Internet of Things password management method and system
By semantic coding and correlation analysis of the passwords of IoT devices, the semantic repetition of passwords is detected and warning prompts are generated, which solves the problem of users using simple or repeated passwords, and improves the password security and overall security of IoT systems.
Patent Information
- Application Number
- CN202311262939.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-09-27
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2043-09-27
AI Technical Summary
In IoT devices and systems, users use simple and easy-to-guess passwords or reuse the same password, resulting in low password security and vulnerability to attacks, and it is difficult for traditional password management systems to effectively monitor and improve password security.
By obtaining the password data to be detected and the historical password database entered by the user, semantic encoding and association analysis are carried out, the semantic repetition of the password is detected, and warning prompts are generated to help users avoid using vulnerable passwords and improve password security.
Effectively detect and warn whether the password entered by the user is too simple or duplicate with the historical password, improve the overall password security and the security of the Internet of Things system, and reduce the risk of unauthorized access.
Smart Images

Figure CN117056902B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of password management, and more specifically, to a password management method and system for the Internet of Things. Background Art
[0002] The Internet of Things (IoT) refers to a wide range of intelligent devices, including physical devices, sensors, vehicles, and appliances, that connect and communicate via the internet. These devices can collect and exchange data and be remotely monitored and managed through cloud computing and automated control systems. With the rapid development of the IoT, an increasing number of devices and systems are being connected to the internet, making password management a critical security issue.
[0003] In the Internet of Things (IoT), devices and systems often require passwords for authentication and data protection. However, due to the large number of users and varying password complexity requirements, many users tend to use simple, easy-to-remember passwords, such as common dictionary words, birthdays, and other easily guessable information. They also reuse the same password across multiple platforms, making passwords more vulnerable to cracking and posing a risk to password security. Furthermore, in traditional password management systems, users often use the same password for extended periods of time and do not change it regularly. This means that if a password is leaked, attackers have more time and opportunity to exploit it for unauthorized access. Therefore, some IoT password management systems require users to change their passwords regularly. However, to make it easier to remember, users often use previously used passwords. This results in new passwords being less secure and more vulnerable to attacks, resulting in losses for users.
[0004] Therefore, an optimized IoT password management solution is desired. Summary of the Invention
[0005] To address the aforementioned technical issues, the present application is proposed. Embodiments of the present application provide an IoT password management method and system. This method can detect the semantic overlap between a user's entered password and historical passwords, thereby determining whether the password is too simple or whether the semantic overlap with historical passwords is too high. This helps users identify and avoid using vulnerable passwords, thereby improving overall password security and the security of the IoT system.
[0006] According to one aspect of the present application, a method for managing passwords for an Internet of Things is provided, comprising:
[0007] Obtaining the password data to be tested input by the user and a database of historical passwords used by the user;
[0008] Semantically encoding each historical password data in the historical password database to obtain a plurality of historical password semantic encoding feature vectors;
[0009] Performing semantic association analysis on the plurality of historical password semantic encoding feature vectors to obtain historical password semantic association topological features;
[0010] Performing semantic encoding on the password data to be detected input by the user to obtain a semantic understanding feature of the input password;
[0011] Mapping the input password semantic understanding feature to the high-dimensional space of the historical password semantic association topological feature to obtain a password semantic association feature; and
[0012] Based on the password semantic association feature, it is determined whether the semantic repetition degree of the password data to be detected exceeds a predetermined threshold and a warning prompt is generated.
[0013] According to another aspect of the present application, an Internet of Things password management system is provided, comprising:
[0014] A data acquisition module is used to obtain the password data to be tested input by the user and a historical password database used by the user;
[0015] A historical data semantic encoding module, configured to perform semantic encoding on each historical password data in the historical password database to obtain a plurality of historical password semantic encoding feature vectors;
[0016] A semantic association analysis module, configured to perform semantic association analysis on the plurality of historical password semantic encoding feature vectors to obtain historical password semantic association topological features;
[0017] An input data semantic encoding module, configured to semantically encode the password data to be detected input by the user to obtain a semantic understanding feature of the input password;
[0018] A mapping module, configured to map the input password semantic understanding feature to the high-dimensional space of the historical password semantic association topology feature to obtain a password semantic association feature; and
[0019] The early warning prompt module is used to determine whether the semantic repetition of the password data to be detected exceeds a predetermined threshold based on the password semantic association feature and generate an early warning prompt.
[0020] Compared with the prior art, the IoT password management method and system provided by this application first obtains the password data to be tested input by the user, as well as a database of historical passwords used by the user. Then, each historical password data in the historical password database is semantically encoded to obtain multiple historical password semantic encoding feature vectors. Then, semantic association analysis is performed on the multiple historical password semantic encoding feature vectors to obtain historical password semantic association topological features. Then, the password data to be tested input by the user is semantically encoded to obtain input password semantic understanding features. Then, the input password semantic understanding features are mapped to the high-dimensional space of the historical password semantic association topological features to obtain password semantic association features. Finally, based on the password semantic association features, it is determined whether the semantic repetition of the password data to be tested exceeds a predetermined threshold and a warning prompt is generated. In this way, password security can be improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for the description of the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present application.
[0022] Figure 1 Flowchart of the Internet of Things password management method according to an embodiment of the present application.
[0023] Figure 2 Schematic diagram of the architecture of the Internet of Things password management method according to an embodiment of the present application.
[0024] Figure 3 This is a flowchart of sub-step S130 of the Internet of Things password management method according to an embodiment of the present application.
[0025] Figure 4 This is a flowchart of sub-step S160 of the Internet of Things password management method according to an embodiment of the present application.
[0026] Figure 5 2 is a block diagram of an Internet of Things password management system according to an embodiment of the present application.
[0027] Figure 6 This is an application scenario diagram of the Internet of Things password management method according to an embodiment of the present application. DETAILED DESCRIPTION
[0028] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without making creative efforts also fall within the scope of protection of this application.
[0029] As used in this application and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not intended to refer to the singular but may include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.
[0030] Although the present application makes various references to certain modules in the system according to embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are illustrative only, and different aspects of the system and method can use different modules.
[0031] Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the various steps may be processed in reverse order or simultaneously, as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.
[0032] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.
[0033] In response to the above technical problems, the technical concept of this application is to collect the password data entered by the user, and introduce data processing and analysis algorithms in the back end to perform semantic correlation analysis on the password data entered by the user and the historical password data, so as to detect the semantic repetition between the password entered by the user and the historical passwords, and thus determine whether the password entered by the user is too simple and whether the semantic repetition with the historical passwords is too high, so as to help users identify and avoid using vulnerable passwords, thereby improving the overall password security and the security of the Internet of Things system.
[0034] Figure 1 Flowchart of the Internet of Things password management method according to an embodiment of the present application. Figure 2 FIG is a schematic diagram of the architecture of the Internet of Things password management method according to an embodiment of the present application. Figure 1 and Figure 2As shown, the Internet of Things password management method according to the embodiment of the present application includes the following steps: S110, obtaining the password data to be detected input by the user, and a historical password database used by the user; S120, semantically encoding each historical password data in the historical password database to obtain multiple historical password semantic encoding feature vectors; S130, performing semantic association analysis on the multiple historical password semantic encoding feature vectors to obtain historical password semantic association topological features; S140, semantically encoding the password data to be detected input by the user to obtain input password semantic understanding features; S150, mapping the input password semantic understanding features to the high-dimensional space of the historical password semantic association topological features to obtain password semantic association features; and, S160, determining whether the semantic repetition of the password data to be detected exceeds a predetermined threshold based on the password semantic association features and generating an early warning prompt.
[0035] Specifically, in the technical solution of the present application, first, the password data to be tested, entered by the user, and a database of historical passwords used by the user are obtained. Next, considering that each historical password data in the historical password database has password semantic features, in order to perform semantic association analysis on each historical password data to compare the hidden semantic similarities and repetitions between historical passwords, thereby detecting whether the user has used similar or repeated passwords and reminding the user to avoid using vulnerable passwords, the technical solution of the present application further requires semantic encoding of each historical password data in the historical password database to obtain multiple historical password semantic encoding feature vectors. In particular, in a specific example of the present application, each historical password data in the historical password database can be semantically encoded using a contextual semantic encoder including an embedding layer, thereby extracting the contextual semantic feature information of each historical password data.
[0036] Then, in order to measure the degree of semantic association between the various historical password data and thereby more accurately detect and evaluate the semantic repetition of the password data to be detected, the technical solution of the present application further calculates the cross-entropy between any two historical password semantic encoding feature vectors among the multiple historical password semantic encoding feature vectors to obtain a historical password semantic association topology matrix. It should be understood that the historical password semantic association topology matrix can represent the degree of semantic association between the various historical password data, wherein each element in the matrix represents the degree of association between two historical passwords, and can be used to determine whether there is similarity or reuse of historical passwords. Specifically, a higher cross-entropy value indicates greater differences between passwords, while a lower cross-entropy value indicates greater similarity between passwords.
[0037] Then, the historical password semantic association topology matrix is subjected to feature mining in a topological feature extractor based on a convolutional neural network model to extract the semantic topological association feature information between the various historical password data, thereby obtaining the historical password semantic association topological feature matrix.
[0038] Furthermore, using the individual historical password semantic encoding feature vectors as feature representations of nodes, and the historical password semantic association topology feature matrix as feature representations of edges between nodes, a historical password global semantic encoding feature matrix obtained by two-dimensionally arranging the multiple historical password semantic encoding feature vectors and the historical password semantic association topology feature matrix are passed through a graph neural network model to obtain a semantic association topology historical password semantic encoding feature matrix. Specifically, the graph neural network model uses learnable neural network parameters to perform graph structure data encoding on the historical password global semantic encoding feature matrix and the historical password semantic association topology feature matrix to obtain the semantic association topology historical password semantic encoding feature matrix containing irregular semantic topology association features between the individual historical password data and semantic feature information of the individual historical password data.
[0039] Furthermore, semantic understanding of the password data to be tested, which is input by the user, is also required to facilitate subsequent exploration of its semantic association with the historical password data. Therefore, in the technical solution of this application, the password data to be tested, which is input by the user, is further encoded using a password semantic understander comprising a word embedding model and a BiLSTM model to extract contextual semantic association feature information of the password data to be tested, thereby obtaining an input password semantic understanding feature vector.
[0040] Next, the semantic topology historical password semantic encoding feature matrix is matrix-multiplied with the input password semantic understanding feature vector to map the semantic understanding features of the user-entered password data to be tested into the high-dimensional space of global semantic association features of each historical password data, thereby obtaining a password semantic association vector. This password semantic association vector reflects the degree of semantic association between the input password and historical passwords, facilitating the detection of semantic overlap between the user-entered password and historical passwords, thereby determining whether the user-entered password is too simple or whether the semantic overlap with historical passwords is too high.
[0041] Accordingly, if Figure 3As shown, performing semantic association analysis on the multiple historical password semantic encoding feature vectors to obtain historical password semantic association topological features includes: S131, calculating the cross entropy between any two historical password semantic encoding feature vectors to obtain a historical password semantic association topological matrix; S132, passing the historical password semantic association topological matrix through a topological feature extractor based on a convolutional neural network model to obtain a historical password semantic association topological feature matrix; and S133, passing the multiple historical password semantic encoding feature vectors and the historical password semantic association topological feature matrix through a graph neural network model to obtain a semantic association topological historical password semantic encoding feature matrix as the historical password semantic association topological feature. It should be understood that in step S131, cross entropy is an indicator used to measure the difference between two probability distributions. Here, the historical password semantic encoding feature vectors are regarded as probability distributions. By calculating the cross entropy between them, the degree of association between them can be obtained. The historical password semantic association topological matrix records the degree of association between each pair of historical password semantic encoding feature vectors. In step S132, a topological feature extractor based on a convolutional neural network model converts the historical password semantic correlation topology matrix into a historical password semantic correlation topology feature matrix. This step uses a convolutional neural network model to extract topological features, converting the correlation information in the historical password semantic correlation topology matrix into more meaningful feature representations. This reduces data dimensionality and extracts more valuable features. In step S133, a graph neural network model is used to combine the historical password semantic encoding feature vectors and the historical password semantic correlation topology feature matrix to obtain a semantic correlation topology historical password semantic encoding feature matrix. Graph neural networks are neural network models suitable for graph-structured data, effectively processing the relationships and connections between nodes. Here, the historical password semantic encoding feature vectors and the historical password semantic correlation topology feature matrix are used as features of the nodes and edges of the graph. The graph neural network model learns the correlations between them to obtain more comprehensive and integrated historical password semantic correlation topology features. Taken together, these three steps aim to extract more meaningful and comprehensive semantic correlation topology features from the historical password semantic encoding feature vectors for use in password-related tasks or analysis. These features can help understand the degree of correlation between passwords, extract useful information, and support applications in password security and cryptography research.
[0042] It's worth mentioning that the Convolutional Neural Network (CNN) is a deep learning model particularly well-suited for processing data with a grid structure. The key features of a CNN are its hierarchical structure and weight sharing mechanism. It consists of multiple convolutional layers, pooling layers, and fully connected layers. Convolutional layers extract features from input data through convolution operations, pooling layers reduce the data's dimensionality through downsampling, and fully connected layers map the extracted features to the final output. The advantage of CNNs lies in their ability to automatically learn features from the data and their invariance to transformations such as translation, scaling, and rotation. Through their local receptive field and weight sharing mechanisms, CNNs can effectively process large amounts of data. Furthermore, they employ a certain degree of parameter sharing during training, reducing the number of parameters required and improving model training efficiency. In summary, CNNs are powerful deep learning models that can automatically learn features from data and excel at processing data with a grid structure.
[0043] A graph neural network (GNN) is a deep learning model specifically designed for processing graph-structured data. Unlike traditional neural network models, which are designed to process data in vector or matrix form, GNNs are able to effectively capture the relationships and connections between nodes in a graph. A key feature of GNNs is their ability to learn representations of nodes and edges through message passing and aggregation. They iteratively update node representations, incorporating information about a node's neighbors into the representation, thereby capturing global graph structure. This enables GNNs to process graphs of varying sizes and structures, while also providing reasonable reasoning and generalization capabilities. GNNs can be used for tasks such as node classification, link prediction, graph generation, and graph clustering. By learning the structure and relationships of a graph, GNNs can extract both global and local features, helping to understand and analyze complex graph data and providing effective solutions to graph-related problems.
[0044] Accordingly, in step S140, semantic encoding is performed on the password data to be tested input by the user to obtain input password semantic understanding features, including: passing the password data to be tested input by the user through a password semantic understander comprising a word embedding model and a BiLSTM model to obtain an input password semantic understanding feature vector as the input password semantic understanding feature. It is worth noting that word embedding models and BiLSTM models are two commonly used models in password semantic understanders, used to convert input password data into semantic understanding feature vectors. A word embedding model is a model used to map discrete password data (such as characters or words in a password) into a continuous vector space. It can capture semantic relationships and similarities between password data and represent them as dense vectors. Common word embedding models include Word2Vec, GloVe, and FastText. By using a word embedding model, password data can be converted into a vector representation with semantic information, thereby better expressing the meaning and characteristics of the password. The BiLSTM model is a bidirectional long short-term memory network (BLSTM) used to process sequence data. It consists of two LSTM networks, one forward and one backward, processing the sequence from both directions, concatenating or merging their outputs. The BiLSTM model captures contextual information in sequence data and considers the order and dependencies of characters in the password when processing password data. By using the BiLSTM model, password data can be converted into a feature representation with sequence information, better capturing the semantics and structure of the password. In password semantic understanders, word embedding models and BiLSTM models are often used together to obtain richer and more accurate features for password semantic understanding. First, the word embedding model maps the password data into a dense vector representation, capturing the password's semantic information. Then, the BiLSTM model further processes the sequence information based on these word embeddings, capturing the context and dependencies in the password. Finally, the output of the BiLSTM model is used as a password semantic understanding feature vector, which can be used in subsequent password analysis, classification, or other tasks. In other words, word embedding models and BiLSTM models play a key role in password semantic understanders, transforming input password data into feature vectors with semantic and sequence information, thereby better understanding and representing the meaning and structure of the password.
[0045] Accordingly, in step S150, the input password semantic understanding features are mapped to the high-dimensional space of the historical password semantic association topology features to obtain password semantic association features, including matrix multiplication of the semantic association topology historical password semantic encoding feature matrix with the input password semantic understanding feature vector to obtain a password semantic association vector as the password semantic association feature. It should be understood that the purpose of matrix multiplication of the semantic association topology historical password semantic encoding feature matrix with the input password semantic understanding feature vector is to obtain a password semantic association vector, which is used as the password semantic association feature. This step combines the semantic association information in the historical password data with the semantic understanding features of the current input password, thereby capturing the semantic association and similarity between passwords. Through matrix multiplication, the semantic association information in the historical password data and the semantic features of the current password are weighted and fused to obtain a password semantic association vector. The password semantic association vector can provide a measure of the degree of semantic correlation between passwords. It can help determine the similarity or degree of correlation between the input password and historical passwords, thereby supporting password-related tasks such as password recommendation, password cracking, and password strength assessment. By using password semantic association vectors as password semantic association features, the representation ability of password data can be improved.
[0046] The password semantic association vector is then passed through a classifier to obtain a classification result. The classification result indicates whether the semantic repetition of the password data to be tested exceeds a predetermined threshold. Specifically, the classification process utilizes the mapping and association feature information between the semantic understanding features of the input password data to be tested and the semantic association features of each historical password data. This is used to detect the semantic repetition between the password input by the user and the historical passwords, determine whether the semantic repetition of the password data to be tested exceeds a predetermined threshold, and generate a warning based on the classification result.
[0047] Accordingly, if Figure 4 As shown, based on the password semantic association features, determining whether the semantic repetition of the password data to be detected exceeds a predetermined threshold and generating a warning prompt, including: S161, performing Hilbert orthogonal space domain representation decoupling on the password semantic association vector to obtain an optimized password semantic association vector; S162, passing the optimized password semantic association vector through a classifier to obtain a classification result, and the classification result is used to indicate whether the semantic repetition of the password data to be detected exceeds a predetermined threshold; and, S163, generating a warning prompt based on the classification result.
[0048] In particular, in the technical solution of the present application, when the multiple historical password semantic encoding feature vectors and the historical password semantic association topology feature matrix are passed through a graph neural network model, the semantic association topology historical password semantic encoding feature matrix can express the topological association representation of the encoding semantics of the historical password data under the semantic distribution similarity topology. Therefore, after matrix multiplication of the semantic association topology historical password semantic encoding feature matrix with the input password semantic understanding feature vector, the password semantic association vector not only includes the short-range and long-range bidirectional context encoding semantic features of the password data to be tested input by the user, but also includes the semantic distribution topological association representation in the historical full semantic space, thereby giving the password semantic association vector a diversified feature expression. In this way, when the password semantic association vector is passed through a classifier, it will affect the generalization effect of the password semantic association vector as a whole within the classification domain, that is, affect the accuracy of the classification result.
[0049] Based on this, when the applicant of this application classifies the password semantic association vector, it is preferred to record the password semantic association vector, for example, as Perform Hilbert orthogonal space domain representation decoupling.
[0050] Accordingly, in a specific example, performing Hilbert orthogonal space domain representation decoupling on the cryptographic semantic association vector to obtain an optimized cryptographic semantic association vector includes: performing Hilbert orthogonal space domain representation decoupling on the cryptographic semantic association vector using the following optimization formula to obtain the optimized cryptographic semantic association vector; wherein the optimization formula is:
[0051] ,
[0052] in, is the crypto-semantic association vector, is the global feature mean of the cryptographic semantic association vector, is the two-norm of the cryptographic semantic association vector, is the length of the crypto-semantic association vector, and is a unit vector, represents vector subtraction, represents the covariance matrix, is the optimized cryptographic semantic association vector.
[0053] Here, the Hilbert orthogonal space domain representation is decoupled for semantic association vectors by emphasizing the cryptographic The essential domain-specific information within the diverse feature expressions of the cryptographic semantics is extracted by using the Hilbert space metric based on the vector self-space metric and the vector self-inner product representation. The orthogonal space domain decoupling of domain-invariant representation is performed within the overall domain representation of to improve the password semantic association vector The domain-adaptive generalization performance within the classification domain improves the accuracy of the classification results obtained by the classifier using the password semantic association vector. This effectively detects the semantic overlap between the user-entered password and historical passwords, thereby determining whether the user-entered password is too simple or whether the semantic overlap with historical passwords is too high. This helps users identify and avoid using vulnerable passwords, improving overall password security and the security of the IoT system.
[0054] Furthermore, in step S162, the optimized password semantic association vector is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the semantic repetition degree of the password data to be detected exceeds a predetermined threshold, including: using the fully connected layer of the classifier to fully connect encode the optimized password semantic association vector to obtain an encoded classification feature vector; and inputting the encoded classification feature vector into the Softmax classification function of the classifier to obtain the classification result.
[0055] That is, in the technical solution disclosed herein, the labels of the classifier include the semantic repetition of the password data to be detected exceeding a predetermined threshold (first label), and the semantic repetition of the password data to be detected not exceeding a predetermined threshold (second label), wherein the classifier determines to which classification label the optimized password semantic association vector belongs by a soft maximum function. It is worth noting that the first label p1 and the second label p2 here do not contain artificially set concepts. In fact, during the training process, the computer model does not have the concept of "whether the semantic repetition of the password data to be detected exceeds the predetermined threshold". It only has two classification labels and the probability of the output feature under these two classification labels, that is, the sum of p1 and p2 is one. Therefore, the classification result of whether the semantic repetition of the password data to be detected exceeds the predetermined threshold is actually converted from the classification label to a binary classification class probability distribution that conforms to natural laws. In essence, what is used is the physical meaning of the natural probability distribution of the label, rather than the linguistic text meaning of "whether the semantic repetition of the password data to be detected exceeds the predetermined threshold."
[0056] It's easy to understand that a classifier uses given categories and known training data to learn classification rules and classifiers, and then classify (or predict) unknown data. Logistic regression and SVM are commonly used to solve binary classification problems. For multi-class classification, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-class classification. However, this approach is error-prone and inefficient. A commonly used multi-classification method is the Softmax classification function.
[0057] In summary, the Internet of Things password management method based on the embodiment of the present application is explained, which can detect the semantic repetition between the password entered by the user and the historical password, so as to determine whether the password entered by the user is too simple and whether the semantic repetition with the historical password is too high, so as to help users identify and avoid using vulnerable passwords, improve the overall password security and the security of the Internet of Things system.
[0058] Figure 5 FIG is a block diagram of an Internet of Things password management system 100 according to an embodiment of the present application. Figure 5 As shown, the Internet of Things password management system 100 according to the embodiment of the present application includes: a data acquisition module 110, which is used to obtain the password data to be detected input by the user, and a historical password database used by the user; a historical data semantic encoding module 120, which is used to semantically encode each historical password data in the historical password database to obtain multiple historical password semantic encoding feature vectors; a semantic association analysis module 130, which is used to perform semantic association analysis on the multiple historical password semantic encoding feature vectors to obtain historical password semantic association topological features; an input data semantic encoding module 140, which is used to semantically encode the password data to be detected input by the user to obtain input password semantic understanding features; a mapping module 150, which is used to map the input password semantic understanding features to the high-dimensional space of the historical password semantic association topological features to obtain password semantic association features; and an early warning prompt module 160, which is used to determine whether the semantic repetition of the password data to be detected exceeds a predetermined threshold based on the password semantic association features and generate an early warning prompt.
[0059] In one example, in the above-mentioned Internet of Things password management system 100, the semantic association analysis module 130 includes: a cross-entropy calculation unit, used to calculate the cross entropy between any two historical password semantic encoding feature vectors among the multiple historical password semantic encoding feature vectors to obtain a historical password semantic association topology matrix; a topological feature extraction unit, used to pass the historical password semantic association topology matrix through a topological feature extractor based on a convolutional neural network model to obtain a historical password semantic association topology feature matrix; and, a graph neural network encoding unit, used to pass the multiple historical password semantic encoding feature vectors and the historical password semantic association topology feature matrix through a graph neural network model to obtain a semantic association topology historical password semantic encoding feature matrix as the historical password semantic association topology feature.
[0060] In one example, in the above-mentioned Internet of Things password management system 100, the input data semantic encoding module 140 is used to: pass the password data to be detected input by the user through a password semantic understander including a word embedding model and a BiLSTM model to obtain an input password semantic understanding feature vector as the input password semantic understanding feature.
[0061] In one example, in the above-mentioned Internet of Things password management system 100, the mapping module 150 is used to: matrix multiply the semantic association topology historical password semantic encoding feature matrix with the input password semantic understanding feature vector to obtain a password semantic association vector as the password semantic association feature.
[0062] Here, those skilled in the art will appreciate that the specific functions and operations of each module in the above-mentioned Internet of Things password management system 100 have been described in detail in the previous reference. Figures 1 to 4 The description of the IoT password management method has been introduced in detail, and therefore, its repeated description will be omitted.
[0063] As described above, the IoT password management system 100 according to an embodiment of the present application can be implemented in various wireless terminals, such as a server equipped with an IoT password management algorithm. In one example, the IoT password management system 100 according to an embodiment of the present application can be integrated into the wireless terminal as a software module and / or hardware module. For example, the IoT password management system 100 can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the IoT password management system 100 can also be one of the many hardware modules of the wireless terminal.
[0064] Alternatively, in another example, the IoT password management system 100 and the wireless terminal may also be separate devices, and the IoT password management system 100 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.
[0065] Figure 6 FIG is an application scenario diagram of the Internet of Things password management method according to an embodiment of the present application. Figure 6 As shown, in this application scenario, first, the password data to be detected input by the user is obtained (for example, Figure 6 D1 shown in FIG), and a historical password database used by the user (for example, Figure 6 Then, the password data to be detected input by the user and the historical password database are input into a server deployed with an Internet of Things password management algorithm (for example, Figure 6 S as shown in ), wherein the server is capable of using the Internet of Things password management algorithm to process the password data to be detected input by the user and the historical password database to obtain a classification result indicating whether the semantic repetition of the password data to be detected exceeds a predetermined threshold.
[0066] In addition, it will be understood by those skilled in the art that various aspects of the present application can be illustrated and described by a number of patentable categories or situations, including any new and useful process, machine, product or combination of substances, or any new and useful improvements thereto. Accordingly, various aspects of the present application can be performed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software may all be referred to as "data blocks", "modules", "engines", "units", "components" or "systems". In addition, various aspects of the present application may be represented as a computer product located in one or more computer-readable media, which includes computer-readable program code.
[0067] Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology and should not be interpreted in an idealized or highly formal sense, unless expressly defined as such herein.
[0068] The above is an illustration of the present invention and should not be considered as limiting thereof. Although several exemplary embodiments of the present invention have been described, it will be readily understood by those skilled in the art that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present invention. Therefore, all such modifications are intended to be included within the scope of the present invention as defined by the claims. It should be understood that the above is an illustration of the present invention and should not be considered as being limited to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present invention is defined by the claims and their equivalents.
Claims
1. A password management method for the Internet of Things, characterized in that: include: Obtaining the password data to be detected input by the user and a historical password database used by the user; Semantically encoding each historical password data in the historical password database to obtain a plurality of historical password semantic encoding feature vectors; Performing semantic association analysis on the plurality of historical password semantic encoding feature vectors to obtain historical password semantic association topological features; Semantically encoding the password data to be detected input by the user to obtain a semantic understanding feature of the input password; Mapping the input password semantic understanding feature to the high-dimensional space of the historical password semantic association topological feature to obtain a password semantic association feature; as well as Based on the password semantic association feature, determine whether the semantic repetition of the password data to be detected exceeds a predetermined threshold and generate a warning prompt; Performing semantic association analysis on the plurality of historical password semantic encoding feature vectors to obtain historical password semantic association topological features includes: Calculating the cross entropy between any two historical password semantic encoding feature vectors among the plurality of historical password semantic encoding feature vectors to obtain a historical password semantic association topology matrix; Passing the historical password semantic association topology matrix through a topology feature extractor based on a convolutional neural network model to obtain a historical password semantic association topology feature matrix; and The multiple historical password semantic encoding feature vectors and the historical password semantic association topology feature matrix are passed through a graph neural network model to obtain a semantic association topology historical password semantic encoding feature matrix as the historical password semantic association topology feature.
2. The Internet of Things password management method according to claim 1, characterized in that: Semantically encoding the password data to be detected input by the user to obtain a semantic understanding feature of the input password includes: The password data to be detected input by the user is passed through a password semantic understander including a word embedding model and a BiLSTM model to obtain an input password semantic understanding feature vector as the input password semantic understanding feature.
3. The Internet of Things password management method according to claim 2, characterized in that: Mapping the input password semantic understanding feature to the high-dimensional space of the historical password semantic association topological feature to obtain the password semantic association feature, including: The semantic association topology historical password semantic encoding feature matrix is matrix-multiplied with the input password semantic understanding feature vector to obtain a password semantic association vector as the password semantic association feature.
4. The Internet of Things password management method according to claim 3, characterized in that: Based on the password semantic association feature, determining whether the semantic repetition of the password data to be detected exceeds a predetermined threshold and generating a warning prompt includes: Performing Hilbert orthogonal space domain representation decoupling on the cryptographic semantic association vector to obtain an optimized cryptographic semantic association vector; Passing the optimized password semantic association vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the semantic repetition degree of the password data to be detected exceeds a predetermined threshold; and Based on the classification result, a warning prompt is generated.
5. The Internet of Things password management method according to claim 4, characterized in that: Performing Hilbert orthogonal space domain representation decoupling on the cryptographic semantic association vector to obtain an optimized cryptographic semantic association vector includes: Decoupling the cryptographic semantic association vector by Hilbert orthogonal space domain representation using the following optimization formula to obtain the optimized cryptographic semantic association vector; Wherein, the optimization formula is: ; in, is the crypto-semantic association vector, is the global feature mean of the password semantic association vector, is the bi-norm of the cryptographic semantic association vector, is the length of the crypto-semantic association vector, and is a unit vector, represents vector subtraction, represents the covariance matrix, is the optimized crypto-semantic association vector.
6. An Internet of Things password management system, characterized in that: include: A data acquisition module, used to acquire the password data to be detected input by the user, and a historical password database used by the user; A historical data semantic coding module, used for semantically coding each historical password data in the historical password database to obtain a plurality of historical password semantic coding feature vectors; A semantic association analysis module, used for performing semantic association analysis on the plurality of historical password semantic encoding feature vectors to obtain historical password semantic association topological features; An input data semantic encoding module, used for semantically encoding the password data to be detected input by the user to obtain input password semantic understanding features; A mapping module, used for mapping the input password semantic understanding feature to the high-dimensional space of the historical password semantic association topological feature to obtain the password semantic association feature; as well as An early warning prompt module, used to determine whether the semantic repetition degree of the password data to be detected exceeds a predetermined threshold and generate an early warning prompt based on the password semantic association feature; The semantic association analysis module comprises: A cross entropy calculation unit, used to calculate the cross entropy between any two historical cryptographic semantic encoding feature vectors among the plurality of historical cryptographic semantic encoding feature vectors to obtain a historical cryptographic semantic association topology matrix; A topological feature extraction unit, used for obtaining a historical password semantic association topological feature matrix by passing the historical password semantic association topological matrix through a topological feature extractor based on a convolutional neural network model; and A graph neural network encoding unit is used to pass the multiple historical password semantic encoding feature vectors and the historical password semantic association topology feature matrix through a graph neural network model to obtain a semantic association topology historical password semantic encoding feature matrix as the historical password semantic association topology feature.
7. The Internet of Things password management system according to claim 6, characterized in that: The input data semantic encoding module is used to: The password data to be detected input by the user is passed through a password semantic understander including a word embedding model and a BiLSTM model to obtain an input password semantic understanding feature vector as the input password semantic understanding feature.
8. The Internet of Things password management system according to claim 7, characterized in that: The mapping module is used to: The semantic association topology historical password semantic encoding feature matrix is matrix-multiplied with the input password semantic understanding feature vector to obtain a password semantic association vector as the password semantic association feature.
