficient of the sliding window in each privacy data encoding sequence at each iteration includes:

[0036] For the sliding coded data sequence in each privacy data coding sequence at each iteration, the average of the differences between the sliding coded data sequence and other privacy data coding sequences is taken as the overall difference coefficient of the sliding window in each privacy data coding sequence at each iteration.

[0037] Furthermore, the global difference evaluation coefficient is formulated as follows:

[0038]

[0039] is the global difference evaluation coefficient of the sliding window u in the address information encoding data sequence K3 at the mth iteration, len is the data length of the plaintext matrix sequence M; Z(X u , M) is the sliding coding data sequence X u The degree of difference between the plaintext matrix sequence M and the degree of difference between the sliding window encoded data sequence and each privacy data encoded sequence is calculated in the same way, P u is the similarity index of the sliding window u in the address information encoding data sequence K3, and ∝ is the preset parameter factor.

[0040] Further, constructing a plaintext matrix according to the loneliness index includes:

[0041] Calculate the loneliness index of the sliding window under all different sliding steps in each privacy data encoding sequence, and sort the loneliness index in descending order. Process the sorted loneliness index according to the encoded data in the corresponding sliding window. The processing process is: process the encoded data in the sliding window corresponding to the loneliness index in descending order, with every four bytes as a row, to construct all the plaintext matrices in each privacy data encoding sequence.

[0042] This application has at least the following beneficial effects:

[0043] This application analyzes the characteristics of repeated characters in privacy business data, clusters the data and calculates the criticality of the clusters, so as to divide the most critical characters in the privacy business data; then constructs the degree of correlation between data sequences to determine whether two privacy business data have a strong correlation; then constructs the similarity index of the sliding window to determine the difference between the sliding window and other windows in the data sequence; further, by constructing the degree of difference, the difference between the sliding window and other data sequences is evaluated; finally, the loneliness index is constructed to locate a group of data with the greatest difference from all data, and a plaintext matrix is ​​constructed. This application takes into account the high correlation and high similarity between privacy business data. In order to further improve the security of data storage, the AES algorithm is improved in the way of sequentially constructing the plaintext matrix. The loneliness index is used to select data to construct the plaintext matrix, which reduces the similarity between the plaintext matrices and improves the security of the AES algorithm when processing highly correlated and highly similar privacy business data. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0045] Figure 1 A flowchart of a method for securely storing private business data in a business process management system provided by this application;

[0046] Figure 2 Flowchart of the method for obtaining the loneliness index. DETAILED DESCRIPTION

[0047] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the specific implementation method, structure, features and effects of the privacy business data security storage method for the business process management system proposed in the present application are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0048] Unless defined otherwise, 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 belongs.

[0049] The specific scheme of the method for securely storing private business data in a business process management system provided by the present application is described in detail below with reference to the accompanying drawings.

[0050] An embodiment of the present application provides a method for securely storing private business data for a business process management system. Specifically, the following method for securely storing private business data for a business process management system is provided. Figure 1 , the method comprises the following steps:

[0051] Step S001, obtaining private business data and performing preliminary processing on the private business data.

[0052] Business data refers to data used to record business events, including customer information, order information, address information, and total consumption. Customer information, order information, address information, and total consumption business data are all private business data. If private data is not encrypted and stored in ciphertext, privacy leakage is very likely to occur, causing significant impact on individuals or enterprises.

[0053] Obtain the privacy business data that needs to be encrypted from the business management system. This application obtains four sensitive data, namely customer information, order information, address information and total consumption. Encode the characters in each type of data through GB2312 encoding, obtain the encoding data of the characters in each type of data, record the sequence composed of customer information encoding data as customer information encoding data sequence K1, the sequence composed of order information encoding data as order information encoding data sequence K2, the sequence composed of address information encoding data as address information encoding data sequence K3, and the sequence composed of total consumption encoding data as total consumption encoding data sequence K4. The above encoding data sequences are collectively referred to as privacy data encoding sequences.

[0054] At this point, the preprocessed private data encoding sequences are obtained.

[0055] Step S002, by analyzing the data features between the privacy business data, calculate the key index of each cluster, and judge the degree of association between the various data sequences based on this; then build a sliding window and calculate the similarity index between the data sequence in which it is located; then build the degree of difference between the sliding window and the data sequence through the correlation index, and build a loneliness index to judge the association between the window and all data.

[0056] The AES (Advanced Encryption Standard) algorithm is a very classic encryption algorithm. AES uses a block cipher, which has the characteristics of fast encryption speed, easy standardization and easy implementation; and has good nonlinear ability, which makes its encryption degree very high and can protect privacy data securely. However, the AES encryption algorithm divides the plaintext data into multiple 16-byte plaintext matrices in sequence, and constructs the plaintext matrix based on the plaintext matrix. When constructing the plaintext matrix for two groups of data with large feature correlation, the correlation of the plaintext matrices at the corresponding positions in the two groups of data will be too high; and if the data similarity within a group of privacy data is also high, the similarity between adjacent bytes in the plaintext matrix constructed by this group of data is also high. If this high correlation and high similarity are exploited by attackers, they can find out the potential rules in the encryption process by observing multiple plaintexts and corresponding ciphertexts, thereby reducing the security of privacy data. Therefore, by analyzing the characteristics of privacy business data, the plaintext matrix construction method of the AES algorithm applied to privacy business data encryption can be further optimized to improve the security of privacy business data.

[0057] Through further analysis of the privacy data in the business data, it is found that the privacy data in the same business management system have a certain correlation, and the update of a certain privacy data will affect the change of other privacy data; at the same time, because they are in the same business management system, the storage format of each privacy data is also of the same type, and the similarity within each privacy data is high. So far, the characteristics of privacy data have been analyzed, and the correlation between customer information encoding data, order information encoding data, address information encoding data, and total consumption encoding data is relatively large, and the data similarity within each data sequence is also high.

[0058] For each private data encoding sequence, some characters are repeated in the private data encoding sequence. The more times a character appears, the more universal it is and the lower its criticality is; when a character appears less often, its distinction is stronger and the more critical it is.

[0059] The number of times the corresponding code of each character appears in its privacy data code sequence is counted, and the number of occurrences is used as input to cluster each character. A clustering algorithm is used to perform clustering. The number of clusters in this embodiment is set to 5.

[0060] It should be noted that clustering algorithms include K-means clustering algorithm, hierarchical clustering algorithm, DBSCAN clustering algorithm, and OPTICS clustering algorithm. The clustering algorithm used in this embodiment is the k-means clustering algorithm. The implementer can choose other clustering methods according to actual conditions; among them, the k-means algorithm is a well-known technology and will not be repeated here.

[0061] For each cluster of each private data encoding sequence, take the vth cluster as an example, and count the number of characters inside it, denoted as num v . Calculate the key index of the vth cluster using the formula:

[0062]

[0063]

[0064] Where Y v is the internal similarity index in the vth cluster; E(Z i ,Z j ) is used to calculate the EDR edit distance between the i-th and j-th encoded data in cluster v, num v is the number of characters in the vth cluster, is the number of pairwise combinations of all encoded data in the vth cluster; the EDR edit distance is a well-known technology and will not be described here.

[0065] E(Z i ,Z j ) is larger, the more it proves that from character Z i To character Z j The more times editing is required, the greater the difference will be; the editing distance is a well-known technology and will not be described in detail here. is the similarity of the jth character. The smaller its value is, the smaller the difference between the jth character and other characters in the cluster is. v The smaller the value of , the higher the consistency of the internal characters, which proves that there are a large number of identical characters in cluster v.

[0066] G v is the key index in the vth cluster; C v It is the mean number of times the encoded data of all characters in cluster v appear in the encoded sequence of the private data in which it is located.

[0067] C vThe larger the value of G, the more times the characters appear in cluster v, and the lower its criticality. v The smaller the value of G, the more identical characters there are in cluster v, and the more the average number of occurrences of its characters, the lower the criticality. v The larger the value of , the stronger the ability to distinguish clusters and the higher the criticality.

[0068] At this point, the key index of each cluster in each private data encoding sequence is obtained.

[0069] Obtain the cluster with the largest key index in each privacy data encoding sequence, obtain the number of times each encoding data in the cluster appears in the privacy data encoding sequence to which it belongs, and arrange the number of times of each character in the cluster in ascending order to obtain a number change sequence of each cluster.

[0070] Taking the order information encoding data sequence K2 and the address information encoding data sequence K3 as examples, the clusters with the largest key index in the order information encoding data sequence K2 and the address information encoding data sequence K3 are respectively denoted as G K2 With G K3 ; Cluster G K2 With cluster G K3 The number of changes in the sequence is recorded as nt K2 and nt K3 .

[0071] Calculate the correlation index between the order information encoding data sequence K2 and the address information encoding data sequence K3. The formula is:

[0072]

[0073] In the formula, G 2,3 It is the correlation index between the order information encoding data sequence K2 and the address information encoding data sequence K3. is the cluster G K2 The key index of is the cluster G K3 Key index of nt K2 is the cluster G K2 The number of changes in the sequence, nt K3 is the cluster G K3 JAC is the Jaccard similarity function. Jaccard similarity is a well-known technology and will not be described in detail in this embodiment.

[0074] The smaller the value, the higher the similarity between the two clusters. K2 ,nt K3 ) is the number of changes in the sequence ntK2 With nt K3 Jaccard similarity between K2 ,G K3 ) is larger, indicating that the two clusters have more common characters, and thus the correlation is greater.

[0075] Similarly, obtain the correlation indexes between other private data encoding sequences.

[0076] At this point, by calculating the correlation index, the degree of influence between multiple sequences can be judged. The larger the value, the greater the influence and the stronger the correlation.

[0077] Taking the address information encoding data sequence K3 as an example, a sliding window with a length of 16 bytes is constructed to slide in the address information encoding data sequence K3, and the sliding step length is 16 bytes.

[0078] The LCS (Longest Common Subsequence) algorithm is used to calculate the longest common subsequence length between each sliding window in the address information encoding data sequence K3. The LCS (Longest Common Subsequence) algorithm is a well-known technology and will not be described here.

[0079] For each sliding window, if the characters corresponding to the encoded data in the sliding window are all located in the same cluster, the key index of the cluster is used as the criticality ratio of each sliding window; if the characters corresponding to the encoded data in the sliding window are located in multiple clusters, the average of the key indexes of the multiple clusters is used as the criticality ratio of each sliding window.

[0080] Taking the u-th sliding window as an example, the similarity index of sliding window u in the address information encoding data sequence K3 is calculated, and the formula is:

[0081]

[0082] Where P u is the similarity index of the sliding window u in the address information encoding data sequence K3; N is the length of the address information encoding data sequence K3, and the number of valid sliding windows that can be formed in the address information encoding data sequence K3 is L(u,i) is the longest common subsequence length between sliding window u and sliding window i, kd u is the criticality ratio of sliding window u, kd i is the criticality ratio of sliding window i; n u is the number of characters included in the coded data in the sliding window u, n iis the number of characters included in the encoded data in the sliding window i; τ is the parameter adjustment coefficient, and in order to avoid the distribution being 0, the value is 1 in this embodiment.

[0083] The larger the value of L(u,i), the longer the length of consecutive bytes between the two sliding windows is, and the greater the similarity is; |G u -G i The larger the value of |, the greater the difference in the criticality, and the greater the difference between the sliding windows u and i; since the byte lengths occupied by Chinese characters and numbers after encoding are inconsistent, |n u -n i The larger the value of |, the greater the difference in the number of corresponding characters; at this time, P u The smaller the value of , the less similar the sliding window u is to other sliding windows in the K3 data, and the greater the difference.

[0084] Similarly, obtain the similarity index of each sliding window in the customer information encoding data sequence K1, the order information encoding data sequence K2, the address information encoding data sequence K3, and the total consumption encoding data sequence K4.

[0085] Taking the sliding window u in the address information encoding data sequence K3 as an example, the sequence composed of all the encoding data in the sliding window u is recorded as the sliding encoding data sequence Calculate the sliding code data sequence X u The EDR edit distance of the coded data corresponding to the i-th character in the order information coded data sequence K2 is recorded as E(X u ,K2 i ). Calculate the sliding coding data sequence X u The degree of difference between the order information encoding data sequence K2 is as follows:

[0086]

[0087] In the formula, Z(X u , K2) is the sliding coding data sequence X u The degree of difference between the order information encoding data sequence K2; G 2,3 is the correlation index between the order information encoding data sequence K2 and the address information encoding data sequence K3, τ is the preset parameter adjustment coefficient, and in order to avoid the denominator being 0, the value of this embodiment is 1; E ave (X u ,K2 i ) is used to calculate the sliding coding data sequence X u The average of the EDR edit distances between all non-repeated coded data in the order information coded data sequence K2 that have the same number of bits as the sliding coded data sequence. The calculation method of the EDR edit distance is a well-known technique and will not be described in detail here.

[0088] G 2,3 The larger the value, the stronger the correlation. u ,K2 i ) value, the greater the difference. u ,K2) value is larger, indicating that X u The greater the difference between the K2 data sequence, the greater the difference between the K2 data sequence and the K2 data sequence. The calculation method of the EDR edit distance is a well-known technology and will not be described in detail here.

[0089] Similarly, we can calculate X u The difference between the customer information coding data sequence K1 and the total consumption coding data sequence K4 is recorded as Z(X u ,K1)、Z(X u ,K4). Construct an empty data sequence, recorded as the plaintext matrix sequence M=[]. At this time, the length of the plaintext matrix sequence is 0.

[0090] Furthermore, the loneliness index of the sliding window u in the address information encoding data sequence K3 at the mth iteration is calculated as follows:

[0091]

[0092]

[0093]

[0094] In the formula, is the overall difference coefficient of sliding window u in the address information encoding data sequence K3 at the mth iteration, Z(X u , K1) is the sliding coding data sequence X u The degree of difference between the obtained customer information encoding data sequence K1, Z(X u , K2) is the sliding coding data sequence X u The degree of difference between the order information encoding data sequence K2, Z(X u ,K4) Sliding coded data sequence X u The degree of difference between the total consumption encoded data sequence K4.

[0095] Among them, Z(X u ,K1)、Z(X u ,K2)、Z(X u , K4) respectively represent the overall difference between the internal elements in the sliding window u in the K3 data and the K1, K2, and K4 data sequences. The larger the value, the stronger the X u The greater the difference between the K1, K2, and K4 data sequences.

[0096] is the global difference evaluation coefficient of the sliding window u in the address information encoding data sequence K3 at the mth iteration, len is the data length of the plaintext matrix sequence M, when no iteration is performed, the data length of M is 0, len = 0; Z(X u , M) is the sliding coding data sequence X u The degree of difference between the plaintext matrix sequence M and the encoding sequence of the sliding window data is calculated in the same way as the degree of difference between the sliding window encoding data sequence and each privacy data encoding sequence. u is the similarity index of the sliding window u in the address information encoding data sequence K3, ∝ is a preset parameter adjustment factor to prevent the denominator from being zero, and the value is 1 in this embodiment.

[0097] is the loneliness index of the sliding window in the address information encoding data sequence K3 at the mth iteration; the larger its value, the more u The greater the difference between the address information encoding data sequence K3 and other elements, the greater the difference between the customer information encoding data sequence K1, the order information encoding data sequence K2, the total consumption encoding data sequence K4, and the plaintext matrix sequence M. The flowchart of the loneliness index acquisition method is as follows: Figure 2 shown.

[0098] According to the above formula, the iteration process is as follows:

[0099] When the length of the plaintext matrix sequence is 0, it is the 0th iteration. At the 0th iteration, calculate At this time, since the length of the plaintext matrix sequence M is 0, The value of Further acquisition The value of is used as the loneliness index of the sliding window u at the 0th iteration; at this time, the encoded data in the sliding window u is assigned to the plaintext matrix sequence M, and the sliding window u is slid back 16 bits for the first iteration.

[0100] During the first iteration, the sliding window after sliding is calculated At this time, since the length of the plaintext matrix sequence M is greater than zero, for To further obtain the value of The value of is used as the loneliness index of the sliding window u in the first iteration; at this time, the plaintext matrix sequence M is cleared, and the encoded data in the sliding window after sliding one bit is assigned to the plaintext matrix sequence M, and the sliding window u is slid back 16 bits for the second iteration.

[0101] The subsequent iteration process is similar and will not be elaborated on here.

[0102] At this point, the loneliness index of the sliding window u in each iteration is obtained.

[0103] By calculating the loneliness index, we can effectively evaluate the similarity between the sliding window u and the data sequence it is in, as well as the overall difference between the window and other data sequences. The larger the loneliness index value, the smaller the correlation and similarity between the internal elements in the sliding window u and all the data, and the more suitable it is for constructing the plaintext matrix.

[0104] Step S003, by calculating the loneliness index, the construction method of the plaintext matrix is ​​optimized to achieve secure storage of private business data.

[0105] Taking the address information encoding data sequence K3 as an example, the loneliness index of all sliding windows in the address information encoding data sequence K3 is calculated, and the loneliness index is sorted in order from large to small, and the encoded data in the sliding window corresponding to the loneliness index is processed in order from large to small. Every four bytes are used as a row to construct all the plaintext matrices in the address information encoding data sequence K3, and all the plaintext matrices are encrypted separately using the key to convert the plaintext into ciphertext to realize the encrypted storage of privacy business data. The subsequent encryption process of the AES encryption algorithm is a well-known technology and will not be repeated here.

[0106] Similarly, the encryption of the customer information coding data sequence K1, the order information coding data sequence K2, and the total consumption coding data sequence K4 is completed.

[0107] At this point, through the calculation of the loneliness index, it can be ensured that the plaintext matrix data used to construct the plaintext matrix has a small correlation and similarity with other data; at the same time, the original data will be scrambled after each construction of the plaintext matrix to further confuse the association between the data. The original data can also be restored through the subscript sequence XB, which improves the robustness of the AES algorithm when processing data with high correlation and similarity, and realizes the secure storage of privacy business data.

[0108] It should be noted that the above sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0109] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0110] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Modifications to the technical solutions recorded in the aforementioned embodiments, or equivalent replacement of some of the technical features therein, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for securely storing private business data in a business process management system, characterized in that: The method comprises the following steps: Collect various privacy data coding sequences in business process management systems; Using a clustering algorithm to obtain clusters of each private data encoding sequence, obtaining a key index of each cluster according to the distance between the encoding data in each cluster; obtaining a correlation index between private data encoding sequences based on the key index and the similarity between the clusters of each private data encoding sequence; Constructing a window of a preset length as a sliding window of each private data encoding sequence, with a sliding step length being a preset step length; obtaining a similarity index of the sliding window of each private data encoding sequence based on the key index and the encoded data in the sliding window of the private data encoding sequence; The sequence composed of all the coded data in the sliding window in each private data coding sequence is used as the sliding coded data sequence, and the difference between the sliding coded data sequence and other private data coding sequences is obtained based on the correlation index between the private data coding sequences and the distance between the sliding coded data sequence and other private data coding sequences; Iteration 0: Initially, an empty set is preset as the plaintext matrix sequence. The loneliness index of the sliding window is obtained according to the similarity index, the difference between the sliding coded data sequence and other private data coded sequences, and the difference between the sliding coded data sequence and the plaintext matrix sequence. At this time, the coded data of the sliding window is assigned to the plaintext matrix sequence, and the sliding window is slid backward. Iteration 1: The loneliness index of the current sliding window is calculated according to the new plaintext matrix sequence and the current sliding coded data sequence. At this time, the plaintext matrix sequence is cleared, the coded data in the current sliding window is assigned to the plaintext matrix sequence, and the sliding window is slid backward. The loneliness index of the sliding window of each iteration is obtained according to the method of the first iteration. The plaintext matrix is ​​constructed according to the loneliness index of all sliding windows in the private data coded sequence, and all plaintext matrices are encrypted respectively using the key, and the plaintext is converted into ciphertext, and the private business data is encrypted and stored. The obtaining of the similarity index of the sliding window of each privacy data encoding sequence includes: Use the LCS algorithm to calculate the longest common subsequence length between each sliding window in each private data encoding sequence; For each sliding window in each private data encoding sequence, if the characters corresponding to the encoded data in the sliding window are all located in the same cluster, the key index of the cluster is used as the criticality ratio of each sliding window; if the characters corresponding to the encoded data in the sliding window are located in multiple clusters, the average of the key indexes of the multiple clusters is used as the criticality ratio of each sliding window; The similarity index is formulated as follows: ； In the formula, Encode sequence for private data Similarity index of sliding window u in ; Data sequence for private data Length; is the longest common subsequence length between sliding window u and sliding window i, is the criticality ratio of sliding window u, is the criticality ratio of sliding window i; is the number of characters included in the encoded data in the sliding window u, is the number of characters included in the encoded data in the sliding window i; is the preset parameter adjustment coefficient; For the similarity index of the sliding window in other private data encoding sequences, the similarity index of the sliding window in the private data encoding sequence is The same method is used to obtain.

2. The method for securely storing private business data in a business process management system according to claim 1, characterized in that: The various privacy data encoding sequences in the collection business process management system include: Customer data, order data, address data and total consumption data are collected and encoded to construct various privacy data encoding sequences. The privacy data encoding sequences include customer information encoding data sequence, order information encoding data sequence, address information encoding data sequence and total consumption encoding data sequence.

3. The method for securely storing private business data in a business process management system according to claim 1, characterized in that: The step of obtaining the key index of each cluster according to the distance between the encoded data in each cluster includes: For each cluster of each private data encoding sequence, calculate the edit distance between each encoding data in the cluster, and use the average of all the edit distances as the internal similarity index of each cluster; The mean of the number of times the encoded data of all characters in the cluster appear in the encoded sequence of the private data is calculated as the first mean, and the ratio of the internal similarity index to the first mean is used as the key index of each cluster.

4. The method for securely storing private business data for a business process management system according to claim 1, characterized in that: The obtaining of the correlation index between the private data encoding sequences includes: Obtaining a cluster with the largest key index in each private data encoding sequence as the first cluster of each private data encoding sequence; For each private data encoding sequence, the number of occurrences of each encoding data in the first cluster in the private data sequence to which it belongs is sorted in ascending order to obtain a number change sequence of the first cluster in each private data encoding sequence; For any two private data encoding sequences, the absolute value of the difference between the key indexes of the first clustering clusters of the two private data encoding sequences is calculated; the similarity between the frequency change sequences of the first clustering clusters of the two private data encoding sequences is calculated, and the product of the absolute value of the difference and the similarity is used as the correlation index between any two private data encoding sequences.

5. The method for securely storing private business data for a business process management system according to claim 1, characterized in that: The step of obtaining the difference between the sliding coded data sequence and other privacy data coded sequences includes: Privacy data encoding sequence The sliding window in is denoted as u, and the sequence composed of all the coded data in the sliding window u is taken as the sliding coded data sequence , sliding coded data sequence Sequence encoded with private data The degree of difference is: ； In the formula, Sliding coded data sequence Sequence encoded with private data the degree of difference between Encode sequence for private data Sequence encoded with private data The relevant index of is the preset parameter adjustment coefficient; To calculate the sliding coding data sequence Sequence encoded with private data The mean of the EDR edit distances between all non-repeating coded data with the same number of bits as the sliding coded data sequence.

6. The method for securely storing private business data for a business process management system according to claim 5, characterized in that: The loneliness index of the sliding window includes: ； ； ； In the formula, Sliding coded data sequence Sequence encoded with private data The degree of difference between Sliding coded data sequence Sequence encoded with private data The degree of difference between is the private data encoding sequence at the mth iteration The overall coefficient of difference of sliding window u in ; Sliding coded data sequence The degree of difference between the plaintext matrix sequence M, is the length of the plaintext matrix sequence M, is the private data encoding sequence at the mth iteration The global difference evaluation coefficient of the sliding window u in ; is the private data encoding sequence at the mth iteration The loneliness index of sliding window u in , It is the preset parameter factor.

7. The method for securely storing private business data for a business process management system according to claim 1, characterized in that: Construct a plaintext matrix, including: For private data encoding sequence , calculate the private data encoding sequence The loneliness index of all sliding windows in the , and sort the loneliness index from large to small, and process the encoded data in the sliding window corresponding to the loneliness index in the order from large to small, with each four bytes as a line, to construct the privacy data encoding sequence All the plaintext matrices in .

Images