Dynamic password generation system based on user behaviors and biological characteristics

Abstract

The invention belongs to the technical field of computer application, and relates to a dynamic password generation system based on user behaviors and biological characteristics. Comprising a data recording module used for collecting and recording user information, an information updating module used for updating the user information in a local database, the local database comprising a user basic information area, a user behavior characteristic area and a user biological characteristic area, and a first verification module used for verifying the user basic information, the dynamic generation module is used for generating a dynamic password by utilizing the behavior characteristics of the user; the second verification module is used for verifying the biological characteristics of the user; according to the method, the security of sensitive data of a user is improved, threats caused by AI generation are greatly reduced, an API interface is provided, the use convenience of the user is improved, the multi-application verification complexity is reduced, and the user experience is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer security technology, and particularly to a dynamic password generation system based on user behavior and biometric characteristics. Background Art

[0002] Traditional static passwords are vulnerable to threats such as phishing and dictionary attacks, so dynamic password systems have been introduced. User identities are verified through dynamic passwords generated within a short period of time. These dynamic passwords are usually only valid for a short time, providing a higher level of security protection because even if the password is intercepted, it is difficult to be used again outside the validity period. Dynamic passwords provide a higher level of security protection for individuals, enterprises, financial institutions, etc., reducing the risk of unauthorized access, curbing various network attack behaviors, and becoming one of the important protection means in the current digital security field. Most of the existing dynamic passwords are generated based on time synchronization algorithms or event synchronization algorithms. Users input the current dynamic password during the login process and compare it with the server side to achieve two-factor authentication, such as SMS verification codes. Another effective verification method is biometric verification, such as face recognition and palmprint recognition.

[0003] With the development of technology, each application requires users to perform biometric identification or dynamic SMS, resulting in the abuse of biometric characteristics, the leakage of personal information, and SMS harassment. Moreover, the rapid development of AI makes biometric information such as faces and palmprints reproducible, causing great inconvenience and threats to users. Summary of the Invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a dynamic password generation system based on user behavior and biometric characteristics.

[0005] To achieve the purpose of the present invention, the following technical solutions will be adopted for implementation.

[0006] A dynamic password generation system based on user behavior and biometric characteristics, the dynamic password generation system includes a data recording module, an information updating module, and a local database. The data recording module is used to collect and record user information. The user information includes user basic information and feature data. The feature data includes user behavior characteristics and user biometric characteristics. At the same time, a privacy protection strategy for user information is designed and the user information is stored in the local database. The information updating module is used to update the user information in the local database. The local database includes an access permission area, and the access permission area is divided into a user basic information area, a user behavior characteristic area, and a user biometric characteristic area. It is characterized in that the dynamic password generation system further includes a first verification module, a second verification module, and a dynamic generation module, wherein:

[0007] The first verification module verifies the basic information of the accessing user by using the user's basic information stored in the user basic information area, and gives a verification result: matching or not matching, where:

[0008] If not matching, access is refused;

[0009] If matching, access continues;

[0010] The dynamic generation module obtains the usage permission of the user behavior feature area according to the verification result of "matching" of the first verification module, obtains the user behavior features, and performs time series analysis using the user behavior features and historical user behavior features to generate a dynamic password;

[0011] The second verification module obtains the usage permission of the biometric feature area according to the generated dynamic password, obtains the user biometric features, and verifies the biometric features of the accessing user to achieve secure access;

[0012] The information update module updates the local database according to the dynamic password generated by the dynamic generation module and the verification result of the second verification module to obtain the latest user behavior features and user biometric features.

[0013] As a preferred solution of the present invention, the user behavior features include the login IP, click frequency, and usage period.

[0014] As a preferred solution of the present invention, the user biometric features include face, palmprint, and voice.

[0015] As a preferred solution of the present invention, the specific operation of the data privacy protection strategy is as follows:

[0016] After the user data is collected and saved in the local database of device A, an access permission area is divided, and the access permission area includes a user basic information area, a user behavior feature area, and a user biometric feature area;

[0017] The data in the user basic information area is used by the first verification module to verify the device basic information;

[0018] According to the verification result of the first verification module, obtain the usage permission of the user behavior feature area for the dynamic generation module;

[0019] According to the dynamic password generated by the dynamic generation module, obtain the usage permission of the user biometric feature area for the second verification module.

[0020] As a preferred solution of the present invention, the modeling process of the time series analysis includes the following steps:

[0021] S51. Construct a time series analysis model:

[0022] S511. Describe the linear relationship between the current observation and past observations:

[0023]

[0024] where X t is the observation at the current time t, φ i is the autoregressive coefficient, ε t is the white noise error, p represents the order of the autoregressive term, indicating that the observation at the current time is related to the observations at the previous p time points, and c is a constant term.

[0025] S512. Describe the linear relationship between the current observation and the random error term:

[0026]

[0027] where X t is the observation at the current time t, θ i is the moving average coefficient, ε t is the white noise error at the current time t, ε t-i is the white noise error at time (t - i), q represents the order of the moving average term, indicating that the observation at the current time is related to the error terms at the previous q time points, and c is a constant term.

[0028] S513. The combined model is expressed as follows:

[0029] Z(p,q) = X t + Y t

[0030] S52. Use the historical behavior feature data to estimate and fit the parameters of the Z(p,q) model;

[0031] S53. Use the fitted Z(p,q) model to predict future values, specifically, generate the predicted value at the current time point using the current data point through the model described in S513.

[0032] As a preferred solution of the present invention, the second verification module verifies the biometric features of the accessing user, including facial similarity recognition, palmprint similarity recognition, and voice similarity recognition. The similarity recognition model is designed as follows:

[0033] The input data are x1, x2, the input feature mapping is expressed as y1 = g(x1), y2 = g(x2), and the similarity metric is expressed as: d = |y1 - y2| 2 , and the loss function is designed as:

[0034]

[0035] Among them, g(.) represents the CNN module with shared weights, N represents the number of training samples, and Y represents the label. When Y = 1, it indicates samples of the same category, and the goal is to make their distances as close as possible; when Y = 0, it indicates samples of different categories, and the goal is to make their distances greater than the margin. The margin is a preset threshold, representing the minimum distance that should exist between samples of different categories.

[0036] A storage structure. The dynamic password generation system based on user behavior and biometric features described in claims 1-6 is implemented through the cloud and encapsulated using DOCKER technology, and can be used as a plugin or interface.

[0037] Beneficial effects

[0038] The present invention has the following advantages: (1) Data is stored locally and divided into permission areas, and data access can only be carried out through verification, improving the security of user sensitive data; (2) The dynamic password generation architecture combines basic data, user behavior feature data, and user biometric data for dynamic identity verification, greatly reducing the threat brought by AI generation; (3) An API interface is provided, improving the convenience of user use, reducing the complexity of multi-application verification, and improving the user experience. Brief description of the drawings

[0039] Figure 1 It is a schematic structural diagram of the system described in the present invention;

[0040] Figure 2 It is a schematic structural diagram of the access permission area described in the present invention;

[0041] Figure 3 It is a schematic architecture diagram of the dynamic password generation system described in the present invention; Detailed implementation manners

[0042] The following specific embodiments illustrate the implementation manners of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0043] As an embodiment of the present invention, as Figure 1As shown in the figure, a dynamic password generation system based on user behavior and biometrics first determines whether the accessing user is a new user. If it is a new user, the data recording module needs to collect the user's information about commonly used devices and website usage, as well as biometric information such as the user's face, palm print, and voice. If it is an old user, the user's identity is verified based on basic information such as the account password. After passing the verification by the first module, the user behavior characteristic data in the local database is called, and a dynamic password is generated through the dynamic generation module. Then, the corresponding biometric information in the local database is obtained based on the dynamic password for verification by the second verification module. After passing the verification by the second verification module, secure access is successfully achieved.

[0044] As an embodiment of the present invention, as Figure 2 shown, the data privacy protection strategy stores the data in the local database. The specific operations are as follows: After the user data is collected, it is saved in the local database of the device, and the access permission area is divided. The access permission area includes the user basic information area, the user behavior characteristic area, and the user biometric area. Among them, the data in the user basic information area is used by the first verification module to verify the device basic information. According to the result of the first verification module, the data usage permission of the user behavior characteristic area is obtained for the dynamic generation module. According to the dynamic password of the dynamic generation module, the data usage permission of the biometric area is obtained for verification by the second verification module.

[0045] As an embodiment of the present invention, as Figure 3 shown, the first verification module uses the data in the basic information area to verify the basic information. According to the result of the first verification module, the data usage permission of the behavior characteristic area is obtained. The dynamic generation module performs time series analysis on the current user behavior characteristic data and historical behavior characteristic data and generates a dynamic password. The time series analysis modeling process is as follows:

[0046] S1. Construct a time series analysis model:

[0047] S11. Describe the linear relationship between the current observation value and past observation values:

[0048]

[0049] Among them, X t is the observation value at the current time t, φ i is the autoregressive coefficient, ε t is the white noise error, p represents the order of the autoregressive term, indicating that the observation value at the current time is related to the observation values at the previous p time points, and c is a constant term.

[0050] S12. Describe the linear relationship between the current observation value and the random error term:

[0051]

[0052] Among them, X t is the observed value at the current moment t, θ i is the moving average coefficient, ε t is the white noise error at the current moment t, ε t-i is the white noise error at the moment (t - i), q represents the order of the moving average term, indicating that the observed value at the current moment is related to the error terms of the previous q time points, and c is a constant term.

[0053] S13. The comprehensive model is expressed as follows:

[0054] Z(p,q) = X t +Y t

[0055] S2. Use the historical behavior feature data to estimate and fit the parameters of the Z(p,q) model;

[0056] S3. Use the fitted Z(p,q) model to predict future values, specifically, use the current data point to generate the predicted value at the current time point through the model described in S513.

[0057] As an embodiment of the present invention, according to the dynamic password generated by the dynamic generation module, obtain the usage permission of the biometric area data, and the second verification module verifies the biometric data. The verification includes facial similarity recognition, palmprint similarity recognition, and voice similarity recognition. The similarity recognition model is designed as follows:

[0058] The input data are x1 and x2, the input feature mapping is expressed as y1 = g(x1), y2 = g(x2), and the similarity metric is expressed as: d = |y1 - y2| 2 , and the loss function is designed as:

[0059]

[0060] Among them, g(.) represents the CNN module with shared weights, N represents the number of training samples, Y represents the label. When Y = 1, it represents samples of the same category, and the goal is to make their distances as close as possible; when Y = 0, it represents samples of different categories, and the goal is to make their distances greater than the margin. The margin is a preset threshold, indicating the minimum distance that should exist between samples of different categories.

[0061] As an embodiment of the present invention, call the interface to log in to a third-party application

[0062] The user logs in to the application program, and the program requires the user to register including an account, password, and biometric data;

[0063] The user selects to jump to this system for generating a dynamic password for login;

[0064] First, verify the user's identity based on basic information such as the account password; after verification by the first module, call the user behavior characteristic data in the local database to generate a dynamic password;

[0065] Generate a dynamic password through the dynamic generation module, and obtain the corresponding biometric information in the local database for verification by the second verification module;

[0066] After verification by the second verification module, secure access is successfully achieved.

[0067] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A dynamic password generation system based on user behavior and biometrics. The dynamic password generation system includes a data recording module, an information updating module, and a local database. The data recording module is used to collect and record user information, and the user information includes user basic information and feature data. The feature data includes user behavior features and user biometric features. At the same time, a privacy protection strategy for user information is designed and the user information is stored in the local database. The information updating module is used to update the user information in the local database. The local database includes an access permission area, and the access permission area is divided into a user basic information area, a user behavior feature area, and a user biometric feature area. It is characterized in that, The dynamic password generation system further includes a first verification module, a second verification module, and a dynamic generation module, where: The first verification module uses the user basic information stored in the user basic information area to verify the basic information of the accessing user and gives a verification result: matching or not matching, where: If not matching, access is refused; If matching, access continues; The dynamic generation module, according to the "matching" verification result of the first verification module, obtains the usage permission of the user behavior feature area, obtains user behavior features, and performs time series analysis using the user behavior features and historical user behavior features to generate a dynamic password; The second verification module, according to the generated dynamic password, obtains the usage permission of the biometric feature area, obtains the user biometric features, and verifies the biometric features of the accessing user to achieve secure access; The information update module updates the local database according to the dynamic password generated by the dynamic generation module and the verification result of the second verification module to obtain the latest user behavior features and user biometric features.

2. The dynamic password generation system based on user behavior and biometrics according to claim 1, characterized in that, The user behavior characteristics include the login IP, click frequency, and usage period 。 3. The dynamic password generation system based on user behavior and biometric features according to claim 1, wherein, The user biometric features include face, palmprint, and voice.

4. A dynamic password generation system based on user behavior and biometric features according to claim 1, characterized in that, The specific operations of the data privacy protection policy are as follows: After the user data is collected, it is saved in the local database of device A, and an access permission area is divided. The access permission area includes a user basic information area, a user behavior feature area, and a user biometric feature area; The data in the user basic information area is used by the first verification module to verify the device basic information; According to the verification result of the first verification module, the usage permission of the user behavior feature area is obtained for the dynamic generation module; According to the dynamic password generated by the dynamic generation module, the usage permission of the user biometric feature area is obtained for the second verification module.

5. A dynamic password generation system based on user behavior and biometrics according to claim 1, characterized in that, The modeling process of the time series analysis includes the following steps: S51. Construct a time series analysis model: S511. Describe the linear relationship between the current observation value and the past observation values: where X t is the observed value at the current time t, φ i is the autoregressive coefficient, ε t is the white noise error, p represents the order of the autoregressive term, indicating that the observed value at the current time is related to the observed values at the previous p time points, and c is a constant term. S512. Describe the linear relationship between the current observation value and the random error term: where X t is the observed value at the current time t, θ i is the moving average coefficient, ε t is the white noise error at the current time t, ε t-i is the white noise error at time (t - i), q represents the order of the moving average term, indicating that the observed value at the current time is related to the error terms at the previous q time points, and c is a constant term. S513. The comprehensive model is expressed as follows: Z(p,q) = X t + Y t S52. Use the historical behavior feature data to estimate and fit the parameters of the Z(p,q) model; S53. Use the fitted Z(p,q) model to predict future values, specifically, use the current data point to generate the predicted value of the current time point through the model described in S513.

6. The dynamic password generation system based on user behavior and biometric features according to claim 1, wherein, The second verification module verifies the biometric features of the accessing user, including face similarity recognition, palmprint similarity recognition, and voice similarity recognition. The model design of the similarity recognition is as follows: The input data is x1, x2, the input feature mapping is represented as y1 = g(x1), y2 = g(x2), and the similarity metric is represented as: d = |y1 - y2| 2 , and the loss function is designed as: Among them, g(.) represents a CNN module with shared weights, N represents the number of training samples, Y represents the label. When Y = 1, it represents samples of the same category, and the goal is to make their distances as close as possible; when Y = 0, it represents samples of different categories, and the goal is to make their distances greater than the margin. The margin is a preset threshold, indicating the minimum distance that should exist between samples of different categories.

7. A storage structure, characterized in that, A dynamic password generation system based on user behavior and biometrics described in claims 1-6 is implemented through the cloud and encapsulated using DOCKER technology, and can be used as a plugin or interface.

Images