A method for exchanging and processing computer data

Abstract

This application discloses a method for exchanging and processing computer data, relating to the field of data processing technology. By segmenting the symmetric encryption key using a key splitting algorithm and combining it with a public key transmission scheme, the difficulty of key cracking increases exponentially, and the method also possesses a certain degree of key loss resistance, greatly improving the security of the computer data exchange process. Furthermore, during the data access process by the data requester, the method performs real-time network security detection, determines the network security detection result, and executes data transmission based on the network security detection result. This effectively prevents unauthorized intrusion into the database, thereby avoiding the loss of computer data and further enhancing the security of the computer data exchange process.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and more specifically, to a method for exchanging and processing computer data. Background Technology

[0002] In today's highly interconnected digital age, data exchange between computer systems has become commonplace. Whether it's different departments within an enterprise, cloud service providers and users, or partner systems across geographical regions, frequent data transfer and sharing are essential. Traditional data exchange methods, such as simple file transfers, direct database connections, or data push using generic APIs (Application Programming Interfaces), often have limitations. They are vulnerable to attacks during transmission. While typical symmetric-key encryption can quickly encrypt large amounts of data, its keys are easily cracked, leading to low data security and the risk of data leakage. Summary of the Invention

[0003] This application aims to provide a method for exchanging and processing computer data, which addresses the problems of low data security and easy leakage in existing computer data exchange processes.

[0004] This application provides a method for exchanging and processing computer data, including:

[0005] Obtain the computer data to be exchanged from the data transmission party, and encrypt the computer data to be exchanged using a symmetric encryption algorithm to determine the encrypted computer data and its corresponding symmetric encryption key;

[0006] The symmetric encryption key is segmented using a key instantaneous segmentation algorithm to determine multiple different symmetric encryption key information;

[0007] Determine the target number corresponding to the symmetric encryption key information, and request multiple public keys corresponding to the target number from the data receiver; wherein, the private key corresponding to each public key is stored locally on the data receiver's local machine;

[0008] The symmetric encryption key information is encrypted using a one-to-one correspondence between the public key and the symmetric encryption key information to determine the encrypted symmetric encryption key information;

[0009] The encrypted computer data and the encrypted symmetric encryption key information are associated and stored to obtain the target exchange data;

[0010] During the process of the data requester requesting data exchange with the target, network security is detected in real time, and the network security detection result is determined; wherein, the network security detection result is either network security or network insecurity;

[0011] If the network security detection result indicates that the network security is secure, the target exchange data is transmitted to the data requester so that the data requester can use the locally stored private key to recover the data.

[0012] In one possible implementation, it also includes:

[0013] If the network security test result is deemed insecure, the data access function of the data requester will be restricted.

[0014] In one possible implementation, the process involves acquiring the computer data to be exchanged from the data transmission party, encrypting the computer data using a symmetric encryption algorithm, and determining the encrypted computer data and its corresponding symmetric encryption key, including:

[0015] Obtain the data to be exchanged from the computer of the data transmission party;

[0016] The AES encryption algorithm is used to encrypt the computer data to be exchanged, and the encrypted computer data and its corresponding symmetric encryption key are determined.

[0017] In one possible implementation, the symmetric encryption key is segmented using a key instantaneous splitting algorithm to determine multiple distinct symmetric encryption key information, including:

[0018] The symmetric encryption key is converted into a decimal number to obtain the target value corresponding to the symmetric encryption key.

[0019] Construct a K-order polynomial with the target value as a constant term, and randomly generate at least 2K different random numbers. Use the random numbers as input to the K-order polynomial and determine the polynomial value corresponding to each random number.

[0020] The random number and its corresponding polynomial value are used together as the symmetric encryption key information to obtain 2K symmetric encryption key information.

[0021] In one possible implementation, real-time network security detection and determination of network security detection results include:

[0022] Network traffic characteristics are collected in real time and input into a network security detection model for identification to determine the network security detection result.

[0023] In one possible implementation, before inputting the network traffic characteristics into the network security detection model for identification, the method further includes:

[0024] A deep learning model is constructed, and an improved chaotic mapping strategy is used to initialize the hyperparameters of the deep learning model to determine multiple different hyperparameter vectors;

[0025] Obtain the loss function value corresponding to each hyperparameter vector, and determine the first optimal vector, the second optimal vector, and the third optimal vector based on the loss function value;

[0026] Based on the first optimal vector, a variety of initial search strategies are used to perform an initial search on the hyperparameter vector to determine the hyperparameter vector after the initial search.

[0027] Based on the first optimal vector, the second optimal vector, and the third optimal vector, a cooperative encirclement search strategy is adopted to conduct a cooperative search on the hyperparameter vector after the initial search, and the hyperparameter vector after the cooperative search is determined.

[0028] A global mutation strategy is used to perform a global search on the hyperparameter vectors after the cooperative search, and the hyperparameter vectors after the global search are determined.

[0029] Determine if the current number of training iterations is greater than or equal to the maximum number of training iterations. If so, determine the target optimal vector based on the hyperparameter vector after global search. Otherwise, return to the steps of determining the first optimal vector, the second optimal vector, and the third optimal vector.

[0030] The hyperparameters in the target optimal vector are used as the final hyperparameters of the deep learning model to obtain a network security detection model with network security identification capabilities.

[0031] In one possible implementation, an improved chaotic mapping strategy is used to initialize the hyperparameters of the deep learning model, determining multiple different hyperparameter vectors, including:

[0032] The hyperparameters of the deep learning model are randomly initialized between the upper and lower bounds, and the initialized hyperparameters are encoded into vectors to determine a basic vector.

[0033] Based on the aforementioned base vector, several different hyperparameter vectors are obtained as follows:

[0034]

[0035] in, Indicates the first i The th hyperparameter vector of the th ... d dimensional hyperparameters, and i When it is 1, Represents the first fundamental vector d dimensional hyperparameters, d =1,2,…,D, where D represents the total dimension of the hyperparameters. Indicates the first i +1 hyperparameter vectors d dimensional hyperparameters, Represents pi (π). This represents the arcsine function.

[0036] In one possible implementation, based on the first optimal vector, a diversified initial search strategy is used to perform an initial search on the hyperparameter vector, and the hyperparameter vector after the initial search is determined as follows:

[0037]

[0038]

[0039]

[0040] in, It represents the training process of the th iteration. j The th hyperparameter vector of the th ... d dimensional hyperparameters, j =1,2,…,M, where M represents the total number of hyperparameter vectors. d =1,2,…,D, where D represents the total dimension of the hyperparameters. Indicates the first j The hyperparameter vector after the initial search is the first... d dimensional hyperparameters, Denotes the first hyperparameter vector of a random hyperparameter vector. d dimensional hyperparameters, Represents the first random number between (0,1). This represents the second random number between (0,1), and | represents the absolute value sign. The first optimal vector represents the first optimal vector. d dimensional hyperparameters, Indicates a variable learning rate. This represents a third random number between (0,1). This represents the fourth random number between (0,1). Represents a random decision factor between (0,1). Represents the nonlinear convergence factor. This represents the fifth random number between (0,1). Let represent the initial value of the nonlinear convergence factor, e represent the natural constant, and T represent the maximum number of training iterations. Represents a constant term. Represents the beta distribution function. Denotes the parameters of the first beta distribution. This represents the second beta distribution parameter.

[0041] In one possible implementation, based on the first optimal vector, the second optimal vector, and the third optimal vector, a cooperative encirclement search strategy is used to conduct a cooperative search on the hyperparameter vectors after the initial search, determining the hyperparameter vectors after the cooperative search, including:

[0042] Based on the first optimal vector, the second optimal vector, and the third optimal vector, the first distance parameter, the second distance parameter, and the third distance parameter between the vector and the hyperparameter vector after the initial search are determined as follows:

[0043]

[0044]

[0045]

[0046] in, It represents the training process of the th iteration. m The hyperparameter vector after the initial search is the first... d dimensional hyperparameters, m =1,2,…,M, express and The first distance parameter between them This represents the sixth random number between (0,1). This represents the seventh random number between (0,1). The second optimal vector represents the first... d dimensional hyperparameters, express and The second distance parameter between them This represents the eighth random number between (0,1). This represents the ninth random number between (0,1). The third optimal vector represents the first... d dimensional hyperparameters, express and The third distance parameter between them Represents the tenth random number between (0,1). Represents the eleventh random number between (0, 1);

[0047] Based on the first distance parameter, the second distance parameter, and the third distance parameter, the first reverse search probability, the second reverse search probability, and the third reverse search probability are determined as follows:

[0048]

[0049]

[0050]

[0051] in, Indicates the probability of the first reverse search. Indicates the probability of the second reverse search. Let represent the probability of the third reverse search, and exp represent an exponential function with the natural constant e as the base.

[0052] Based on the first optimal vector, the second optimal vector, the third optimal vector, the first reverse search probability, the second reverse search probability, and the third reverse search probability, the first cooperation information, the second cooperation information, and the third cooperation information are determined as follows:

[0053]

[0054]

[0055]

[0056] in, Indicates the first collaborative information. The first optimal vector represents the backward learning vector of the first optimal vector. d dimensional hyperparameters, This represents the twelfth random number between (0,1). Indicates the second collaborative information. The second optimal vector represents the backward learning vector of the second optimal vector. d dimensional hyperparameters, This represents the thirteenth random number between (0,1). Indicates third-party collaborative information. The third optimal vector represents the backward learning vector of the third optimal vector. d dimensional hyperparameters, This represents the fourteenth random number between (0, 1);

[0057] Based on the first cooperation information, the second cooperation information, and the third cooperation information, the hyperparameter vector after the initial search is determined to be the following through cooperative search:

[0058]

[0059] in, Indicates the first m The hyperparameter vector after the first collaborative search is the first... d Dimensional hyperparameters.

[0060] In one possible implementation, a global mutation strategy is used to perform a global search on the hyperparameter vectors after the cooperative search, and the hyperparameter vectors after the global search are determined, including:

[0061] For any hyperparameter vector obtained after a collaborative search, two other hyperparameter vectors are randomly matched to obtain the first differential evolution vector and the second differential evolution vector corresponding to the hyperparameter vector.

[0062] Based on the first and second differential evolution vectors, a global search is performed on the hyperparameter vectors after the cooperative search to determine the global search vector as follows:

[0063]

[0064] in, Indicates the first iter During the training process, the first n The hyperparameter vector after the first collaborative search is the first... d dimensional hyperparameters, n =1,2,…,M, Indicates the first n A global search vector, Indicates the first n The first difference evolution vector corresponding to the hyperparameter vector after the collaborative search is the first... d dimensional hyperparameters, Indicates the first n The second differential evolution vector corresponding to the hyperparameter vector after the first collaborative search. d dimensional hyperparameters, This represents the fifteenth random number between (0,1). This represents the sixteenth random number between (0, 1). Indicates the first scaling factor. Indicates the second scaling factor;

[0065] Determine if the loss function value of the global search vector is less than the loss function value of the hyperparameter vector after the corresponding cooperative search. If so, use the global search vector as the hyperparameter vector after the global search; otherwise, determine the probability of accepting the inferior solution as follows:

[0066]

[0067] in, This represents the probability of accepting an inferior solution. This represents the maximum probability of accepting a suboptimal solution. This represents the minimum probability of accepting a suboptimal solution. Represents the natural constant;

[0068] Based on the acceptance probability of the inferior solution, a roulette wheel strategy is adopted to use the hyperparameter vector after the global search or the hyperparameter vector after the cooperative search as the hyperparameter vector after the global search.

[0069] Beneficial effects:

[0070] This application provides a method for exchanging and processing computer data. By segmenting the symmetric encryption key using a key splitting algorithm and combining it with a public key transmission scheme, the difficulty of key cracking increases exponentially, and the method also possesses a certain degree of key loss resistance. This greatly enhances the security of the computer data exchange process. Furthermore, during the data access process by the data requester, the method performs real-time network security detection, determines the network security detection result, and executes data transmission based on the network security detection result. This effectively prevents unauthorized intrusion into the database, thereby avoiding the loss of computer data and further enhancing the security of the computer data exchange process. Attached Figure Description

[0071] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application 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.

[0072] Figure 1 This is a flowchart of a computer data exchange and processing method proposed in an embodiment of this application.

[0073] Figure 2 This is a flowchart of obtaining a network security detection model according to an embodiment of this application. Detailed Implementation

[0074] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0075] like Figure 1 As shown in the figure, this application provides a method for exchanging and processing computer data. This method is applied in a cloud server or a data storage server and may include:

[0076] S101. Obtain the computer data to be exchanged from the data transmission party, and encrypt the computer data to be exchanged using a symmetric encryption algorithm, and determine the encrypted computer data and its corresponding symmetric encryption key.

[0077] Obtaining the computer data to be exchanged from the data transmission party refers to the process by which the data transmission party uploads the computer data to be exchanged. The data encryption method used in this upload process can be the same as the data encryption method used when transmitting data to the data receiver, and will not be elaborated here.

[0078] After encrypting the computer data to be exchanged using a symmetric encryption algorithm, the encrypted computer data is determined, which can achieve fast encryption of large amounts of data and ensure data transmission latency.

[0079] S102. The symmetric encryption key is segmented using a key instantaneous segmentation algorithm to determine multiple different symmetric encryption key information.

[0080] In existing technologies, the symmetric encryption key is usually directly encrypted using the public key of a cloud server or data storage server before transmission. Although this can achieve high-speed encryption of data and secure transmission of the symmetric encryption key to a certain extent, it still has the following shortcomings: the symmetric encryption key may still be cracked during transmission, and data may be lost during transmission, resulting in the inability to decrypt successfully after data transmission.

[0081] Therefore, the embodiments of this application use a key instantaneous segmentation algorithm to segment the symmetric encryption key, which can effectively prevent the symmetric encryption key from being cracked in one go, and increase the difficulty for unauthorized attackers to crack it.

[0082] S103. Determine the target number corresponding to the symmetric encryption key information, and request multiple public keys corresponding to the target number from the data receiver; wherein, the private key corresponding to each public key is stored locally on the data receiver's machine.

[0083] Public and private keys can be generated using some common asymmetric encryption algorithms.

[0084] S104. Encrypt the symmetric encryption key information using a one-to-one correspondence between the public key and the symmetric encryption key information, and determine the encrypted symmetric encryption key information.

[0085] By segmenting the symmetric encryption key using a key splitting algorithm and combining it with a public key transmission scheme, the difficulty of cracking the key increases exponentially, and the key also has a certain degree of resistance to loss, greatly improving the security of computer data exchange.

[0086] S105. The encrypted computer data and the encrypted symmetric encryption key information are associated and stored to obtain the target exchange data.

[0087] Optionally, since there are multiple encrypted symmetric encryption key information, the association between the public key and the encrypted symmetric encryption key information can also be stored, thereby facilitating the subsequent decryption of the encrypted symmetric encryption key information.

[0088] S106. During the process of the data requester requesting data exchange with the target, network security is detected in real time, and the network security detection result is determined; wherein the network security detection result is network security or network insecurity.

[0089] Deep learning algorithms can be used to detect network security in real time, thereby determining the network security detection results. These results often indicate whether the data requester is accessing data normally, providing a basis for the management of the exchange process.

[0090] S107. If the network security detection result is network security, the target exchange data is transmitted to the data requester so that the data requester can use the locally stored private key to recover the data.

[0091] For example, multiple encrypted symmetric encryption key information can be decrypted using a locally stored private key to obtain multiple symmetric encryption key information, and the symmetric encryption key can be recovered from the symmetric encryption key information. Finally, the encrypted computer data can be decrypted using the symmetric encryption key to obtain the plaintext data of the computer data to be exchanged.

[0092] This application provides a method for exchanging and processing computer data. By segmenting the symmetric encryption key using a key splitting algorithm and combining it with a public key transmission scheme, the difficulty of key cracking increases exponentially, and the method also possesses a certain degree of key loss resistance. This greatly enhances the security of the computer data exchange process. Furthermore, during the data access process by the data requester, the method performs real-time network security detection, determines the network security detection result, and executes data transmission based on the network security detection result. This effectively prevents unauthorized intrusion into the database, thereby avoiding the loss of computer data and further enhancing the security of the computer data exchange process.

[0093] In one possible implementation, it also includes:

[0094] If the network security test result is deemed insecure, the data access function of the data requester will be restricted.

[0095] The function of restricting data access for data requesters can generally be: restricting data requesters' access for a certain period of time or permanently restricting data requesters' access.

[0096] In one possible implementation, the process involves acquiring the computer data to be exchanged from the data transmission party, encrypting the computer data using a symmetric encryption algorithm, and determining the encrypted computer data and its corresponding symmetric encryption key, including:

[0097] Obtain the computer data to be exchanged from the data transmission party.

[0098] The computer data to be exchanged is encrypted using the AES (Advanced Encryption Standard) encryption algorithm to determine the encrypted computer data and its corresponding symmetric encryption key.

[0099] It is worth noting that using the AES encryption algorithm to encrypt the computer data to be exchanged is merely an example of an embodiment of this application; other symmetric encryption algorithms can also be used for encryption.

[0100] In one possible implementation, the symmetric encryption key is segmented using a key instantaneous splitting algorithm to determine multiple distinct symmetric encryption key information, including:

[0101] The symmetric encryption key is converted into a decimal number to obtain the target value corresponding to the symmetric encryption key.

[0102] Construct a K-order polynomial with the target value as a constant term, and randomly generate at least 2K different random numbers. Use these random numbers as input to the K-order polynomial and determine the polynomial value corresponding to each random number.

[0103] The random number and its corresponding polynomial value are used together as the symmetric encryption key information to obtain 2K symmetric encryption key information.

[0104] As demonstrated by the above scheme, since the K-order polynomial contains only one variable (i.e., a random number) and one dependent variable (i.e., the polynomial value), only K+1 symmetric encryption key information is needed to recover the K-order polynomial, thereby obtaining the target value corresponding to the symmetric encryption key. Finally, the target value corresponding to the symmetric encryption key can be recovered to the symmetric encryption key. Therefore, during the transmission of the encrypted symmetric encryption key information, the number of keys can be freely chosen from K+1 to 2K, and decryption can be achieved in all cases.

[0105] In one possible implementation, real-time network security detection and determination of network security detection results include:

[0106] Network traffic characteristics are collected in real time and input into a network security detection model for identification to determine the network security detection result.

[0107] like Figure 2As shown, before inputting the network traffic characteristics into the network security detection model for identification, the following steps are also included:

[0108] S201. Construct a deep learning model and use an improved chaotic mapping strategy to initialize the hyperparameters of the deep learning model, thereby determining multiple different hyperparameter vectors.

[0109] For example, convolutional neural networks can be used to build deep learning models. However, it is worth noting that, in addition to convolutional neural networks, other deep learning models with recognition capabilities can also be used.

[0110] In the prior art, random initialization or chaotic mapping strategies are generally used to initialize the hyperparameters of deep learning models. Although these strategies can achieve the initialization effect, the initial solutions are poorly distributed in the solution space. Therefore, this application provides an improved chaotic mapping strategy to improve the uniformity of the initial solution distribution and enhance the training effect.

[0111] S202. Obtain the loss function value corresponding to each hyperparameter vector, and determine the first optimal vector, the second optimal vector, and the third optimal vector based on the loss function value.

[0112] For example, the KDD CUP 99 dataset can be used as training data. After applying the hyperparameters contained in the hyperparameter vector to the deep learning model, the root mean square loss function value or cross-entropy loss function value corresponding to the hyperparameter vector can be obtained using the training data. The loss function value corresponding to the hyperparameter vector can then be determined. The hyperparameter vector with the smallest loss function value is then determined as the first optimal vector, the hyperparameter vector with the second smallest loss function value is determined as the second optimal vector, and the hyperparameter vector with the third smallest loss function value is determined as the third optimal vector.

[0113] S203. Based on the first optimal vector, a diversified initial search strategy is adopted to perform an initial search on the hyperparameter vector to determine the hyperparameter vector after the initial search.

[0114] S204. Based on the first optimal vector, the second optimal vector, and the third optimal vector, a cooperative encirclement search strategy is adopted to conduct a cooperative search on the hyperparameter vectors after the initial search, and the hyperparameter vectors after the cooperative search are determined.

[0115] S205. Use a global mutation strategy to perform a global search on the hyperparameter vector after the cooperative search, and determine the hyperparameter vector after the global search.

[0116] S206. Determine whether the current number of training iterations is greater than or equal to the maximum number of training iterations. If so, determine the target optimal vector based on the hyperparameter vector after global search. Otherwise, return to the steps of determining the first optimal vector, the second optimal vector, and the third optimal vector.

[0117] S207. The hyperparameters in the target optimal vector are used as the final hyperparameters of the deep learning model to obtain a network security detection model with network security identification capabilities.

[0118] Optionally, after the hyperparameter vector changes, boundary overflow handling can be performed to ensure that the hyperparameter vector always remains within its corresponding upper and lower bounds. For example, hyperparameters exceeding the upper bound can be set to their corresponding upper bound value, hyperparameters exceeding the lower bound can be set to their corresponding lower bound value, or hyperparameters exceeding the upper bound can be randomly regenerated within their upper and lower bounds.

[0119] In existing technologies, gradient descent optimization algorithms are generally used to optimize the hyperparameters of deep learning models. While this can achieve some optimization effect, it often suffers from getting trapped in local optima, leading to poor training results and consequently, the trained network security detection model being unable to accurately identify network security threats. Therefore, this application proposes a multi-strategy optimization algorithm consisting of an improved chaotic mapping strategy, a diversified initial search strategy, a cooperative encirclement search strategy, and a global mutation strategy. This multi-strategy optimization algorithm is used to train the hyperparameters of the deep learning model, thereby avoiding getting trapped in local optima, improving the training effect and accuracy of the algorithm, and ultimately enhancing the identification accuracy of the network security detection model.

[0120] In one possible implementation, an improved chaotic mapping strategy is used to initialize the hyperparameters of the deep learning model, determining multiple different hyperparameter vectors, including:

[0121] The hyperparameters of the deep learning model are randomly initialized between the upper and lower bounds, and the initialized hyperparameters are encoded into vectors to determine a base vector.

[0122] Based on the aforementioned base vector, several different hyperparameter vectors are obtained as follows:

[0123]

[0124] in, Indicates the first i The th hyperparameter vector of the th ... d dimensional hyperparameters, and i When it is 1, Represents the first fundamental vector d dimensional hyperparameters, d =1,2,…,D, where D represents the total dimension of the hyperparameters. Indicates the first i +1 hyperparameter vectors d dimensional hyperparameters, Represents pi (π). This represents the arcsine function.

[0125] While existing chaotic mapping algorithms (such as logistic mapping) offer slightly better results than random initialization, they also result in a higher probability of hyperparameters taking values ​​at their upper or lower bounds, which is detrimental to finding the global optimum. Therefore, this application provides an improved chaotic mapping strategy that distributes hyperparameters more evenly during initialization, thereby enhancing global search capabilities.

[0126] In one possible implementation, based on the first optimal vector, a diversified initial search strategy is used to perform an initial search on the hyperparameter vector, and the hyperparameter vector after the initial search is determined as follows:

[0127]

[0128]

[0129]

[0130] in, It represents the training process of the th iteration. j The th hyperparameter vector of the th ... d dimensional hyperparameters, j =1,2,…,M, where M represents the total number of hyperparameter vectors. d =1,2,…,D, where D represents the total dimension of the hyperparameters. Indicates the first j The hyperparameter vector after the initial search is the first... d dimensional hyperparameters, Denotes the first hyperparameter vector of a random hyperparameter vector. d dimensional hyperparameters, Represents the first random number between (0,1). This represents the second random number between (0,1), and | represents the absolute value sign. The first optimal vector represents the first optimal vector. d dimensional hyperparameters, Indicates a variable learning rate. This represents a third random number between (0,1). This represents the fourth random number between (0,1). Represents a random decision factor between (0,1). Represents the nonlinear convergence factor. This represents the fifth random number between (0,1). The initial value of the nonlinear convergence factor can be set to 2.5. e represents the natural constant, and T represents the maximum number of training iterations. This represents a constant term and can be set to 0.05. Represents the beta distribution function. This represents the first beta distribution parameter, which can be set to 4. This represents the second beta distribution parameter, which can be set to 1.

[0131] The diverse initial search strategies provided in this application expand the search range by searching towards the location of random hyperparameter vectors with a certain probability, thereby increasing the probability of finding food. As the algorithm converges, all hyperparameter vectors gradually converge together, ensuring convergence even in the later stages. Furthermore, by setting a non-linear convergence factor, the algorithm gains more refined search capabilities in the later stages, improving convergence accuracy.

[0132] In one possible implementation, based on the first optimal vector, the second optimal vector, and the third optimal vector, a cooperative encirclement search strategy is used to conduct a cooperative search on the hyperparameter vectors after the initial search, determining the hyperparameter vectors after the cooperative search, including:

[0133] Based on the first optimal vector, the second optimal vector, and the third optimal vector, the first distance parameter, the second distance parameter, and the third distance parameter between the vector and the hyperparameter vector after the initial search are determined as follows:

[0134]

[0135]

[0136]

[0137] in, It represents the training process of the th iteration. m The hyperparameter vector after the initial search is the first... d dimensional hyperparameters, m =1,2,…,M, express and The first distance parameter between them This represents the sixth random number between (0,1). This represents the seventh random number between (0,1). The second optimal vector represents the first... d dimensional hyperparameters, express and The second distance parameter between them This represents the eighth random number between (0,1). This represents the ninth random number between (0,1). The third optimal vector represents the first... d dimensional hyperparameters, express and The third distance parameter between them Represents the tenth random number between (0,1). Represents the eleventh random number between (0,1).

[0138] Based on the first distance parameter, the second distance parameter, and the third distance parameter, the first reverse search probability, the second reverse search probability, and the third reverse search probability are determined as follows:

[0139]

[0140]

[0141]

[0142] in, Indicates the probability of the first reverse search. Indicates the probability of the second reverse search. Let represent the probability of the third reverse search, and exp represent the exponential function with the natural constant e as the base.

[0143] Based on the first optimal vector, the second optimal vector, the third optimal vector, the first reverse search probability, the second reverse search probability, and the third reverse search probability, the first cooperation information, the second cooperation information, and the third cooperation information are determined as follows:

[0144]

[0145]

[0146]

[0147] in, Indicates the first collaborative information. The first optimal vector represents the backward learning vector of the first optimal vector. d dimensional hyperparameters, This represents the twelfth random number between (0,1). Indicates the second collaborative information. The second optimal vector represents the backward learning vector of the second optimal vector. d dimensional hyperparameters, This represents the thirteenth random number between (0,1). Indicates third-party collaborative information. The third optimal vector represents the backward learning vector of the third optimal vector. d dimensional hyperparameters, This represents the fourteenth random number between (0,1).

[0148] Based on the first cooperation information, the second cooperation information, and the third cooperation information, the hyperparameter vector after the initial search is determined to be the following through cooperative search:

[0149]

[0150] in, Indicates the first m The hyperparameter vector after the first collaborative search is the first... d Dimensional hyperparameters.

[0151] The cooperative encirclement search strategy provided in this application embodiment can search within the region formed by the first optimal vector, the second optimal vector, and the third optimal vector. During the search process, the reverse position information of the first optimal vector, the second optimal vector, and the third optimal vector in the solution space can be obtained with a certain probability, thereby increasing the ability to escape local optima. As the algorithm progresses, the probability of selecting reverse position information will gradually decrease, and the search accuracy will increase, ensuring the convergence of the algorithm.

[0152] In one possible implementation, a global mutation strategy is used to perform a global search on the hyperparameter vectors after the cooperative search, and the hyperparameter vectors after the global search are determined, including:

[0153] For any hyperparameter vector obtained after a collaborative search, two other hyperparameter vectors are randomly matched to obtain the first differential evolution vector and the second differential evolution vector corresponding to the hyperparameter vector.

[0154] Based on the first and second differential evolution vectors, a global search is performed on the hyperparameter vectors after the cooperative search to determine the global search vector as follows:

[0155]

[0156] in, Indicates the first iter During the training process, the first n The hyperparameter vector after the first collaborative search is the first... d dimensional hyperparameters, n =1,2,…,M, Indicates the first n A global search vector, Indicates the first n The first difference evolution vector corresponding to the hyperparameter vector after the collaborative search is the first... d dimensional hyperparameters, Indicates the first n The second differential evolution vector corresponding to the hyperparameter vector after the first collaborative search. d dimensional hyperparameters, This represents the fifteenth random number between (0,1). This represents the sixteenth random number between (0, 1). This represents the first scaling factor, which can be set to 0.5. This represents the second scaling factor, which can be set to 0.5.

[0157] Determine if the loss function value of the global search vector is less than the loss function value of the hyperparameter vector after the corresponding cooperative search. If so, use the global search vector as the hyperparameter vector after the global search; otherwise, determine the probability of accepting the inferior solution as follows:

[0158]

[0159] in, This represents the probability of accepting an inferior solution. This represents the maximum probability of accepting a suboptimal solution, and can be set to 0.9. This represents the minimum probability of accepting a suboptimal solution, which can be set to 0.1. Represents the natural constant.

[0160] Based on the acceptance probability of the inferior solution, a roulette wheel strategy is adopted to use the hyperparameter vector after the global search or the hyperparameter vector after the cooperative search as the hyperparameter vector after the global search.

[0161] The global mutation strategy provided in this application embodiment utilizes the principle of differential evolution to achieve global search. In the early stage of the algorithm, it has a high probability of accepting inferior solutions, thus ensuring the algorithm's global search capability. In the later stage of the algorithm, the probability of accepting inferior solutions gradually decreases, thus ensuring the convergence of the algorithm.

[0162] Compared with the prior art, the multi-strategy optimization algorithm provided in this application comprehensively enhances the global search capability without reducing search accuracy and training speed, enabling the trained network security detection model to accurately identify network security.

[0163] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0164] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, apparatuses, electronic devices, and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0165] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0166] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0167] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0168] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0169] The above provides a detailed description of the hierarchical relationship analysis method for influencing factors provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for exchanging and processing computer data, characterized in that, include: The process involves acquiring the computer data to be exchanged from the data transmission party, encrypting the computer data using a symmetric encryption algorithm, and determining the encrypted computer data and its corresponding symmetric encryption key. The symmetric encryption key is segmented using a key instantaneous segmentation algorithm to determine multiple different symmetric encryption key information, including: converting the symmetric encryption key into a decimal number to obtain the target value corresponding to the symmetric encryption key; constructing a K-order polynomial with the target value as a constant term, and randomly generating at least 2K different random numbers, using the random numbers as input to the K-order polynomial, and determining the polynomial value corresponding to each random number; using the random numbers and the corresponding polynomial values ​​together as symmetric encryption key information to obtain 2K symmetric encryption key information. Determine the target number corresponding to the symmetric encryption key information, and request multiple public keys corresponding to the target number from the data receiver; wherein, the private key corresponding to each public key is stored locally on the data receiver's local machine; The symmetric encryption key information is encrypted using a one-to-one correspondence between the public key and the symmetric encryption key information to determine the encrypted symmetric encryption key information; The encrypted computer data and the encrypted symmetric encryption key information are associated and stored to obtain the target exchange data; During the process of the data requester requesting data exchange with the target, network security is detected in real time, and the network security detection result is determined. This includes: collecting network traffic characteristics in real time, inputting the network traffic characteristics into the network security detection model for identification, and determining the network security detection result; wherein, the network security detection result is either network security or network insecurity, where network insecurity means that the network is under attack, and network security means that the network is not under attack. If the network security detection result indicates that the network security is secure, the target exchange data will be transmitted to the data requester so that the data requester can use the locally stored private key to recover the data. Data recovery using locally stored private keys includes: decrypting multiple encrypted symmetric encryption key information using the locally stored private key to obtain multiple symmetric encryption key information; recovering a K-order polynomial using K+1 symmetric encryption key information to obtain the target value corresponding to the symmetric encryption key, and recovering the target value corresponding to the symmetric encryption key as the symmetric encryption key; and decrypting encrypted computer data using the symmetric encryption key to obtain the plaintext data of the computer data to be exchanged.

2. The computer data exchange and processing method according to claim 1, characterized in that, Also includes: If the network security test result is deemed insecure, the data access function of the data requester will be restricted.

3. The computer data exchange and processing method according to claim 1, characterized in that, The process involves acquiring the computer data to be exchanged from the data transmission provider, encrypting the computer data using a symmetric encryption algorithm, and determining the encrypted computer data and its corresponding symmetric encryption key, including: Obtain the data to be exchanged from the computer of the data transmission party; The AES encryption algorithm is used to encrypt the computer data to be exchanged, and the encrypted computer data and its corresponding symmetric encryption key are determined.

4. The computer data exchange and processing method according to claim 1, characterized in that, Before inputting the network traffic characteristics into the network security detection model for identification, the following steps are also included: A deep learning model is constructed, and an improved chaotic mapping strategy is used to initialize the hyperparameters of the deep learning model to determine multiple different hyperparameter vectors; Obtain the loss function value corresponding to each hyperparameter vector, and determine the first optimal vector, the second optimal vector, and the third optimal vector based on the loss function value; Based on the first optimal vector, a variety of initial search strategies are used to perform an initial search on the hyperparameter vector to determine the hyperparameter vector after the initial search. Based on the first optimal vector, the second optimal vector, and the third optimal vector, a cooperative encirclement search strategy is adopted to conduct a cooperative search on the hyperparameter vector after the initial search, and the hyperparameter vector after the cooperative search is determined. A global mutation strategy is used to perform a global search on the hyperparameter vectors after the cooperative search, and the hyperparameter vectors after the global search are determined. Determine if the current number of training iterations is greater than or equal to the maximum number of training iterations. If so, determine the target optimal vector based on the hyperparameter vector after global search. Otherwise, return to the steps of determining the first optimal vector, the second optimal vector, and the third optimal vector. The hyperparameters in the target optimal vector are used as the final hyperparameters of the deep learning model to obtain a network security detection model with network security identification capabilities.

5. The computer data exchange and processing method according to claim 4, characterized in that, An improved chaotic mapping strategy is used to initialize the hyperparameters of the deep learning model, determining multiple different hyperparameter vectors, including: The hyperparameters of the deep learning model are randomly initialized between the upper and lower bounds, and the initialized hyperparameters are encoded into vectors to determine a basic vector. Based on the aforementioned base vector, several different hyperparameter vectors are obtained as follows: in, Indicates the first i The th hyperparameter vector of the th ... d dimensional hyperparameters, and i When it is 1, Represents the first fundamental vector d dimensional hyperparameters, d =1,2,…,D, where D represents the total dimension of the hyperparameters. Indicates the first i +1 hyperparameter vectors d dimensional hyperparameters, Represents pi (π). This represents the arcsine function.

6. The computer data exchange and processing method according to claim 4, characterized in that, Based on the first optimal vector, a variety of initial search strategies are used to perform an initial search on the hyperparameter vector, and the hyperparameter vector after the initial search is determined as follows: in, It represents the th training iteration during the th iteration. j The th hyperparameter vector of the th ... d dimensional hyperparameters, j =1,2,…,M, where M represents the total number of hyperparameter vectors. d =1,2,…,D, where D represents the total dimension of the hyperparameters. Indicates the first j The hyperparameter vector after the initial search is the first... d dimensional hyperparameters, Denotes the first hyperparameter vector of a random hyperparameter vector. d dimensional hyperparameters, Represents the first random number between (0,1). This represents the second random number between (0,1), and | represents the absolute value sign. The first optimal vector represents the first optimal vector. d dimensional hyperparameters, Indicates a variable learning rate. This represents a third random number between (0,1). This represents the fourth random number between (0,1). Represents a random decision factor between (0,1). Represents the nonlinear convergence factor. This represents the fifth random number between (0,1). Let represent the initial value of the nonlinear convergence factor, e represent the natural constant, and T represent the maximum number of training iterations. Represents a constant term. Represents the beta distribution function. Denotes the parameters of the first beta distribution. This represents the second beta distribution parameter.

7. The computer data exchange and processing method according to claim 6, characterized in that, Based on the first, second, and third optimal vectors, a cooperative encirclement search strategy is employed to conduct a cooperative search of the hyperparameter vectors after the initial search, determining the hyperparameter vectors after the cooperative search, including: Based on the first optimal vector, the second optimal vector, and the third optimal vector, the first distance parameter, the second distance parameter, and the third distance parameter between the vector and the hyperparameter vector after the initial search are determined as follows: in, It represents the th training iteration during the th iteration. m The hyperparameter vector after the initial search is the first... d dimensional hyperparameters, m =1,2,…,M, express and The first distance parameter between them This represents the sixth random number between (0,1). This represents the seventh random number between (0,1). The second optimal vector represents the first... d dimensional hyperparameters, express and The second distance parameter between them This represents the eighth random number between (0,1). This represents the ninth random number between (0,1). The third optimal vector represents the first... d dimensional hyperparameters, express and The third distance parameter between them Represents the tenth random number between (0,1). Represents the eleventh random number between (0, 1); Based on the first distance parameter, the second distance parameter, and the third distance parameter, the first reverse search probability, the second reverse search probability, and the third reverse search probability are determined as follows: in, Indicates the probability of the first reverse search. Indicates the probability of the second reverse search. Let represent the probability of the third reverse search, and exp represent an exponential function with the natural constant e as the base. Based on the first optimal vector, the second optimal vector, the third optimal vector, the first reverse search probability, the second reverse search probability, and the third reverse search probability, the first cooperation information, the second cooperation information, and the third cooperation information are determined as follows: in, Indicates the first collaborative information. The first optimal vector represents the backward learning vector of the first optimal vector. d dimensional hyperparameters, This represents the twelfth random number between (0,1). Indicates the second collaborative information. The second optimal vector represents the backward learning vector of the second optimal vector. d dimensional hyperparameters, This represents the thirteenth random number between (0,1). Indicates third-party collaborative information. The third optimal vector represents the backward learning vector of the third optimal vector. d dimensional hyperparameters, This represents the fourteenth random number between (0, 1); Based on the first cooperation information, the second cooperation information, and the third cooperation information, the hyperparameter vector after the initial search is determined to be the following through cooperative search: in, Indicates the first m The hyperparameter vector after the first collaborative search is the first... d Dimensional hyperparameters.

8. The computer data exchange and processing method according to claim 7, characterized in that, A global mutation strategy is employed to perform a global search on the hyperparameter vectors obtained after the cooperative search, determining the hyperparameter vectors after the global search, including: For any hyperparameter vector obtained after a collaborative search, two other hyperparameter vectors are randomly matched to obtain the first differential evolution vector and the second differential evolution vector corresponding to the hyperparameter vector. Based on the first and second differential evolution vectors, a global search is performed on the hyperparameter vectors after the cooperative search to determine the global search vector as follows: in, Indicates the first iter During the training process, the first n The hyperparameter vector after the first collaborative search is the first... d dimensional hyperparameters, n =1,2,…,M, Indicates the first n A global search vector, Indicates the first n The first difference evolution vector corresponding to the hyperparameter vector after the collaborative search is the first difference evolution vector. d dimensional hyperparameters, Indicates the first n The second differential evolution vector corresponding to the hyperparameter vector after the first collaborative search. d dimensional hyperparameters, This represents the fifteenth random number between (0,1). This represents the sixteenth random number between (0,1). Indicates the first scaling factor. Indicates the second scaling factor; Determine if the loss function value of the global search vector is less than the loss function value of the hyperparameter vector after the corresponding cooperative search. If so, use the global search vector as the hyperparameter vector after the global search; otherwise, determine the probability of accepting the inferior solution as follows: in, This represents the probability of accepting a suboptimal solution. This represents the maximum probability of accepting a suboptimal solution. This represents the minimum probability of accepting a suboptimal solution. Represents the natural constant; Based on the acceptance probability of the inferior solution, a roulette wheel strategy is adopted to use the hyperparameter vector after the global search or the hyperparameter vector after the cooperative search as the hyperparameter vector after the global search.

Images