A secure transmission method for enterprise procurement data based on big data

Abstract

The present invention relates to the field of data security technology, and in particular to a method for securely transmitting enterprise procurement data based on big data, the method comprising: obtaining procurement data to be transmitted by the enterprise; constructing a neighbor sequence of each character according to the position difference between binary data of any two characters in the procurement data and the degree of repetition of each character; obtaining a limited threshold according to the median distribution of the neighbor sequence of each character; obtaining the cluster center selectivity of each character based on the global difference separation degree of each character and the limited threshold; optimizing the clustering algorithm according to the maximum cluster center selectivity and the minimum density distance to obtain a difference cluster; constructing a private key with randomly selected elements in the difference cluster, and outputting encrypted data using an ECC algorithm for the procurement data to be transmitted by the enterprise. The present invention aims to screen characters with relatively large overall differences, thereby improving the security of the ECC algorithm.

Description

Technical Field

[0001] The present application relates to the field of data security technology, and in particular to a method for securely transmitting enterprise procurement data based on big data. Background Art

[0002] Big data refers to data sets that are large in scale, grow rapidly, are diverse in variety, require high processing timeliness, and are difficult to capture, manage, and analyze with traditional data processing tools, such as corporate procurement data. Corporate procurement data refers to various electronic records and information generated by companies during their daily operations and procurement activities. These data may include but are not limited to highly sensitive information such as supplier information, purchase orders, transaction records, and strategic planning. In order to protect companies from the leakage of procurement information, the transmitted data needs to be encrypted during the transmission of corporate procurement data.

[0003] The ECC algorithm is the main encryption algorithm used in traditional transmission processes. However, since the private key of the ECC algorithm is generated by a pseudo-random number, there are certain security risks, and the data of enterprise procurement data is highly obfuscated. Therefore, the present invention obtains the private key by performing data analysis using the enterprise procurement data to improve the security of the ECC algorithm. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method for securely transmitting enterprise procurement data based on big data to solve the existing problems.

[0005] The secure transmission method of enterprise procurement data based on big data of the present invention adopts the following technical solutions:

[0006] An embodiment of the present invention provides a method for securely transmitting enterprise procurement data based on big data, the method comprising the following steps:

[0007] Obtain the procurement data to be transmitted by the enterprise;

[0008] For each character in the procurement data, the position difference between any two characters is obtained based on the position difference between the binary data of any two characters; the global difference separation of each character is obtained based on the position difference between each character and the remaining characters and the degree of repetition of each character; the neighbor sequence of each character is obtained based on the global difference separation of each character; the median set and the quartile distribution mean are obtained based on the median distribution of the neighbor sequence of each character; the limiting threshold is obtained by combining the quartile distribution mean and the element distribution in the median set;

[0009] Based on the global difference separation of each character, the density distance of each character is obtained; based on the limited threshold, the density distance of each character and the global difference separation, the cluster center selectivity of each character is obtained; based on the maximum cluster center selectivity and the minimum density distance, the clustering algorithm is optimized to obtain the difference cluster;

[0010] The characters corresponding to the randomly selected elements in the difference cluster are combined into binary data in the order of selection. The binary data is used as the private key and the ECC algorithm is used to encrypt the procurement data to be transmitted by the enterprise, thereby completing the secure transmission of the enterprise procurement data.

[0011] Preferably, obtaining the position difference between any two characters according to the position difference between binary data of any two characters comprises:

[0012] The binary data of any two characters are calculated by using an XOR operation, and the calculation result is recorded as a binary difference string; the subscript index values ​​of all elements in the binary difference string whose values ​​are 1 are obtained to form a non-zero subscript sequence;

[0013] For each element of the non-zero subscript sequence, calculate the difference between the element and the left adjacent element, use the difference as the exponent of an exponential function with a natural constant as the base, calculate the sum of the calculation results of the exponential function for all elements of the non-zero subscript sequence, and record it as the first sum;

[0014] The number of elements in the non-zero subscript sequence is obtained; and the product of the first sum value and the number of elements is used as the position difference between any two characters.

[0015] Preferably, the step of obtaining the global difference separation degree of each character according to the position difference degree between each character and the remaining characters and the repetition degree of each character includes:

[0016] The sum of the position differences between each character and all remaining characters is obtained and recorded as a second sum; the frequency of occurrence of each character in the procurement data is obtained; and the ratio of the second sum to the frequency is used as the global difference separation degree of each character.

[0017] Preferably, the step of obtaining a neighbor sequence of each character based on the global difference separation degree of each character includes:

[0018] Calculate the absolute value of the difference between the global difference separation of each character and the remaining characters in the purchased data;

[0019] All the absolute values ​​of the differences of each character are sorted from small to large, and the global difference separation degrees of the first m elements after sorting are arranged in order from small to large to form the neighbor sequence of each character.

[0020] Preferably, the step of obtaining a median set and a quartile distribution mean according to the median distribution of each character neighbor sequence includes:

[0021] Get the median of the neighbor sequence of each character, and form a median set with the medians of all characters;

[0022] Get the mean of all data between the first quartile and the third quartile in the median set, recorded as the quartile distribution mean.

[0023] Preferably, the step of obtaining a limited threshold by combining the quartile distribution mean and the element distribution in the median set includes:

[0024] Get the maximum, minimum and median of the median set;

[0025] Calculate the difference between the median and the minimum value as the first difference; calculate the difference between the maximum value and the median as the second difference;

[0026] The ratio of the first difference to the second difference is calculated, and the product of the ratio and the mean of the quartile distribution is used as a limiting threshold.

[0027] Preferably, the step of obtaining the density distance of each character based on the global difference separation degree of each character includes:

[0028] Calculate the absolute value of the difference between each character and the global difference separation degree of all remaining characters in the purchased data;

[0029] All the absolute values ​​of the differences of each character are sorted from small to large, and the Kth absolute value of the difference in the sorting result is divided by the value K to obtain the density distance of each character.

[0030] Preferably, the step of obtaining the cluster center selectivity of each character according to a defined threshold value and the density distance and global difference separation degree of each character includes:

[0031] The global difference separation degree of the character to which the minimum density distance belongs is used as the first initial cluster center; the density distances of all the remaining characters except the character to which the minimum density distance belongs are used to form a density distance sequence;

[0032] For each element of the density distance sequence, the selection component of each character is obtained by combining the limited threshold and the difference between the global difference separation degree of each element and the first initial cluster center; the ratio of the selection component of each character to the density distance is used as the cluster center selectivity of each character.

[0033] Preferably, the step of obtaining the selection component of each character by combining the defined threshold and the difference between the global difference separation degree of each element and the first initial cluster center comprises:

[0034] Calculate the absolute value of the difference between the global difference separation degree of each character and the first initial cluster center value as the distance value of each character;

[0035] When the distance value is less than or equal to a specified threshold, the distance value of the character is used as the selection component of the corresponding character; otherwise, the difference between the specified threshold and the distance value of the character is used as the exponent of an exponential function with a natural constant as the base, and the calculation result of the exponential function is used as the selection component of the corresponding character.

[0036] Preferably, the step of optimizing the clustering algorithm according to the maximum cluster center selectivity and the minimum density distance to obtain the difference clusters includes:

[0037] The global difference separation degree of the character with the minimum density distance is recorded as the first center; the global difference separation degree of the character with the maximum cluster center selectivity is recorded as the second center;

[0038] The global difference separation of all characters is used as the input of the clustering algorithm, the first center and the second center are used as the initial clustering centers of the clustering algorithm, and the output of the clustering algorithm is two clusters; the cluster with the largest element mean is recorded as the difference cluster.

[0039] The present invention has at least the following beneficial effects:

[0040] The present invention obtains the position difference degree by calculating the corresponding difference position relationship between the binary data of any two characters, and reflects the complexity of constructing the private key based on the position difference situation between any two characters, which is conducive to constructing a private key with higher complexity through the position difference relationship between the characters, enhancing the difficulty of cracking the private key and improving the security of the data; combining the position difference degree between the characters with the frequency of the characters appearing in the procurement data, the global difference separation degree of the characters is calculated, and the global difference separation degree not only takes into account the number of times the characters appear in the data, that is, the characters are relatively discrete relative to the normal mode of the entire data set, but also takes into account the distance between the binary data differences of the characters, which has the beneficial effect of improving the accuracy of calculating the difference degree of the characters in the global, so that the consideration direction is more comprehensive;

[0041] At the same time, the present invention constructs a neighbor sequence of each character based on the global difference separation degree of each character, and through the median distribution in the neighbor sequence, explores the overall distribution size characteristics of the data, distinguishes the concentration degree between the data and the maximum and minimum values, thereby affecting the limited threshold of the global difference separation degree, avoiding the initial clustering cluster center in the subsequent clustering process from seriously deviating from the mainstream data, and reducing the influence of extreme data values ​​on data calculation; the present invention calculates the density distance of each character based on the K nearest neighbor data between the global difference separation degrees of each character, and constructs the cluster center selectivity of each character in combination with the limited threshold and density distance, and jointly evaluates the selection of each character as the next character from the threshold level and the spatial distribution density level. The possibility of clustering centers is increased, making the evaluation result more reasonable; by taking the character with the smallest density distance among all characters as the cluster center, and considering the distance and outlier factors at the same time, the character with the largest cluster center selectivity is taken as another cluster center, thereby obtaining the optimized clustering effect, which greatly reduces the number of iterations of the clustering algorithm and improves the calculation speed. Moreover, through the restrictions on density and distance, the K-means clustering effect can be improved, thereby screening out characters with large differences from the overall characters and improving the security of the encryption algorithm; based on randomly extracting characters from the difference clusters obtained by clustering to construct private keys, it has the characteristics of real-time, randomness and high obfuscation, which improves the security of the ECC algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention 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 invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1 A flow chart of a method for securely transmitting enterprise procurement data based on big data provided by the present invention;

[0044] Figure 2 Build a flow chart for the metrics that encrypt data. DETAILED DESCRIPTION

[0045] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation method, structure, features and effects of the secure transmission method of enterprise procurement data based on big data proposed by the present invention 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.

[0046] 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 invention belongs.

[0047] The specific scheme of the method for securely transmitting enterprise procurement data based on big data provided by the present invention is described in detail below with reference to the accompanying drawings.

[0048] An embodiment of the present invention provides a method for securely transmitting enterprise procurement data based on big data.

[0049] Specifically, the following method for secure transmission of enterprise procurement data based on big data is provided. Please refer to Figure 1 , the method comprises the following steps:

[0050] Step S001, obtaining the enterprise's procurement data to be transmitted and preprocessing the data.

[0051] In order to achieve secure transmission of enterprise procurement data, this embodiment first obtains the enterprise's procurement data to be transmitted through the enterprise's procurement system database, and re-encodes the procurement data using UTF-16 encoding to obtain pre-processed procurement data.

[0052] Step S002, improving the initial cluster center of the clustering algorithm by the density and distance difference of each character in the procurement data, thereby screening out characters with greater differences relative to the overall.

[0053] Since the procurement data of enterprises is highly private, the data is usually encrypted before data transmission. When processing the procurement data of enterprises, sometimes it is necessary to replace the data for the purpose of privacy protection, data desensitization or pre-processing analysis. For example, replace the data characters according to the relationship between the similarities of the two in the overall data. When cracking the password, the characters are close in binary data, which makes the cracking more difficult and more secure.

[0054] Count all kinds of characters in the enterprise procurement data, form the binary data of all character categories into a character binary set BD, and count the frequency of each element in the character binary set in the procurement data to form a character frequency sequence NB.

[0055] Since the similarity between characters is calculated, the binary difference string between any two characters is obtained through XOR operation, which is recorded as lbd ij , whose length is 32. Get lbd ij The subscript index values ​​of all elements in the array whose values ​​are 1 are recorded as the non-zero subscript sequence y ijBased on this, the position difference between any two characters is calculated through the difference position of the binary data of the characters.

[0056]

[0057] In the formula, CPd ij Indicates the position difference between the i-th character and the j-th character, c ij Represents the non-zero subscript sequence y ij The number of elements in; exp() represents the exponential function with the natural constant e as the base; y ij,l ,y ij,l-1 represents the lth and l-1th element values ​​in the non-zero subscript sequence, wherein, in this embodiment, Recorded as the first sum.

[0058] It should be noted that the higher the difference between two characters, the more non-zero elements there are in the binary difference string obtained by the two characters, that is, c ij The larger the value, the greater the difference between the two at the bit level. The richness of its non-zero elements reflects the heterogeneity of the two at the information level. The relative position between non-zero elements also affects the evaluation of the difference. The greater the distance between non-zero elements, the greater the structural deviation between characters i and j. And the degree of difference increases exponentially with the increase of the distance between non-zero elements, that is, exp(y ij,l -y ij,l-1 ). The larger the value of the position difference between the i-th character and the j-th character calculated based on this, the more difficult it is to crack the private key constructed by these two characters, and the better the data security.

[0059] Due to CPd ij It only reflects the degree of difference between the i-th character and the j-th character. To find the global difference separation of the i-th character in the procurement data, it is necessary to calculate the position difference between the i-th character and all other characters. The global difference separation of the characters is calculated by all position differences and the frequency of occurrence of the characters in the procurement data.

[0060]

[0061] In the formula, gdd i represents the global difference separation of the i-th character, NB i represents the frequency of the i-th character in the procurement data; n represents the number of character types in the procurement data; CPd ij represents the position difference between the i-th character and the j-th character, where Recorded as the second sum.

[0062] It should be noted that the greater the difference between the i-th character and other characters, the greater the difference between the i-th character and other characters. The larger the value, the lower the number of occurrences of the i-th character in the enterprise procurement data. The character is more discrete relative to the normal mode of the entire data set, indicating that the character has a higher degree of difference in the world. The global difference separation degree gdd of the i-th character in the enterprise procurement transmission data is calculated. i The larger the value of .

[0063] In order to obtain character data in the purchase data that is more complex and has higher security for constructing private key data, this embodiment clusters the characters in all the purchase data, thereby clustering the characters with greater differences into one category, which is convenient for constructing private keys. However, when the K-means algorithm is used to classify character data, since there may be relatively scattered outliers in the data, if the outliers are selected as the initial clustering center, it will have a greater impact on the clustering result. Therefore, it is necessary to calculate a threshold to reduce the impact of outliers on the selection of the initial clustering center, so as to improve the accuracy of the algorithm.

[0064] Calculate the neighbor sequence of the global difference separation of all characters, where the neighbor sequence of each character is composed of the global difference separation of the first m characters with the smallest difference between each character and all the remaining characters, sorted from small to large in numerical value. In this embodiment, m is set to 10, and the implementer can set it according to the actual situation.

[0065] At the same time, select the median of all neighbor sequences, and then form the median set MS with all the medians. Calculate the mean of the data between the first and third quartiles of the median set MS, and record it as the quartile distribution mean IRQ μ The quartile calculation method is a well-known technique and will not be described in detail in this embodiment. The threshold a of the global difference separation degree of characters is calculated through the above steps.

[0066]

[0067] Where a represents the limit threshold; MS represents the median set; IRQ μ It represents the mean of the data between the first quartile and the third quartile in the median set MS, that is, the mean of the quartile distribution; max(), min(), medi() represent the maximum value, minimum value, and median function respectively. Among them, medi(MS)-min(MS) is recorded as the first difference, and max(MS)-medi(MS) is recorded as the second difference.

[0068] It should be noted that the quartiles of the data reflect the remainder of the concentrated part of the data. When the data is between the first quartile and the third quartile, the mean IRQ of the data μThe larger it is, the larger the data values ​​are. When the difference between the median and the minimum value of the data is smaller than the difference between the median and the maximum value, that is, The smaller it is, the higher the data concentration is between the minimum value and the median, and the calculated threshold a of the global difference separation of characters should be biased towards the minimum value, and vice versa.

[0069] Calculate the density distance of the global difference separation of all characters. The calculation method is: first obtain the Kth neighbor distance of the global difference separation of the characters, K takes an empirical value of 5, and then divide the Kth neighbor distance by K to obtain the density distance of the global difference separation of the characters.

[0070] Select the global difference separation degree with the smallest density distance, and use it as the first initial cluster center of the data. Calculate the distance value between the global difference separation degree of the remaining characters and it, where the distance value is obtained by the absolute value after subtracting the two values. The distance values ​​of all remaining characters are combined into a set L; the density distances of the global difference separation degrees of all remaining characters are combined into a density distance sequence DS. The cluster center selectivity of each character is calculated through the above steps.

[0071]

[0072]

[0073] Where, L ′ m Indicates the selection component of the mth character; L m represents the distance between the global difference separation of the mth character and the first initial cluster center; a represents the limiting threshold; exp() represents an exponential function with the natural constant e as the base.

[0074] csd m Indicates the cluster center selection degree of the mth character; DS m Represents the density distance of the mth character in the density distance sequence.

[0075] It should be noted that when the distance between the global difference separation degree of the remaining characters except the first initial cluster center in the enterprise's procurement data and the value of the first initial cluster center does not exceed the threshold, it is more likely to be selected as the next cluster center; and when the global difference separation degree of the characters in the enterprise's procurement data is in a more concentrated area, the DS of its density distance m The smaller the value, the more likely it is to be selected as the next cluster center.

[0076] Select the character with the smallest density distance among all characters, and record the global difference separation degree of this character as the first center A; select the character with the largest cluster center selectivity among all characters, and record the global difference separation degree of this character as the second center B.

[0077] It should be noted that since A is the value with the smallest density distance in the global distribution, it means that A has the highest density in the global distribution, which means that A must be in a cluster, not an outlier. B is chosen because the farther away from A, the more likely it is to be the next cluster center, but considering that this point may be an outlier, the distance between A and B, as well as the density of A are taken into account. The calculated A removes the influence of outliers, making A farther from B and in a region with higher density, which reduces the number of iterations of the algorithm during clustering and increases the calculation speed.

[0078] Furthermore, the global difference separation of all characters is used as the input of the improved K-means algorithm, and points A and B are used as the initial clustering centers of the K-means algorithm to improve the K-means algorithm, and the output results are two clusters. Among them, the cluster with the largest element mean in the cluster is recorded as the difference cluster. The characters corresponding to the elements in the difference cluster are greatly different from the overall global ones, and the security is higher.

[0079] Step S003, the elements obtained through the difference cluster correspond to the corresponding characters, construct the private key of the ECC algorithm, and complete the encryption of the enterprise procurement data to be transmitted.

[0080] By selecting the difference cluster obtained in the above steps, 16 characters are randomly extracted from the difference cluster, and the extracted characters are combined into a 256-bit binary data in the order of extraction.

[0081] Furthermore, the enterprise procurement data to be transmitted is used as the input of the improved ECC algorithm, and the 256-bit binary data is used as the private key of the ECC algorithm to improve the ECC algorithm. The output of the algorithm is encrypted data. The indicator construction flow chart of encrypted data is as follows: Figure 2 As shown in the figure, secure transmission of corporate procurement data is ensured by transmitting encrypted data.

[0082] At this point, this embodiment is completed.

[0083] In summary, the embodiment of the present invention obtains the position difference by calculating the corresponding difference position relationship between the binary data of any two characters, and reflects the complexity of constructing the private key based on the position difference between any two characters, which is conducive to constructing a private key with higher complexity through the position difference relationship between the characters, enhancing the difficulty of cracking the private key and improving the security of the data; combining the position difference between the characters with the frequency of the characters appearing in the procurement data, the global difference separation of the characters is calculated, and the global difference separation not only takes into account the number of times the characters appear in the data, that is, the characters are relatively discrete relative to the normal mode of the entire data set, but also takes into account the distance between the binary data differences of the characters, which has the beneficial effect of improving the accuracy of calculating the difference of the characters in the global, making the consideration direction more comprehensive;

[0084] At the same time, the embodiment of the present invention constructs a neighbor sequence of each character based on the global difference separation degree of each character, and mines the overall distribution size characteristics of the data through the median distribution in the neighbor sequence, distinguishes the concentration degree between the data and the maximum value and the minimum value, thereby affecting the limited threshold of the global difference separation degree, avoiding the initial clustering cluster center in the subsequent clustering process from seriously deviating from the mainstream data, and reducing the influence of extreme data values ​​on data calculation; the embodiment of the present invention calculates the density distance of each character based on the K nearest neighbor data between the global difference separation degrees of each character, and constructs the cluster center selectivity of each character in combination with the limited threshold and density distance, and jointly evaluates the selection of each character from the threshold level and the spatial distribution density level. The possibility of being the next cluster center makes the evaluation result more reasonable; by taking the character with the smallest density distance among all characters as the cluster center, and considering the distance and outlier factors at the same time, the character with the largest cluster center selectivity is taken as another cluster center, so as to obtain the optimized clustering effect, which greatly reduces the number of iterations of the clustering algorithm and improves the calculation speed. Moreover, through the restrictions on density and distance, the K-means clustering effect can be improved, so that the characters with large differences from the overall characters can be screened out, thereby improving the security of the encryption algorithm; based on the construction of private keys by randomly extracting characters from the difference clusters obtained by clustering, it has the characteristics of real-time, randomness and high obfuscation, which improves the security of the ECC algorithm.

[0085] It should be noted that the sequence of the above embodiments of the present invention is only for description 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.

[0086] 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.

[0087] 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 secure transmission of enterprise procurement data based on big data, characterized in that: The method comprises the following steps: Obtain the procurement data to be transmitted by the enterprise; For each character in the procurement data, the position difference between any two characters is obtained based on the position difference between the binary data of any two characters; the global difference separation of each character is obtained based on the position difference between each character and the remaining characters and the degree of repetition of each character; the neighbor sequence of each character is obtained based on the global difference separation of each character; the median set and the quartile distribution mean are obtained based on the median distribution of the neighbor sequence of each character; the limiting threshold is obtained by combining the quartile distribution mean and the element distribution in the median set; Based on the global difference separation of each character, the density distance of each character is obtained; based on the limited threshold, the density distance of each character and the global difference separation, the cluster center selectivity of each character is obtained; based on the maximum cluster center selectivity and the minimum density distance, the clustering algorithm is optimized to obtain the difference cluster; The characters corresponding to the randomly selected elements in the difference cluster are combined into binary data in the order of selection. The binary data is used as the private key and the ECC algorithm is used to encrypt the procurement data to be transmitted by the enterprise, thus completing the secure transmission of the enterprise procurement data. The step of obtaining the position difference between any two characters according to the position difference between binary data of any two characters comprises: The binary data of any two characters are calculated by using an XOR operation, and the calculation result is recorded as a binary difference string; the subscript index values ​​of all elements in the binary difference string whose values ​​are 1 are obtained to form a non-zero subscript sequence; For each element of the non-zero subscript sequence, calculate the difference between the element and the left adjacent element, use the difference as the exponent of an exponential function with a natural constant as the base, calculate the sum of the calculation results of the exponential function for all elements of the non-zero subscript sequence, and record it as the first sum; The number of elements in the non-zero subscript sequence is obtained; and the product of the first sum value and the number of elements is used as the position difference between any two characters.

2. The method for securely transmitting enterprise procurement data based on big data as claimed in claim 1, characterized in that: The step of obtaining the global difference separation degree of each character according to the position difference degree between each character and the remaining characters and the repetition degree of each character includes: The sum of the position differences between each character and all remaining characters is obtained and recorded as a second sum; the frequency of occurrence of each character in the procurement data is obtained; and the ratio of the second sum to the frequency is used as the global difference separation degree of each character.

3. The method for securely transmitting enterprise procurement data based on big data as claimed in claim 2, characterized in that: The step of obtaining a neighbor sequence of each character based on the global difference separation degree of each character includes: Calculate the absolute value of the difference between the global difference separation of each character and the remaining characters in the purchased data; All the absolute values ​​of the differences of each character are sorted from small to large, and the global difference separation degrees of the first m elements after sorting are arranged in order from small to large to form the neighbor sequence of each character.

4. The method for securely transmitting enterprise procurement data based on big data as claimed in claim 3, characterized in that: The method of obtaining a median set and a quartile distribution mean according to the median distribution of each character neighbor sequence includes: Get the median of the neighbor sequence of each character, and form a median set with the medians of all characters; Get the mean of all data between the first quartile and the third quartile in the median set, recorded as the quartile distribution mean.

5. The method for secure transmission of enterprise procurement data based on big data as claimed in claim 1, characterized in that: The step of obtaining a limiting threshold by combining the quartile distribution mean and the element distribution in the median set includes: Get the maximum, minimum and median of the median set; Calculate the difference between the median and the minimum value as the first difference; calculate the difference between the maximum value and the median as the second difference; The ratio of the first difference to the second difference is calculated, and the product of the ratio and the mean of the quartile distribution is used as a limiting threshold.

6. The method for secure transmission of enterprise procurement data based on big data as claimed in claim 1, characterized in that: The step of obtaining the density distance of each character based on the global difference separation degree of each character includes: Calculate the absolute value of the difference between each character and the global difference separation degree of all remaining characters in the purchased data; All the absolute values ​​of the differences of each character are sorted from small to large, and the Kth absolute value of the difference in the sorting result is divided by the value K to obtain the density distance of each character.

7. The method for securely transmitting enterprise procurement data based on big data as claimed in claim 1, characterized in that: The step of obtaining the cluster center selectivity of each character according to the defined threshold value, the density distance of each character, and the global difference separation degree includes: The global difference separation degree of the character to which the minimum density distance belongs is used as the first initial cluster center; the density distances of all the remaining characters except the character to which the minimum density distance belongs are used to form a density distance sequence; For each element of the density distance sequence, the selection component of each character is obtained by combining the limited threshold and the difference between the global difference separation degree of each element and the first initial cluster center; the ratio of the selection component of each character to the density distance is used as the cluster center selectivity of each character.

8. The method for securely transmitting enterprise procurement data based on big data as claimed in claim 7, characterized in that: The step of combining the defined threshold and the difference between the global difference separation degree of each element and the first initial cluster center to obtain the selection component of each character includes: Calculate the absolute value of the difference between the global difference separation degree of each character and the first initial cluster center value as the distance value of each character; When the distance value is less than or equal to a specified threshold, the distance value of the character is used as the selection component of the corresponding character; otherwise, the difference between the specified threshold and the distance value of the character is used as the exponent of an exponential function with a natural constant as the base, and the calculation result of the exponential function is used as the selection component of the corresponding character.

9. The method for secure transmission of enterprise procurement data based on big data as claimed in claim 1, characterized in that: The method of optimizing the clustering algorithm according to the maximum cluster center selection degree and the minimum density distance to obtain the difference clusters includes: The global difference separation degree of the character with the minimum density distance is recorded as the first center; the global difference separation degree of the character with the maximum cluster center selectivity is recorded as the second center; The global difference separation of all characters is used as the input of the clustering algorithm, the first center and the second center are used as the initial clustering centers of the clustering algorithm, and the output of the clustering algorithm is two clusters; the cluster with the largest element mean is recorded as the difference cluster.

Images