Internet data analysis system and method thereof

Abstract

The invention discloses an Internet data analysis system and method, and relates to the technical field of data analysis, and the method comprises the steps: collecting the original data of the Internet, carrying out the encryption through quantum key distribution and a post-quantum cryptography algorithm, and generating an encryption-protected data package; decrypting the data packet through a decryption engine, recovering an original data format by using a quantum key, cleaning, and generating a structured data set; based on the data set, loading an adversarial generative network model, creating a virtual sample, and generating an extended data set; performing feature extraction on the extended version data set by adopting a deep learning framework to form a feature vector; building an analysis model through a vector machine according to the feature vectors; the analysis model is applied to predict to-be-analyzed data, a conversion result is in a visual form, and visual display is generated. By integrating the quantum security technology and the artificial intelligence algorithm, the efficiency and accuracy of data analysis are remarkably improved while the data security is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and particularly to an Internet data analysis system and method thereof. Background Art

[0002] With the rapid development of the Internet, the acquisition and analysis of data have become an important part of modern information society. Traditional data analysis methods mostly rely on centralized data storage and processing. However, in the face of massive and diverse Internet data, traditional methods are unable to cope effectively in terms of data security, privacy protection, real-time performance, etc. Especially during the data transmission process, the risks of data being maliciously tampered with and leaked are becoming increasingly prominent, which not only affects the integrity and credibility of the data, but also limits the application scenarios of data analysis.

[0003] The existing technologies mainly adopt symmetric encryption and asymmetric encryption algorithms to ensure the security of data transmission. However, in the face of the rapid development of quantum computing, the security of traditional encryption algorithms faces severe challenges. In addition, existing data cleaning and feature extraction methods often fail to effectively handle data heterogeneity and scale, resulting in the accuracy and effectiveness of analysis results being affected. Therefore, there is an urgent need for a new data analysis method that can improve the efficiency and accuracy of analysis while ensuring data security. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an Internet data analysis method to solve the problems of insufficient data security and low processing efficiency in the process of Internet data analysis.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides an Internet data analysis method, which includes collecting the original data of the Internet, encrypting it using quantum key distribution and post-quantum cryptography algorithms to generate an encrypted protected data packet; Decrypting the data packet through a decryption engine, restoring the original data format using the quantum key and cleaning it to generate a structured data set; Based on the data set, loading an adversarial generative network model, creating virtual samples, and generating an extended data set; Performing feature extraction on the extended data set using a deep learning framework to form feature vectors; According to the feature vectors, constructing an analysis model through a vector machine; Applying the analysis model to predict the data to be analyzed, converting the result into an intuitive form, and generating a visual display.

[0007] As a preferred embodiment of the Internet data analysis method of the present invention, the steps are as follows: collect the original data of the Internet, encrypt it using quantum key distribution and post-quantum cryptography algorithms to generate encrypted protected data packets. Specifically: Use API interfaces and web crawler tools to capture text, images, and video content from Internet resources; Take the text, images, and video content as the original data and transfer it to a temporary storage area; Start the quantum key distribution protocol, and use the AES-256 algorithm and quantum key to encrypt the original data in the temporary storage area block by block; On the basis of encryption, introduce the NTRU post-quantum cryptography algorithm to encrypt the original data again to generate encrypted data packets; Pack the encrypted data packets into the TLS 1.3 protocol and send them to the decryption engine through an HTTPS connection.

[0008] As a preferred embodiment of the Internet data analysis method of the present invention, the steps are as follows: decrypt the data packets through the decryption engine, use the quantum key to restore the original data format and clean it to generate a structured data set. Specifically: The decryption engine receives the encrypted data packets transmitted through HTTPS; Decrypt the encrypted data in the encrypted data packets using the AES-256 symmetric decryption algorithm to generate partially decrypted data blocks; Apply the NTRU post-quantum cryptography algorithm to decrypt the partially decrypted data blocks to restore the original data; Convert the original data back to the original text, image, and video formats; Clean the original data in the restored format, process missing values, duplicates, and outliers, and organize it into an original data set.

[0009] As a preferred embodiment of the Internet data analysis method of the present invention, the steps are as follows: based on the data set, load the adversarial generative network model, create virtual samples, and generate an extended data set. Specifically: Perform quantization processing on the original data set, and represent the data samples in the original data set as quantum states through quantum superposition; In the loaded adversarial network framework model, use quantum computing to optimize the generative network and the discriminative network; The generative network and the discriminative network compete against each other through quantum optimization to create virtual samples; The virtual samples pass through the quantum extension function and are added to the original data set, so that the data samples in the original data set are combined with quantum states. The expression of the quantum extension function is: ; ; Among them, is the original data set, is the quantized sample set, is the number of generated virtual samples, is the transformation function of the quantum generation network, is for the th transformation parameter of the sample, is the [[ID=1⑨]]th quantum state of the virtual sample, is the sample index variable in the original data set, is the Hamiltonian of the quantum operation; Adjust the probability amplitude of the quantum state using the transformation function to generate quantum transformation virtual samples; Fuse the quantum transformation virtual samples with the original data set to generate an extended data set.

[0010] As a preferred solution of the Internet data analysis method described in the present invention, among them: Use a deep learning framework to extract features from the extended data set to form feature vectors. The specific steps are as follows. Select a deep learning framework based on the data sample type in the extended data set; Input the extended data set into the deep learning framework, process each data sample one by one, and extract high-level semantic features; Through global pooling operations, convert the extracted high-level semantic features into fixed-length feature vectors.

[0011] As a preferred solution of the Internet data analysis method described in the present invention, among them: According to the feature vectors, construct an analysis model through a vector machine. The specific steps are as follows. Map the feature vectors to the quantum state space to form a superposition state of qubits; Use the superposition state of qubits to construct a quantum optimized support vector machine model; In the quantum optimized support vector machine model, introduce the objective function of the support vector machine into quantum computing for optimization, and solve for the optimal hyperplane in the quantum state space, and the expression is: ; Among them, represents the inner product of the sample in the quantum space, is the bias term, is the normal vector of the hyperplane; For the quantum optimized support vector machine model, use the quantum gradient descent algorithm for training to optimize the normal vector of the hyperplane and the bias term , and the quantum gradient descent formula is: ; Among them, is the loss function, is the learning rate, is the gradient of the loss function with respect to the hyperplane normal vector, is the gradient of the loss function with respect to the bias term, represents the update amount of the bias term, represents the gradient of the hyperplane normal vector, represents the gradient of the bias term; After the training is completed, the quantum optimized support vector machine model classifies and predicts the data samples of the feature vectors through the discriminant function, and the discriminant function expression is: ; Among them, is the sign function, is the th sample label; Verify the generalization ability based on the prediction results to complete the construction of the analysis model.

[0012] As a preferred solution of the Internet data analysis method described in the present invention, wherein: the analysis model is applied to predict the data to be analyzed, the conversion result is in an intuitive form, and a visual display is generated. The specific steps are as follows: The input data undergoes the same preprocessing and feature extraction process as the training data to obtain the data to be analyzed and the feature vectors; Input the data to be analyzed and the feature vectors into the analysis model, perform forward propagation calculation, and obtain the prediction results of each sample; Post-process the prediction results to obtain the characteristics of the prediction results; According to the characteristics of the prediction results, select Plotly as the visualization tool and design a visualization scheme; Based on the visualization scheme, use Plotly to convert the prediction results into intuitive charts and graphs.

[0013] In a second aspect, the present invention provides an Internet data analysis system, including a data collection and encryption module, a data decryption module, a sample expansion module, a feature extraction module, a model construction module, and a result visualization module; the data collection and encryption module is used to collect the original data of the Internet, encrypt it using quantum key distribution and post-quantum cryptography algorithms, and generate encrypted and protected data packets; the data decryption module is used to decrypt the data packets through a decryption engine, recover the original data format using quantum keys and clean it, and generate a structured data set; the sample expansion module is used to, based on the data set, load an adversarial generative network model, create virtual samples, and generate an expanded data set; the feature extraction module is used to extract features from the expanded data set using a deep learning framework to form feature vectors; the model construction module is used to, according to the feature vectors, construct an analysis model through a vector machine; the result visualization module is used to apply the analysis model to predict the data to be analyzed, convert the result into an intuitive form, and generate a visual display.

[0014] In a third aspect, the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program is executed by the processor, any step of the Internet data analysis method as described in the first aspect of the present invention is implemented.

[0015] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program is executed by the processor, any step of the Internet data analysis method as described in the first aspect of the present invention is implemented.

[0016] The beneficial effects of the present invention are as follows: By introducing quantum key distribution and post-quantum cryptography algorithms, the present invention significantly improves the security during data transmission. This dual encryption mechanism not only effectively resists traditional attack means but also provides a guarantee against future security threats brought by quantum computing. In addition, the method of creating an expanded data set by combining an adversarial generative network model can generate virtual samples in the case of scarce data, enrich the training data, and improve the generalization ability of the model. In the feature extraction stage, using a deep learning framework can mine high-level semantic features in the data to form representative feature vectors. By visualizing the analysis results, users can intuitively understand the information behind the data. Through this series of innovative steps, the present invention not only solves the core problems in the current data analysis field but also produces significant beneficial effects in terms of data security, processing efficiency, and result visualization. Description of the Drawings

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0018] Figure 1 It is a flowchart of the Internet data analysis method in Embodiment 1.

[0019] Figure 2 It is a module diagram of the Internet data analysis system in Embodiment 1. Detailed implementation manners

[0020] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific implementation manners of the present invention with reference to the drawings in the specification.

[0021] In the following description, many specific details are set forth to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0022] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that mutually excludes other embodiments.

[0023] Embodiment 1, referring to Figure 1 and Figure 2 , is the first embodiment of the present invention. This embodiment provides an Internet data analysis method, including the following steps: S1. Collect the original data of the Internet, encrypt it using quantum key distribution and post-quantum cryptography algorithms, and generate encrypted protected data packets; Furthermore, grab text, images, and video content from Internet resources through API interfaces and web crawler tools; Through multi-channel data collection, ensure the diversity and comprehensiveness of the data. Use web crawler tools to grab the latest information in real time to keep the data set updated, improving timeliness and relevance. After this step, various types of data, including text, images, and videos, can be efficiently obtained from the Internet. This diverse source of original data helps to build a rich data set, providing a broad information foundation.

[0024] Transfer text, image, and video content as raw data to a temporary storage area; Specifically, the temporary storage area is selected to use a database for centralized storage and management. A storage structure is designed according to the data type (text, image, video), and the data is directly transferred to the temporary storage area using an API interface. Data management is implemented within the temporary storage area, permissions are set to ensure that only authorized users can access and modify the data in the storage area, and the data in the temporary storage is backed up regularly to prevent loss or damage.

[0025] Start the quantum key distribution protocol and use the AES-256 algorithm and quantum key to encrypt the raw data in the temporary storage area block by block; Preferably, quantum key distribution and the AES-256 encryption algorithm are introduced, providing double protection for data security. Through quantum key distribution, the secure generation and transmission of keys are ensured, resisting future quantum computing attacks and providing advanced security guarantees. Using the AES-256 algorithm for block-by-block encryption can improve processing efficiency and reduce the risk of being intercepted during data transmission.

[0026] On the basis of encryption, introduce the NTRU post-quantum cryptography algorithm to encrypt the raw data again to generate encrypted data packets; Among them, the NTRU algorithm is designed specifically to resist quantum computing attacks, enhancing the security of data in future environments. Through secondary encryption using the NTRU post-quantum cryptography algorithm, the security of the data is further enhanced. The double encryption means form a more complex encrypted data packet, increasing the difficulty of data decryption and thus enhancing the overall security.

[0027] Pack the encrypted data packets into the TLS 1.3 protocol and send them to the decryption engine through an HTTPS connection.

[0028] S2. Decrypt the data packets through the decryption engine, use the quantum key to restore the original data format and clean it, and generate a structured data set; Furthermore, the decryption engine receives the encrypted data packets transmitted through HTTPS; Through HTTPS transmission, it is ensured that the data is not stolen or tampered with during transmission, guaranteeing the confidentiality and integrity of the data. Using the standard HTTPS protocol ensures that the decryption engine can communicate effectively with various clients and servers, enhancing the applicability of the system.

[0029] Decrypt the encrypted data in the encrypted data packets using the AES-256 symmetric decryption algorithm to generate partially decrypted data blocks; Specifically, in this process, the AES-256 decryption algorithm is used to process the encrypted data packet. As a symmetric encryption algorithm, AES-256 has a high decryption speed, can quickly generate partially decrypted data blocks, reduce latency, and its strong encryption strength ensures that the security of the data is maintained even during partial decryption, providing a guarantee for subsequent decryption.

[0030] Apply the NTRU post-quantum cryptography algorithm to decrypt the partially decrypted data blocks and restore the original data; Convert the original data back to the original text, image, and video formats; Among them, converting the data back to the original format enables the data to be effectively utilized by subsequent applications and analysis tools, enhancing the practicality of the data. Realizing the data conversion from the encrypted state to the available state allows the data to be correctly interpreted and utilized.

[0031] Clean the original data in the restored format, handle missing values, duplicates, and outliers, and organize it into an original data set.

[0032] Specifically, in the process of cleaning the original data in the restored format, first identify and handle missing values, and solve the missing problems by deleting and filling the mean, median, and mode; then find and delete duplicates, keep unique records or merge records with different values; then identify outliers and use statistical methods to handle them, and handle outliers by correction, deletion, and replacement. Finally, organize the cleaned data into a structured data set, clarify the characteristics and data types of each column, and at the same time conduct result verification and documentation to ensure data quality and facilitate subsequent analysis.

[0033] S3. Based on the data set, load the adversarial generative network model, create virtual samples, and generate an extended data set; Furthermore, perform quantization processing on the original data set, and represent the data samples in the original data set as quantum states through quantum superposition; Among them, quantization processing is the process of converting binary-represented data into quantum states, and uses quantum characteristics such as quantum superposition and entanglement to improve data processing capabilities. Quantum superposition is the phenomenon of being able to be in multiple states simultaneously, allowing parallel processing of multiple data samples and increasing the expressive ability of information. Through quantum superposition processing, data samples can be represented in multiple states simultaneously, enhancing the representation ability and complexity of data samples. Quantized data can better capture the complex relationships between data, providing a richer information basis for subsequent model training.

[0034] In loading the adversarial network framework model, use quantum computing to optimize the generative network and discriminative network; Specifically, the adversarial network framework model is a deep learning model composed of a generator network and a discriminator network. The generator network attempts to create real samples, while the discriminator network is responsible for judging the authenticity of the samples. The introduction of quantum computing makes the training process of the generator network and the discriminator network more efficient, enabling a higher convergence speed in a short time. The optimized generator network and discriminator network can generate more realistic samples, improving the overall performance of the model.

[0035] The generator network and the discriminator network compete against each other through quantum optimization to create virtual samples; It should be noted that the adversarial process between the generator network and the discriminator network is strengthened by quantum optimization. The generator network attempts to create virtual samples as realistic as possible to deceive the discriminator network, while the discriminator network tries to distinguish between real and fake. This quantum-enhanced adversarial mechanism improves the quality and diversity of virtual samples, promotes the co-evolution between the two networks, helps to explore a wider area of the data distribution, and achieves the effect of generating more complex and diverse virtual samples.

[0036] The virtual samples are added to the original dataset through the quantum expansion function so that the data samples in the original dataset are combined with quantum states. The expression of the quantum expansion function is: ; ; where is the original dataset, is the set of quantized samples, is the number of generated virtual samples, is the transformation function of the quantum generator network, is for the transformation parameter of the th sample, is the quantum state of the th virtual sample, is the sample index variable in the original dataset, is the Hamiltonian of the quantum operation;

[0037] The probability amplitude of the quantum state is adjusted using the transformation function to generate quantum-transformed virtual samples; Preferably, the probability amplitude of the quantum state is adjusted through a transformation function to achieve fine control of the characteristics of the virtual samples. This step allows for flexible control of the characteristics of the virtual samples according to requirements, fine-tuning the probability amplitude to generate quantum transformation virtual samples that meet the requirements of specific application scenarios.

[0038] Fuse the quantum transformation virtual samples with the original data set to generate an extended data set.

[0039] Specifically, in the process of fusing the quantum transformation virtual samples with the original data set, first make the diversity and quality of the quantum transformation virtual samples consistent with the original data set, then directly add the virtual samples to the original data set through row splicing and column splicing methods, then clean the extended data set to remove duplicate records, verify data integrity and detect outliers, and finally save the generated extended data set.

[0040] S4. Use a deep learning framework to extract features from the extended data set to form feature vectors; Furthermore, based on the data sample types in the extended data set, select a deep learning framework; According to the feature types of the extended data set, select an appropriate deep learning framework to optimize the training effect. The selected framework can maximize its performance while matching the feature types of the extended data set. For image data, a convolutional neural network is selected, and for sequence data, a recurrent neural network and a transformer architecture are selected. By choosing the appropriate deep learning framework, the efficiency and accuracy of the model can be improved.

[0041] Input the extended data set into the deep learning framework, process each data sample one by one, and extract high-level semantic features; Among them, high-level semantic features refer to high-level information extracted from data samples. Processing data samples one by one can make full use of the capabilities of the deep learning model to extract high-level semantic features from the samples. These features can capture the complex patterns and internal relationships of the data, providing strong support for subsequent decision-making. The extracted high-level semantic features can be used for various tasks such as classification and regression, further improving the intelligent level.

[0042] Through global pooling operations, convert the extracted high-level semantic features into fixed-length feature vectors.

[0043] Preferably, global pooling operations are a downsampling technique. By globally aggregating the feature map, they can convert the multi-dimensional feature map into a fixed-length feature vector, simplify the complexity of subsequent processing, reduce the risk of overfitting, and improve computational efficiency. The fixed-length feature vector is convenient for subsequent machine learning tasks.

[0044] S5. According to the feature vectors, construct an analysis model through a vector machine; Furthermore, the feature vector is mapped to the quantum state space to form a superposition state of qubits; Among them, the quantum state space refers to the space where qubits are located. Through characteristics such as superposition and entanglement, the quantum state can represent information of multiple states. A qubit is the basic unit in quantum computing, similar to a bit in classical computing, but it can be in multiple states simultaneously and has stronger computing power.

[0045] Use the superposition state of qubits to construct a quantum optimized support vector machine model; Specifically, by applying the superposition state characteristics of qubits, the quantization of the classical support vector machine model is carried out to create a support vector machine that can operate within the quantum state space, making full use of the powerful parallelism and fast convergence ability of quantum computing. As a result, a more efficient, accurate and forward-looking classification prediction model is established, which is particularly suitable for solving complex pattern recognition problems and improving the learning efficiency and generalization ability.

[0046] In the quantum optimized support vector machine model, the objective function of the vector machine is introduced into quantum computing for optimization, and the optimal hyperplane is solved within the quantum state space. The expression is: ; Among them, represents the inner product of the sample in the quantum space, is the bias term, is the normal vector of the hyperplane; It should be noted that by incorporating the objective function of the vector machine into the quantum computing framework, the optimization of the support vector machine hyperplane is realized, and the optimal hyperplane is solved within the quantum state space, aiming to minimize the loss function. This approach not only shortens the training time but also enables the quantum optimized support vector machine model to search for the best solution in a wider parameter space. Ultimately, the classification accuracy and robustness are improved.

[0047] For the quantum optimized support vector machine model, use the quantum gradient descent algorithm for training to optimize the normal vector of the hyperplane and the bias term . The quantum gradient descent formula is: ; Among them, is the loss function, is the learning rate, is the gradient of the loss function with respect to the normal vector of the hyperplane, is the gradient of the loss function with respect to the bias term, represents the update amount of the bias term, represents the gradient of the normal vector of the hyperplane, represents the gradient of the bias term; It should be noted that quantum gradient descent is an optimization algorithm based on quantum computing that uses quantum state information to update model parameters and improve training efficiency. The quantum gradient descent algorithm is introduced to optimize the normal vector of the hyperplane and the bias term . Taking advantage of quantum computing, the quantum gradient descent algorithm can quickly find the minimum value of the loss function. Ensure that the quantum optimized support vector machine model can effectively learn from the training data, thereby improving the classification decision boundary, enhancing the classification accuracy, and strengthening the model generalization ability.

[0048] After training, the quantum optimized support vector machine model classifies and predicts the data samples of the feature vector through the discriminant function. The expression of the discriminant function is: ; where is the sign function, is the label of the th sample; Specifically, the discriminant function is a mathematical function for classification tasks that generates an output value based on the input feature vector to determine the class to which the sample belongs. In the quantum optimized support vector machine model, the discriminant function generates the classification result by calculating the inner product of the feature vector and the model parameters and adding the bias term.

[0049] Verify the generalization ability based on the prediction results and complete the construction of the analysis model.

[0050] Preferably, by analyzing the prediction results, the generalization ability of the model is evaluated to ensure its stable and reliable performance on unseen data, demonstrating the effectiveness and practicality of the proposed quantum optimized support vector machine. An efficient and reliable analysis model is constructed, providing solid theoretical and technical support for practical applications.

[0051] S6. Use the analysis model to predict the data to be analyzed, convert the results into an intuitive form, and generate a visual display.

[0052] Furthermore, the input data undergoes the same preprocessing and feature extraction processes as the training data to obtain the data to be analyzed and the feature vector; By applying the same preprocessing and feature extraction processes as the training data, the input data is standardized. The input data can be processed and feature extracted in the same way, avoiding biases caused by different data formats and feature representations. Provide a unified data basis for subsequent model predictions, ensuring the reliability and consistency of the prediction results. Improve the prediction accuracy and stability, enabling the data to be analyzed to seamlessly integrate into the analysis model.

[0053] Input the data to be analyzed and the feature vectors into the analysis model, perform forward propagation calculations, and obtain the prediction results for each sample; Among them, in a neural network, forward propagation calculation refers to the process of processing the input features through each layer of the model and finally outputting the prediction results. By passing the data to be analyzed to each layer of neurons, forward propagation calculation can effectively generate the prediction results for each sample. This process utilizes the complex structure of the neural network to capture the non-linear relationships and patterns in the data, achieving the effects of efficient and real-time prediction, and providing users with immediate insights and support. Perform post-processing on the prediction results to obtain the characteristics of the prediction results; Specifically, post-processing is to further analyze and process the prediction results output by the model to extract useful information and characteristics. Calculate the confidence of each prediction result, usually expressed as the probability of the predicted class, identify possible outliers or noises in the prediction results, perform feature importance analysis through the LIME (Local Interpretable Model-agnostic Explanation) method, and use the output of the model to evaluate the contribution degree of each input feature to the prediction results.

[0054] Select Plotly as the visualization tool based on the characteristics of the prediction results and design a visualization scheme; Among them, Plotly is an open-source interactive chart library that supports multiple programming languages and can create static and dynamic charts, including scatter plots, line charts, bar charts, and heatmaps. Customize the visualization scheme according to the characteristics of the prediction results. For classification problems, select the confusion matrix and ROC curve. For regression problems, use the residual plot and the comparison plot with actual values. Use Plotly's Python library to create the charts and embed the generated charts into web applications and Jupyter Notebook pages for sharing and demonstration.

[0055] Based on the visualization scheme, use Plotly to convert the prediction results into intuitive charts and graphs.

[0056] Preferably, by implementing the designed visualization scheme, use Plotly to convert the prediction results into intuitive charts and graphs. Convert the abstract data and predictions into visual expressions for easy observation and analysis. Plotly simplifies the communication of complex information, enhances the ability to tell data stories, makes the prediction results more acceptable and understandable, and promotes decision-makers to make wise choices. Present the complex prediction results to users in an intuitive and easy-to-understand form, greatly enhancing the value and influence of the data.

[0057] This embodiment also provides an Internet data analysis system, including: a data collection and encryption module, a data decryption module, a sample expansion module, a feature extraction module, a model construction module, and a result visualization module; the data collection and encryption module is used to collect the original data of the Internet, encrypt it using quantum key distribution and post-quantum cryptography algorithms, and generate an encrypted protected data packet; the data decryption module is used to decrypt the data packet through a decryption engine, restore the original data format using a quantum key and clean it, and generate a structured data set; the sample expansion module is used to load an adversarial generation network model based on the data set, create virtual samples, and generate an expanded data set; the feature extraction module is used to extract features from the expanded data set using a deep learning framework to form feature vectors; the model construction module is used to construct an analysis model according to the feature vectors through a vector machine; the result visualization module is used to apply the analysis model to predict the data to be analyzed, convert the result into an intuitive form, and generate a visual display.

[0058] This embodiment also provides a computer device, applicable to the situation of the Internet data analysis method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the Internet data analysis method proposed in the above embodiment.

[0059] This computer device can be a terminal. This computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of this computer device is used to provide computing and control capabilities. The memory of this computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of this computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be achieved through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of this computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of this computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0060] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the Internet data analysis method proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, abbreviated as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, abbreviated as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, abbreviated as EPROM), programmable read-only memory (Programmable Red-Only Memory, abbreviated as PROM), read-only memory (Read-Only Memory, abbreviated as ROM), magnetic memory, flash memory, magnetic disk or optical disc.

[0061] In summary, the present invention: by introducing quantum key distribution and post-quantum cryptography algorithms, significantly improves the security during data transmission. This dual encryption mechanism not only effectively resists traditional attack means, but also provides protection against future security threats brought by quantum computing. In addition, the method for creating an extended dataset in combination with the adversarial generative network model can generate virtual samples in the case of scarce data, enrich the training data, and improve the generalization ability of the model. In the feature extraction stage, the deep learning framework can be used to mine high-level semantic features in the data and form representative feature vectors. By visualizing the analysis results, users can intuitively understand the information behind the data. Through this series of innovative steps, the present invention not only solves the core problems in the current data analysis field, but also produces significant beneficial effects in terms of data security, processing efficiency and result visualization.

[0062] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. An Internet data analysis method, characterized in that: Including, Collecting the original data from the Internet, encrypting it using quantum key distribution and post-quantum cryptography algorithms to generate encrypted protected data packets; Decrypting the data packets through a decryption engine, using quantum keys to restore the original data format and cleaning it to generate a structured data set; Based on the data set, loading an adversarial generative network model, creating virtual samples, and generating an extended data set; Performing feature extraction on the extended data set using a deep learning framework to form feature vectors; According to the feature vectors, constructing an analysis model through a vector machine; Applying the analysis model to predict the data to be analyzed, converting the results into an intuitive form, and generating a visual display.

2. The Internet data analysis method according to claim 1, characterized in that: The specific steps for collecting the original data from the Internet, encrypting it using quantum key distribution and post-quantum cryptography algorithms to generate encrypted protected data packets are as follows. Grabbing text, images, and video content from Internet resources through an API interface and a web crawler tool; Taking the text, images, and video content as the original data and transmitting it to a temporary storage area; Starting the quantum key distribution protocol, using the AES-256 algorithm and quantum keys to encrypt the original data in the temporary storage area block by block; On the basis of encryption, introducing the NTRU post-quantum cryptography algorithm to encrypt the original data again to generate encrypted data packets; Packing the encrypted data packets into the TLS 1.3 protocol and sending them to the decryption engine through an HTTPS connection.

3. The Internet data analysis method according to claim 2, characterized in that: The specific steps for decrypting the data packets through a decryption engine, using quantum keys to restore the original data format and cleaning it to generate a structured data set are as follows. The decryption engine receives the encrypted data packets transmitted through HTTPS; Decrypting the encrypted data in the encrypted data packets using the AES-256 symmetric decryption algorithm to generate partially decrypted data blocks; Applying the NTRU post-quantum cryptography algorithm to decrypt the partially decrypted data blocks to restore the original data; Converting the original data back to the original text, image, and video formats; Cleaning the original data in the restored format, handling missing values, duplicates, and outliers, and organizing it into an original data set.

4. The Internet data analysis method according to claim 3, wherein: The specific steps for loading an adversarial generative network model based on the data set, creating virtual samples, and generating an extended data set are as follows. Performing quantization processing on the original data set, representing the data samples in the original data set as quantum states through quantum superposition; In the loaded adversarial network framework model, using quantum computing to optimize the generative network and the discriminative network; The generative network and the discriminative network compete against each other through quantum optimization to create virtual samples; The virtual samples are added to the original dataset through the quantum expansion function to enable the data samples in the original dataset to combine quantum states. The expression of the quantum expansion function is as follows: ； ； Among them, is the original data set, is the quantized sample set, is the number of generated virtual samples, is the transformation function of the quantum generation network, is the transformation parameter of the th sample, is the quantum state of the th sample in the original data set, is the Hamiltonian of the quantum operation; Using a transformation function to adjust the probability amplitude of the quantum state to generate quantum transformation virtual samples; Fusing the quantum transformation virtual samples with the original data set to generate an extended data set.

5. The Internet data analysis method according to claim 4, wherein: The specific steps for performing feature extraction on the extended data set using a deep learning framework to form feature vectors are as follows. Selecting a deep learning framework based on the data sample types in the extended data set; Inputting the extended data set into the deep learning framework, processing each data sample one by one, and extracting high-level semantic features; Through global pooling operations, converting the extracted high-level semantic features into fixed-length feature vectors.

6. The Internet data analysis method according to claim 5, characterized in that: According to the feature vectors, an analysis model is constructed through a vector machine. The specific steps are as follows: Map the feature vectors to the quantum state space to form a superposition state of qubits; Use the superposition state of qubits to construct a quantum optimized support vector machine model; In the quantum optimized support vector machine model, introduce the objective function of the vector machine into quantum computing for optimization, and solve for the optimal hyperplane in the quantum state space. The expression is: ； Among them, represents the inner product of the sample in the quantum space, is the bias term, is the hyperplane normal vector; For the quantum optimized support vector machine model, the quantum gradient descent algorithm is used for training to optimize the normal vector of the hyperplane and the bias term , and the quantum gradient descent formula is: ； Among them, is the loss function, is the learning rate, is the gradient of the loss function with respect to the hyperplane normal vector, is the gradient of the loss function with respect to the bias term, represents the update amount of the bias term, represents the gradient of the hyperplane normal vector, represents the gradient of the bias term; After training, the quantum optimized support vector machine model classifies and predicts the data samples of the feature vectors through a discriminant function. The expression of the discriminant function is: ； Among them, is the sign function, is the label of the th sample; Verify the generalization ability based on the prediction results to complete the construction of the analysis model.

7. The Internet data analysis method according to claim 6, wherein: Apply the analysis model to predict the data to be analyzed, convert the result into an intuitive form, and generate a visual display. The specific steps are as follows: The input data undergoes the same preprocessing and feature extraction processes as the training data to obtain the data to be analyzed and the feature vectors; Input the data to be analyzed and the feature vectors into the analysis model, perform forward propagation calculations, and obtain the prediction results for each sample; Post-process the prediction results to obtain the characteristics of the prediction results; According to the characteristics of the prediction results, select Plotly as the visualization tool and design a visualization scheme; Based on the visualization scheme, use Plotly to convert the prediction results into intuitive charts and graphs.

8. An Internet data analysis system, based on the Internet data analysis method according to any one of claims 1 to 7, characterized in that: Including a data collection and encryption module, a data decryption module, a sample expansion module, a feature extraction module, a model construction module, and a result visualization module; The data collection and encryption module is used to collect the original data from the Internet, encrypt it using quantum key distribution and post-quantum cryptography algorithms, and generate encrypted protected data packets; The data decryption module is used to decrypt the data packets through a decryption engine, restore the original data format using quantum keys and clean it, and generate a structured data set; The sample expansion module is used to load an adversarial generation network model based on the data set, create virtual samples, and generate an expanded data set; The feature extraction module is used to extract features from the expanded data set using a deep learning framework to form feature vectors; The model construction module is used to construct an analysis model through a vector machine according to the feature vectors; The result visualization module is used to apply the analysis model to predict the data to be analyzed, convert the result into an intuitive form, and generate a visual display.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the Internet data analysis method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the Internet data analysis method according to any one of claims 1 to 7.

Images