A deep learning-based password allocation method and system

Abstract

The application provides a password allocation method and system based on deep learning. The password allocation method comprises: collecting user information for each user, and pre-processing the user information to obtain valid information data corresponding to the user; training and testing a deep learning model using the valid information data to obtain a completed test deep learning model, wherein the deep learning model adopts a convolutional neural network model structure; analyzing the correlation degree between users using the completed test deep learning model to obtain a correlation degree result between the users; extracting a user group with a correlation degree lower than a preset correlation degree threshold from the correlation degree result, and allocating the same public key and private key between the user group with a correlation degree lower than the preset correlation degree threshold. The system comprises modules corresponding to the method steps.

Description

Technical Field

[0001] This invention proposes a cryptographic allocation method and system based on deep learning, belonging to the field of cryptographic allocation technology. Background Technology

[0002] Current authentication technologies, such as CA certificates, require different public and private keys for different clients to perform digital signatures and other operations. Assigning a set of public and private keys to each user wastes computing power and storage resources. Summary of the Invention

[0003] This invention provides a deep learning-based cryptographic allocation method and system to address the problem in existing technologies where allocating a single public and private key set to each user wastes both computing power and storage resources. A deep learning-based password allocation method, the password allocation method comprising: Collect user information for each user and preprocess the user information to obtain valid information data corresponding to the user; Using the effective information data, the deep learning model is trained and tested to obtain a deep learning model that has completed the test, wherein the deep learning model adopts a convolutional neural network model structure; Using the deep learning model that has completed the test, the correlation between users is analyzed to obtain the results of the degree of correlation between the users; Extract user groups whose association degree is lower than a preset association degree threshold from the association degree results, and assign the same public and private keys to user groups whose association degree is lower than the preset association degree threshold.

[0004] Furthermore, user information is collected for each user, and the user information is preprocessed to obtain valid information data corresponding to the user, including: Collect user information, wherein the user information includes user identity information and data information on user interaction behavior; The user information is preprocessed to obtain valid information data corresponding to the user. The preprocessing includes data cleaning, data denoising, and data normalization.

[0005] Furthermore, using the deep learning model that has completed the test, the correlation between users is analyzed to obtain the results of the degree of correlation between the users, including: The tested deep learning model is loaded into the system memory, and user data is encoded and features are extracted according to the requirements of the deep learning model to obtain encoded feature data; wherein, the feature data includes user attribute features, user behavior features, social network features, text features, time features and image features. The encoded feature data is input into a deep learning model, and the deep learning model analyzes the feature data to obtain the correlation between users.

[0006] Further, extract user groups whose correlation scores are lower than a preset correlation threshold from the correlation results, and assign the same public and private keys to user groups whose correlation scores are lower than the preset correlation threshold, including: Filter by relevance and set a relevance threshold; Extract the correlation degree between users output by the deep learning model, and compare the correlation degree with the correlation degree threshold; Extract users whose relevance is lower than the relevance threshold and their corresponding relevance; For users whose correlation with each other is lower than the correlation threshold, create the same public and private keys, and assign the public and private keys to the corresponding users.

[0007] A deep learning-based cryptographic distribution system, the cryptographic distribution system comprising: The user information collection module is used to collect user information for each user and preprocess the user information to obtain valid information data corresponding to the user. The training and testing module is used to train and test the deep learning model using the effective information data to obtain a deep learning model that has completed the test, wherein the deep learning model adopts a convolutional neural network model structure. The correlation analysis module is used to analyze the correlation between users using the deep learning model that has completed the test, and to obtain the correlation results between the users. The public and private key allocation module is used to extract user groups whose correlation between users is lower than a preset correlation threshold from the correlation results, and to allocate the same public and private keys among user groups whose correlation is lower than the preset correlation threshold.

[0008] Furthermore, the user information collection module includes: The information collection module is used to collect user information, wherein the user information includes user identity information and interaction behavior data. The preprocessing module is used to preprocess the user information to obtain valid information data corresponding to the user. The preprocessing includes data cleaning, data denoising, and data normalization.

[0009] Furthermore, the correlation analysis module includes: The feature extraction module is used to load the tested deep learning model into the system memory, and encode and extract features from the user data according to the requirements of the deep learning model to obtain encoded feature data; wherein, the feature data includes user attribute features, user behavior features, social network features, text features, time features and image features; The correlation acquisition module is used to input the encoded feature data into a deep learning model, and to obtain the correlation between users by analyzing the feature data through the deep learning model.

[0010] Furthermore, the public and private key distribution module includes: The correlation threshold setting module is used to set the correlation threshold based on the requirements of the correlation filter. The comparison module is used to extract the correlation degree between users output by the deep learning model and compare the correlation degree with the correlation degree threshold. The correlation threshold extraction module is used to extract users whose correlation is lower than the correlation threshold and their corresponding correlation. The public and private key creation and allocation module is used to create the same public and private keys for users whose correlation with each other is lower than the correlation threshold, and to allocate the public and private keys to the corresponding users.

[0011] Beneficial effects of this invention: This invention proposes a deep learning-based cryptographic allocation method and system. This method uses deep learning to analyze the correlation between users. For users with low correlation, the same public and private keys can be assigned, reducing the computational power required for key generation and minimizing storage resource management. The method and system study the connections between users; for users with correlation below a threshold, assigning a single public and private key is harmless and will not cause adverse consequences, significantly reducing resource consumption. The technical solution of this invention achieves the analysis of user correlation and key allocation through steps such as user information collection, deep learning model training and testing, correlation analysis, and key allocation. This technical solution quantifies and analyzes user correlation, identifies user groups with low correlation, and provides these users with the same key to ensure the security and confidentiality of data transmission. In this way, the security and privacy protection level of user data can be improved, while providing users with an efficient and secure encrypted communication environment. Attached Figure Description

[0012] Figure 1 This is a flowchart of the method described in this invention; Figure 2 This is a system block diagram of the system described in this invention. Detailed Implementation

[0013] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0014] This invention proposes a deep learning-based cryptographic allocation method, such as... Figure 1 As shown, the password allocation method includes: S1. Collect user information for each user and preprocess the user information to obtain valid information data corresponding to the user; S2. Using the effective information data, train and test the deep learning model to obtain a deep learning model that has completed the test, wherein the deep learning model adopts a convolutional neural network model structure. S3. Using the deep learning model that has completed the test, analyze the correlation between users to obtain the correlation results between the users; S4. Extract user groups whose correlation with each other is lower than a preset correlation threshold from the correlation results, and assign the same public and private keys to user groups whose correlation is lower than the preset correlation threshold.

[0015] The working principle of the above technical solution is as follows: S1: Collect user information for each user and preprocess the user information to obtain the corresponding valid information data. User information collection may include personal data, behavioral data, interests, etc. The preprocessing steps may involve data cleaning, feature extraction, and normalization to obtain valid user information data.

[0016] S2: Train and test the deep learning model using effective information data to obtain a tested deep learning model. Here, a convolutional neural network is used as the structure of the deep learning model. Through training and testing, an optimized and validated deep learning model is obtained.

[0017] S3: Utilize the tested deep learning model to analyze the correlation between users and obtain the degree of correlation results. By inputting user information data into the deep learning model, the model can learn the correlation patterns between users, such as interest similarity, behavioral similarity, etc., and output the correlation results between users.

[0018] S4: Extract user groups from the correlation results where the correlation between users is lower than a preset correlation threshold, and assign the same public and private keys to these user groups. Based on the correlation results, identify user groups with correlations below the preset threshold; these users may lack a clear connection. To protect user privacy and data security, assign the same public and private keys to these users to achieve key sharing and encrypted communication.

[0019] The technical effects of the above-described technical solution are as follows: This embodiment achieves the analysis of the correlation between users and the allocation of keys through steps such as user information collection, deep learning model training and testing, correlation analysis, and key distribution. This embodiment quantifies and analyzes user correlation, identifies user groups with low correlation, and provides these users with the same key to ensure the security and confidentiality of data transmission. In this way, the security and privacy protection level of user data can be improved, while providing users with an efficient and secure encrypted communication environment.

[0020] In one embodiment of the present invention, user information is collected for each user, and the user information is preprocessed to obtain valid information data corresponding to the user, including: S101. Collect user information, wherein the user information includes user identity information and interaction behavior data; S102. The user information is preprocessed to obtain valid information data corresponding to the user, wherein the preprocessing includes data cleaning, data denoising, and data normalization.

[0021] The working principle of the above technical solution is as follows: S101: Collect user information, including user identity information and interaction behavior data. User identity information may include basic information such as name, age, and gender, and interaction behavior data may include user click records, purchase history, browsing behavior, etc.

[0022] S102: Preprocess the collected user information to obtain valid user information data. Preprocessing includes data cleaning, noise reduction, and normalization. Data cleaning aims to remove noise, invalid information, and duplicate data from the data. Noise reduction can employ techniques such as filtering to reduce interference and noise in the data. Data normalization standardizes the data, making it comparable and interpretable on the same scale.

[0023] The technical effects of the above-described technical solution are as follows: This embodiment obtains valid user information data by collecting and preprocessing user information. The solution extracts useful information about user identity and interaction behavior, and preprocesses this information for subsequent data analysis, mining, and application. Preprocessing removes invalid and redundant data, reduces interference and noise, and improves data quality and usability. This approach allows for a better understanding of user needs, behaviors, and preferences, providing a foundation and support for applications such as personalized recommendations and precision marketing.

[0024] In one embodiment of the present invention, the deep learning model that has completed the test is used to analyze the correlation between users and obtain the correlation degree result between the users, including: S301. Load the completed deep learning model into the system memory, and encode and extract features from the user data according to the requirements of the deep learning model to obtain encoded feature data; wherein, the feature data includes user attribute features, user behavior features, social network features, text features, time features and image features. S302. Input the encoded feature data into a deep learning model, and analyze the feature data through the deep learning model to obtain the correlation between users.

[0025] The working principle of the above technical solution is as follows: S301: Load the tested deep learning model into the system memory, and encode and extract features from the user data according to the model requirements to obtain encoded feature data. User data can include various types of information such as user attribute features (e.g., age, gender), user behavior features (e.g., click records, purchase history), social network features (e.g., friend relationships, social interactions), text features (e.g., user comments, article content), time features (e.g., access time, active time period), and image features (e.g., user avatar, image tags).

[0026] S302: Input the encoded feature data into the deep learning model, and the model analyzes the feature data to obtain the correlation between users. The deep learning model can learn and capture the correlation patterns and interactions between users, and calculate the correlation between users by analyzing and calculating the input feature data.

[0027] The technical effects of the above solution are as follows: This embodiment uses a deep learning model that has completed testing to encode, extract features, and analyze the correlation between user data. By inputting user data into the deep learning model, the model can learn the correlation patterns and feature expressions between users, thereby deriving the correlation between users. This embodiment achieves the quantification and analysis of the correlation between users, enabling the identification of relationships and mutual influences between them. Through this technical solution, applications such as personalized recommendations, social network analysis, and precision marketing can be provided, offering users more personalized and accurate services and experiences.

[0028] In one embodiment of the present invention, user groups with a correlation score lower than a preset correlation score threshold are extracted from the correlation results, and the same public and private keys are allocated among user groups with a correlation score lower than the preset correlation score threshold, including: S401. Filter requirements based on relevance and set relevance threshold; S402. Extract the correlation degree between users output by the deep learning model, and compare the correlation degree with the correlation degree threshold; S403. Extract users whose correlation degree is lower than the correlation degree threshold and their corresponding correlation degrees; S404. For users whose correlation with each other is lower than the correlation threshold, create the same public key and private key, and assign the public key and private key to the corresponding user.

[0029] The working principle of the above technical solution is as follows: S401: Set the relevance threshold according to requirements, as a standard for filtering relevance. The relevance threshold can be set according to specific application requirements and system requirements, and is used to determine the threshold of the degree of relevance between users.

[0030] S402: Extract the correlation between users from the deep learning model output and compare the correlation with a correlation threshold. The correlation between users can be obtained through the output of the deep learning model, and these values ​​are then compared with a preset correlation threshold.

[0031] S403: Extract users whose relevance is below the relevance threshold and their corresponding relevance. Based on the comparison results, extract users whose relevance is below the threshold and their corresponding relevance. These users are considered to have low or no relevance.

[0032] S404: For users with a correlation level below the correlation threshold, create identical public and private keys and assign them to the corresponding users. To protect user privacy and data security, assigning the same public and private keys to users with low correlation levels enables key sharing and encrypted communication.

[0033] The technical effect of the above solution is as follows: This embodiment filters out users with a correlation level below a threshold based on the correlation degree, and creates the same public and private keys for these users. By setting a correlation degree threshold, this embodiment divides users into two groups: high and low correlation. A unified key is used for users with low correlation, improving system efficiency and security. In this way, the privacy and data security of users with low correlation can be protected, while reducing the complexity of key management.

[0034] This invention proposes a deep learning-based cryptographic allocation system, such as... Figure 2 As shown, the cryptographic distribution system includes: The user information collection module is used to collect user information for each user and preprocess the user information to obtain valid information data corresponding to the user. The training and testing module is used to train and test the deep learning model using the effective information data to obtain a deep learning model that has completed the test, wherein the deep learning model adopts a convolutional neural network model structure. The correlation analysis module is used to analyze the correlation between users using the deep learning model that has completed the test, and to obtain the correlation results between the users. The public and private key allocation module is used to extract user groups whose correlation between users is lower than a preset correlation threshold from the correlation results, and to allocate the same public and private keys among user groups whose correlation is lower than the preset correlation threshold.

[0035] The working principle of the above technical solution is as follows: First, user information is collected for each user through the user information collection module, and the user information is preprocessed to obtain the valid information data corresponding to the user; Then, the deep learning model is trained and tested using the effective information data through the training and testing module to obtain a deep learning model that has completed the test, wherein the deep learning model adopts a convolutional neural network model structure. Then, the correlation analysis module uses the deep learning model that has completed the test to analyze the correlation between users and obtain the correlation results between the users. Finally, the public and private key allocation module extracts user groups whose association degree is lower than a preset association degree threshold from the association degree results, and allocates the same public and private keys among user groups whose association degree is lower than the preset association degree threshold.

[0036] The technical effects of the above-described technical solution are as follows: This embodiment achieves the analysis of the correlation between users and the allocation of keys through steps such as user information collection, deep learning model training and testing, correlation analysis, and key distribution. This embodiment quantifies and analyzes user correlation, identifies user groups with low correlation, and provides these users with the same key to ensure the security and confidentiality of data transmission. In this way, the security and privacy protection level of user data can be improved, while providing users with an efficient and secure encrypted communication environment.

[0037] In one embodiment of the present invention, the user information collection module includes: The information collection module is used to collect user information, wherein the user information includes user identity information and interaction behavior data. The preprocessing module is used to preprocess the user information to obtain valid information data corresponding to the user. The preprocessing includes data cleaning, data denoising, and data normalization.

[0038] The working principle of the above technical solution is as follows: First, user information is collected through the information collection module, wherein the user information includes the user's identity information and interactive behavior data; Then, the user information is preprocessed using a preprocessing module to obtain valid information data corresponding to the user. The preprocessing includes data cleaning, data denoising, and data normalization.

[0039] The technical effects of the above-described technical solution are as follows: This embodiment obtains valid user information data by collecting and preprocessing user information. The solution extracts useful information about user identity and interaction behavior, and preprocesses this information for subsequent data analysis, mining, and application. Preprocessing removes invalid and redundant data, reduces interference and noise, and improves data quality and usability. This approach allows for a better understanding of user needs, behaviors, and preferences, providing a foundation and support for applications such as personalized recommendations and precision marketing.

[0040] In one embodiment of the present invention, the correlation analysis module includes: The feature extraction module is used to load the tested deep learning model into the system memory, and encode and extract features from the user data according to the requirements of the deep learning model to obtain encoded feature data; wherein, the feature data includes user attribute features, user behavior features, social network features, text features, time features and image features; The correlation acquisition module is used to input the encoded feature data into a deep learning model, and to obtain the correlation between users by analyzing the feature data through the deep learning model.

[0041] The working principle of the above technical solution is as follows: First, the deep learning model that has completed the test is loaded into the system memory through the feature extraction module, and the user data is encoded and features are extracted according to the requirements of the deep learning model to obtain encoded feature data; wherein, the feature data includes user attribute features, user behavior features, social network features, text features, time features and image features; Finally, the encoded feature data is input into the deep learning model through the correlation acquisition module, and the deep learning model analyzes the feature data to obtain the correlation between users.

[0042] The technical effects of the above solution are as follows: This embodiment uses a deep learning model that has completed testing to encode, extract features, and analyze the correlation between user data. By inputting user data into the deep learning model, the model can learn the correlation patterns and feature expressions between users, thereby deriving the correlation between users. This embodiment achieves the quantification and analysis of the correlation between users, enabling the identification of relationships and mutual influences between them. Through this technical solution, applications such as personalized recommendations, social network analysis, and precision marketing can be provided, offering users more personalized and accurate services and experiences.

[0043] In one embodiment of the present invention, the public and private key allocation module includes: The correlation threshold setting module is used to set the correlation threshold based on the requirements of the correlation filter. The comparison module is used to extract the correlation degree between users output by the deep learning model and compare the correlation degree with the correlation degree threshold. The correlation threshold extraction module is used to extract users whose correlation is lower than the correlation threshold and their corresponding correlation. The public and private key creation and allocation module is used to create the same public and private keys for users whose correlation with each other is lower than the correlation threshold, and to allocate the public and private keys to the corresponding users.

[0044] The working principle of the above technical solution is as follows: First, the correlation threshold setting module sets the correlation threshold according to the correlation requirements; Then, the comparison module is used to extract the correlation degree between users output by the deep learning model, and the correlation degree is compared with the correlation degree threshold. Subsequently, the correlation threshold extraction module extracts users whose correlation scores are lower than the correlation threshold and their corresponding correlation scores. Finally, the public and private key creation and allocation module creates the same public and private keys for users whose correlation with each other is lower than the correlation threshold, and allocates the public and private keys to the corresponding users.

[0045] The technical effect of the above solution is as follows: This embodiment filters out users with a correlation level below a threshold based on the correlation degree, and creates the same public and private keys for these users. By setting a correlation degree threshold, this embodiment divides users into two groups: high and low correlation. A unified key is used for users with low correlation, improving system efficiency and security. In this way, the privacy and data security of users with low correlation can be protected, while reducing the complexity of key management.

[0046] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A deep learning-based password allocation method, characterized in that, The password allocation method includes: Collect user information for each user and preprocess the user information to obtain valid information data corresponding to the user; Using the effective information data, the deep learning model is trained and tested to obtain a deep learning model that has completed the test, wherein the deep learning model adopts a convolutional neural network model structure; Using the deep learning model that has completed the test, the correlation between users is analyzed to obtain the results of the degree of correlation between the users; Extract user groups whose correlation with each other is lower than a preset correlation threshold from the correlation results, and assign the same public and private keys to user groups whose correlation is lower than the preset correlation threshold.

2. The cryptographic allocation method according to claim 1, characterized in that, Collect user information for each user, and preprocess the user information to obtain valid information data corresponding to the user, including: Collect user information, wherein the user information includes user identity information and data information on user interaction behavior; The user information is preprocessed to obtain valid information data corresponding to the user. The preprocessing includes data cleaning, data denoising, and data normalization.

3. The cryptographic allocation method according to claim 1, characterized in that, Using the deep learning model that has completed the test, the correlation between users is analyzed to obtain the correlation results between the users, including: The tested deep learning model is loaded into the system memory, and user data is encoded and features are extracted according to the requirements of the deep learning model to obtain encoded feature data; wherein, the feature data includes user attribute features, user behavior features, social network features, text features, time features and image features. The encoded feature data is input into a deep learning model, and the deep learning model analyzes the feature data to obtain the correlation between users.

4. The cryptographic allocation method according to claim 1, characterized in that, Extract user groups whose correlation with each other is lower than a preset correlation threshold from the correlation results, and assign the same public and private keys to user groups whose correlation is lower than the preset correlation threshold, including: Set a relevance threshold based on the relevance filtering requirements; Extract the correlation degree between users output by the deep learning model, and compare the correlation degree with the correlation degree threshold; Extract users whose relevance is lower than the relevance threshold and their corresponding relevance; For users whose correlation with each other is lower than the correlation threshold, create the same public and private keys, and assign the public and private keys to the corresponding users.

5. A deep learning-based cryptographic distribution system, characterized in that, The cryptographic distribution system includes: The user information collection module is used to collect user information for each user and preprocess the user information to obtain valid information data corresponding to the user. The training and testing module is used to train and test the deep learning model using the effective information data to obtain a deep learning model that has completed the test, wherein the deep learning model adopts a convolutional neural network model structure. The correlation analysis module is used to analyze the correlation between users using the deep learning model that has completed the test, and to obtain the correlation results between the users. The public and private key allocation module is used to extract user groups whose correlation between users is lower than a preset correlation threshold from the correlation results, and to allocate the same public and private keys among user groups whose correlation is lower than the preset correlation threshold.

6. The cryptographic distribution system according to claim 5, characterized in that, The user information collection module includes: The information collection module is used to collect user information, wherein the user information includes user identity information and interaction behavior data. The preprocessing module is used to preprocess the user information to obtain valid information data corresponding to the user. The preprocessing includes data cleaning, data denoising, and data normalization.

7. The cryptographic distribution system according to claim 5, characterized in that, The correlation analysis module includes: The feature extraction module is used to load the tested deep learning model into the system memory, and encode and extract features from the user data according to the requirements of the deep learning model to obtain encoded feature data; wherein, the feature data includes user attribute features, user behavior features, social network features, text features, time features and image features; The correlation acquisition module is used to input the encoded feature data into a deep learning model, and to obtain the correlation between users by analyzing the feature data through the deep learning model.

8. The cryptographic distribution system according to claim 5, characterized in that, The public and private key distribution module includes: The correlation threshold setting module is used to set the correlation threshold according to the correlation filtering requirements; The comparison module is used to extract the correlation degree between users output by the deep learning model and compare the correlation degree with the correlation degree threshold. The correlation threshold extraction module is used to extract users whose correlation is lower than the correlation threshold and their corresponding correlation. The public and private key creation and allocation module is used to create the same public and private keys for users whose correlation with each other is lower than the correlation threshold, and to allocate the public and private keys to the corresponding users.

Images