Electronic signature security management method and system based on artificial intelligence

Abstract

The invention relates to an electronic signature security management method and system based on artificial intelligence. The method comprises the following steps: collecting related data in an electronic signature process and carrying out common space embedding mapping to obtain an embedded signature image semantic coding feature vector, an embedded timestamp one-hot coding feature vector, an embedded IP address one-hot coding feature vector and an embedded user behavior semantic coding feature vector; and carrying out multi-mode feature embedding interaction modeling on the signature images to obtain space-time anchoring signature image-user behavior semantic interaction feature vectors so as to determine a signature behavior identification result. Therefore, the security and reliability of the electronic signature system can be improved, various security threats and fraudulent behaviors can be effectively handled, and more reliable electronic signature authentication service is provided for a digital office environment.

Description

Technical Field

[0001] The present disclosure relates to the technical field of electronic signatures, and in particular, to an artificial intelligence-based electronic signature security management method and system. Background Art

[0002] With the popularization of digital office systems and the widespread use of electronic documents, electronic signatures, as an important authentication method, are widely used in scenarios such as contract signing and document approval.

[0003] However, the security of electronic signatures has always been a concern. Traditional electronic signature systems have certain limitations in ensuring data security and preventing fraud, so more intelligent and efficient security management methods are needed. Summary of the invention

[0004] This summary is provided to introduce concepts in a brief form that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0005] In a first aspect, the present disclosure provides an artificial intelligence-based electronic signature security management method, the method comprising:

[0006] Collecting data related to the electronic signature process, wherein the data related to the electronic signature process includes signature image, timestamp, IP address and user behavior;

[0007] Performing co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp one-hot encoding feature vector, an embedded IP address one-hot encoding feature vector, and an embedded user behavior semantic encoding feature vector;

[0008] Performing multi-modal feature embedding interaction modeling on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector, and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector;

[0009] The recognition result of the signature behavior is determined based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0010] Optionally, a co-space embedding mapping is performed on each data item in the data related to the electronic signature process to obtain an embedded signature image semantic coding feature vector, an embedded timestamp one-hot coding feature vector, an embedded IP address one-hot coding feature vector and an embedded user behavior semantic coding feature vector, including: encoding each data item in the data related to the electronic signature process respectively to obtain a signature image semantic coding feature vector, a timestamp one-hot coding feature vector, an IP address one-hot coding feature vector and a user behavior semantic coding feature vector; passing the signature image semantic coding feature vector, the timestamp one-hot coding feature vector, the IP address one-hot coding feature vector and the user behavior semantic coding feature vector through a co-space embedder based on a multi-layer perceptron to obtain the embedded signature image semantic coding feature vector, the embedded timestamp one-hot coding feature vector, the embedded IP address one-hot coding feature vector and the embedded user behavior semantic coding feature vector.

[0011] Optionally, the signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector are passed through a co-space embedder based on a multi-layer perceptron to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector, including: fully connecting the signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector with multiple fully connected layers of the co-space embedder based on the multi-layer perceptron to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector.

[0012] Optionally, multimodal feature embedding interaction modeling is performed on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including: constructing a spatiotemporal anchored signature image semantic encoding feature vector and a spatiotemporal anchored user behavior semantic encoding feature vector based on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector; using a feature embedding interaction module to perform feature interaction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0013] Optionally, a spatiotemporal anchored signature image semantic encoding feature vector and a spatiotemporal anchored user behavior semantic encoding feature vector are constructed based on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector, including: concatenating the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector and the embedded IP address one-hot encoding feature vector to obtain the spatiotemporal anchored signature image semantic encoding feature vector; concatenating the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector to obtain the spatiotemporal anchored user behavior semantic encoding feature vector.

[0014] Optionally, a feature embedding interaction module is used to perform feature interaction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including: processing the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector using the following feature embedding interaction formula to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector; wherein the feature embedding interaction formula is:

[0015]

[0016] in, is the semantic encoding feature vector of the spatiotemporal anchored signature image, is the spatiotemporal anchored user behavior semantic encoding feature vector, v g is the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, t∈[0,1], t is the threshold value, represents the concatenation of the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector, W g is the transformation vector, b is the bias parameter, and sigmoid represents the activation function.

[0017] Optionally, the recognition result of the signing behavior is determined based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including: passing the spatiotemporal anchored signature image-user behavior semantic interaction feature vector through a classifier-based identifier to obtain the recognition result, and the recognition result is used to indicate whether the signing behavior is abnormal.

[0018] Optionally, the spatiotemporal anchored signature image-user behavior semantic interaction feature vector is passed through a classifier-based identifier to obtain the recognition result, and the recognition result is used to indicate whether the signing behavior is abnormal, including: using multiple fully connected layers of the classifier to fully connect encode the spatiotemporal anchored signature image-user behavior semantic interaction feature vector to obtain an encoded classification feature vector; and, passing the encoded classification feature vector through the Softmax classification function of the classifier to obtain the classification result.

[0019] In a second aspect, the present disclosure provides an artificial intelligence-based electronic signature security management system, the system comprising:

[0020] An electronic signature process related data collection module, used to collect electronic signature process related data, wherein the electronic signature process related data includes signature image, timestamp, IP address and user behavior;

[0021] A co-space embedding mapping module is used to perform co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp one-hot encoding feature vector, an embedded IP address one-hot encoding feature vector, and an embedded user behavior semantic encoding feature vector;

[0022] A multi-modal feature embedding interaction module, used for performing multi-modal feature embedding interaction modeling on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector;

[0023] The signature behavior recognition result determination module is used to determine the signature behavior recognition result based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0024] By adopting the above technical solution, data related to the electronic signature process is collected and co-space embedding mapping is performed to obtain the semantic encoding feature vector of the embedded signature image, the unique hot encoding feature vector of the embedded timestamp, the unique hot encoding feature vector of the embedded IP address, and the semantic encoding feature vector of the embedded user behavior; multi-modal feature embedding interaction modeling is performed to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector to determine the recognition result of the signature behavior. In this way, not only can the security and reliability of the electronic signature system be improved, but also various security threats and fraudulent behaviors can be effectively dealt with, providing a more reliable electronic signature authentication service for the digital office environment.

[0025] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale. In the drawings:

[0027] Figure 1 The present invention is a flowchart of an artificial intelligence-based electronic signature security management method according to an exemplary embodiment.

[0028] Figure 2 It is a block diagram of an electronic signature security management system based on artificial intelligence according to an exemplary embodiment.

[0029] Figure 3 It is a block diagram of an electronic device according to an exemplary embodiment.

[0030] Figure 4 It is an application scenario diagram of an artificial intelligence-based electronic signature security management method according to an exemplary embodiment. DETAILED DESCRIPTION

[0031] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0032] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0033] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0034] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0035] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0036] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0037] First of all, in response to the above technical problems, the technical concept of this application is: to use intelligent algorithms to conduct in-depth analysis of key data in the electronic signature process to identify potential security threats. Specifically, collect and analyze data related to the electronic signature process, such as signature images, timestamps, IP addresses, and user behavior information, and perform multimodal fusion and co-space embedding comparison on these data to construct a comprehensive feature representation of the electronic signature process, and perform correlation analysis and fusion of signature images and user behaviors while anchoring time and space information, so as to achieve abnormal identification of electronic signature behavior. In this way, not only can the security and reliability of the electronic signature system be improved, but also various security threats and fraudulent behaviors can be effectively responded to, providing a more reliable electronic signature authentication service for the digital office environment.

[0038] In order to solve the above problems, the present disclosure provides an artificial intelligence-based electronic signature security management method and system, which collects data related to the electronic signature process and performs co-space embedding mapping to obtain embedded signature image semantic coding feature vectors, embedded timestamp unique hot coding feature vectors, embedded IP address unique hot coding feature vectors and embedded user behavior semantic coding feature vectors; multi-modal feature embedding interaction modeling is performed to obtain spatiotemporal anchored signature image-user behavior semantic interaction feature vectors to determine the recognition results of signature behavior. In this way, not only can the security and reliability of the electronic signature system be improved, but also various security threats and fraudulent behaviors can be effectively dealt with, providing a more reliable electronic signature authentication service for the digital office environment.

[0039] The specific embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0040] Figure 1 is a flow chart of an artificial intelligence-based electronic signature security management method according to an exemplary embodiment. Figure 1 As shown, the method includes:

[0041] Step 101: Collecting data related to the electronic signature process, wherein the data related to the electronic signature process includes signature image, timestamp, IP address and user behavior;

[0042] Step 102, performing co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp one-hot encoding feature vector, an embedded IP address one-hot encoding feature vector, and an embedded user behavior semantic encoding feature vector;

[0043] Step 103, performing multi-modal feature embedding interaction modeling on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector, and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector;

[0044] Step 104: Determine the recognition result of the signature behavior based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0045] Based on this, in the technical solution of the present application, first, data related to the electronic signature process is collected, wherein the data related to the electronic signature process includes a signature image, a timestamp, an IP address, and a user behavior. Here, the signature image and the user behavior can provide identity information about the signatory; the timestamp can provide time information of the signature behavior and anchor the time information of the signature behavior; the IP address can trace the source of the signature behavior and anchor the spatial information of the signature behavior. In particular, since the signature image and the user behavior are forgeable, further analysis and processing thereof can identify abnormal patterns therefrom, thereby realizing abnormal identification of electronic signature behavior. For example, the user behavior data characterizes the user's signature habits and behavior patterns. If abnormal behavioral characteristics appear, it may indicate that there is an abnormality in the electronic signature behavior.

[0046] In one embodiment of the present disclosure, a co-space embedding mapping is performed on each data item in the electronic signature process related data to obtain an embedded signature image semantic coding feature vector, an embedded timestamp one-hot coding feature vector, an embedded IP address one-hot coding feature vector and an embedded user behavior semantic coding feature vector, including: encoding each data item in the electronic signature process related data respectively to obtain a signature image semantic coding feature vector, a timestamp one-hot coding feature vector, an IP address one-hot coding feature vector and a user behavior semantic coding feature vector; passing the signature image semantic coding feature vector, the timestamp one-hot coding feature vector, the IP address one-hot coding feature vector and the user behavior semantic coding feature vector through a co-space embedder based on a multi-layer perceptron to obtain the embedded signature image semantic coding feature vector, the embedded timestamp one-hot coding feature vector, the embedded IP address one-hot coding feature vector and the embedded user behavior semantic coding feature vector.

[0047] Then, each data item in the electronic signature process related data is encoded to obtain a signature image semantic encoding feature vector, a timestamp one-hot encoding feature vector, an IP address one-hot encoding feature vector, and a user behavior semantic encoding feature vector. Those skilled in the art should understand that one-hot encoding is a commonly used data encoding technology for converting data into a form that can be better understood by a machine learning algorithm. In one-hot encoding, each different value is represented as a vector with a length equal to the total number of categories, in which only one element is 1 and the other elements are 0. Such an encoding method can effectively represent different values ​​of variables and avoid the model from misinterpreting discrete variables as continuous variables. In the abnormal identification process of electronic signatures, each data item in the electronic signature process related data is one-hot encoded, and the data structure that is difficult to be read and recognized by the computer in each data item can be converted into a vector representation for subsequent analysis and processing.

[0048] Considering that the signature image semantic encoding feature vector, the timestamp unique hot encoding feature vector, the IP address unique hot encoding feature vector and the user behavior semantic encoding feature vector are derived from data of different modalities, their feature spaces may be highly inconsistent. If they are directly fused and compared, it may lead to mismatch of feature space, affecting the learning of features by the model. In addition, since data of different modalities may have different feature scales, direct fusion will lead to scale differences between features, and may also lead to loss or confusion of information, and fail to retain the unique information and features of each modality data well. Therefore, in order to avoid or alleviate the occurrence of the above problems, in the technical solution of the present application, the signature image semantic encoding feature vector, the timestamp unique hot encoding feature vector, the IP address unique hot encoding feature vector and the user behavior semantic encoding feature vector are obtained by a co-space embedder based on a multi-layer perceptron to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp unique hot encoding feature vector, an embedded IP address unique hot encoding feature vector and an embedded user behavior semantic encoding feature vector. In this way, mapping multimodal encoded data into a common space can realize unified processing of data and provide a richer and more effective feature representation for subsequent data analysis.

[0049] In one embodiment of the present disclosure, the signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector are passed through a co-space embedder based on a multi-layer perceptron to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector, including: using multiple fully connected layers of the co-space embedder based on the multi-layer perceptron to fully connect encode the signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector.

[0050] Furthermore, multimodal feature embedding interaction modeling is performed on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including: constructing a spatiotemporal anchored signature image semantic encoding feature vector and a spatiotemporal anchored user behavior semantic encoding feature vector based on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector; using a feature embedding interaction module to perform feature interaction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0051] Subsequently, the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector and the embedded IP address one-hot encoding feature vector are concatenated to obtain a spatiotemporal anchored signature image semantic encoding feature vector. The timestamp and IP address provide information about the time and location of the signature behavior. By concatenating this information with the semantic encoding features of the signature image, the spatiotemporal information can be combined to more comprehensively describe the background and environment of the signature image.

[0052] At the same time, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector are spliced ​​to obtain a spatiotemporal anchored user behavior semantic encoding feature vector. Similarly, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector respectively provide the time, place and user behavior content information of the electronic signature behavior. By splicing this information, the spatiotemporal information can be combined with the user behavior information to characterize and describe the user behavior characteristics in a more comprehensive and integrated manner, which helps to better understand the user's behavior patterns and characteristics.

[0053] In one embodiment of the present disclosure, a spatiotemporal anchored signature image semantic encoding feature vector and a spatiotemporal anchored user behavior semantic encoding feature vector are constructed based on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector, including: concatenating the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector and the embedded IP address one-hot encoding feature vector to obtain the spatiotemporal anchored signature image semantic encoding feature vector; concatenating the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector to obtain the spatiotemporal anchored user behavior semantic encoding feature vector.

[0054] Furthermore, a feature embedding interaction module is used to perform feature interaction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector. Here, considering that both signature image information and user behavior information contain important features and information about signature behavior, by interactively fusing the two, the complex associations and dependencies between the two information sources can be fully utilized to form a comprehensive representation of electronic signature behavior. Among them, during the process of feature interactive fusion, the feature embedding interaction module adaptively balances the importance of the fused information from the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector through a threshold value, so that the model can flexibly adjust the attention to information from different modalities and the importance of energy consumption data evolution characteristics at different periodic time scales, thereby optimizing the fusion process of comprehensive feature representation.

[0055] In one embodiment of the present disclosure, a feature embedding interaction module is used to perform feature interaction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including: processing the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector using the following feature embedding interaction formula to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector; wherein the feature embedding interaction formula is:

[0056]

[0057] in, is the semantic encoding feature vector of the spatiotemporal anchored signature image, is the spatiotemporal anchored user behavior semantic encoding feature vector, v g is the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, t∈[0,1], t is the threshold value, represents the concatenation of the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector, W g is the transformation vector, b is the bias parameter, and sigmoid represents the activation function.

[0058] Furthermore, the recognition result of the signing behavior is determined based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including: passing the spatiotemporal anchored signature image-user behavior semantic interaction feature vector through a classifier-based identifier to obtain the recognition result, and the recognition result is used to indicate whether the signing behavior is abnormal.

[0059] In one embodiment of the present disclosure, the spatiotemporal anchored signature image-user behavior semantic interaction feature vector is passed through a classifier-based recognizer to obtain the recognition result, and the recognition result is used to indicate whether the signature behavior is abnormal, including: using multiple fully connected layers of the classifier to fully connect encode the spatiotemporal anchored signature image-user behavior semantic interaction feature vector to obtain an encoded classification feature vector; and passing the encoded classification feature vector through the Softmax classification function of the classifier to obtain the classification result.

[0060] In a preferred embodiment, the spatiotemporal anchored signature image-user behavior semantic interaction feature vector is passed through a classifier-based recognizer to obtain a recognition result, including:

[0061] Determine a spatiotemporal anchored signature image-user behavior semantic interaction probability value corresponding to the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, wherein the spatiotemporal anchored signature image-user behavior semantic interaction probability value indicates a probability of abnormal signature behavior;

[0062] Calculate the product of the first spatiotemporal anchored signature image-user behavior semantic interaction feature value of the spatiotemporal anchored signature image-user behavior semantic interaction feature vector and the spatiotemporal anchored signature image-user behavior semantic interaction probability value to obtain a first spatiotemporal anchored signature image-user behavior semantic interaction feature probability product;

[0063] Calculate the product of the second spatiotemporal anchored signature image-user behavior semantic interaction feature value of the spatiotemporal anchored signature image-user behavior semantic interaction feature vector and the difference between one and one minus the spatiotemporal anchored signature image-user behavior semantic interaction probability value to obtain a second spatiotemporal anchored signature image-user behavior semantic interaction feature inverse probability product;

[0064] Calculating the quotient of the first spatiotemporal anchored signature image-user behavior semantic interaction feature value divided by one minus the difference between the spatiotemporal anchored signature image-user behavior semantic interaction probability value to obtain a first spatiotemporal anchored signature image-user behavior semantic interaction feature inverse probability quotient;

[0065] Calculating the quotient of the second spatiotemporal anchored signature image-user behavior semantic interaction feature value and the spatiotemporal anchored signature image-user behavior semantic interaction probability value to obtain a second spatiotemporal anchored signature image-user behavior semantic interaction feature probability quotient;

[0066] After adding the first spatiotemporal anchored signature image-user behavior semantic interaction feature probability product and the second spatiotemporal anchored signature image-user behavior semantic interaction feature inverse probability product, subtracting the difference between the first spatiotemporal anchored signature image-user behavior semantic interaction feature inverse probability quotient and the second spatiotemporal anchored signature image-user behavior semantic interaction feature probability quotient to obtain a spatiotemporal anchored signature image-user behavior semantic interaction correction feature value;

[0067] For each first spatiotemporal anchored signature image-user behavior semantic interaction feature value and second spatiotemporal anchored signature image-user behavior semantic interaction feature value of the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, matrix multiplying the spatiotemporal anchored signature image-user behavior semantic interaction correction feature matrix composed of the spatiotemporal anchored signature image-user behavior semantic interaction correction feature values ​​with the spatiotemporal anchored signature image-user behavior semantic interaction feature vector to obtain a corrected spatiotemporal anchored signature image-user behavior semantic interaction feature vector;

[0068] The corrected spatiotemporal anchored signature image-user behavior semantic interaction feature vector is passed through a classifier-based recognizer to obtain a recognition result.

[0069] The correction process of the spatiotemporal anchored signature image-user behavior semantic interaction feature vector V is expressed as:

[0070]

[0071] m i,j ∈M,v i ∈V,v j(j≠i) ∈V

[0072]

[0073] Where V represents the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, v i The first spatiotemporal anchored signature image-user behavior semantic interaction feature value representing the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, v j The second spatiotemporal anchored signature image-user behavior semantic interaction feature value representing the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, p represents the probability of abnormal signature behavior, m i,j represents the spatiotemporal anchored signature image-user behavior semantic interaction correction feature value, M represents the spatiotemporal anchored signature image-user behavior semantic interaction correction feature matrix composed of the spatiotemporal anchored signature image-user behavior semantic interaction correction feature value, represents matrix multiplication, and V' represents the corrected spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0074] In the above technical solution, the spatiotemporally anchored signature image semantic encoding feature vector and the spatiotemporally anchored user behavior semantic encoding feature vector respectively express the spatiotemporally anchored user behavior semantic features in the embedding space and the image semantic features of the spatiotemporally anchored signature image in the embedding space. However, considering that the signature image and the user behavior are source data of different modalities, this will lead to structural differences in the feature distribution between the spatiotemporally anchored signature image semantic encoding feature vector and the spatiotemporally anchored user behavior semantic encoding feature vector. When the feature embedding interaction module is used to perform feature interaction on the spatiotemporally anchored signature image semantic encoding feature vector and the spatiotemporally anchored user behavior semantic encoding feature vector, the obtained spatiotemporally anchored signature image-user behavior semantic interaction feature vector will also have semantic representation diversity due to the difference in feature distribution structure, thereby affecting the iterative consistency of the classification mapping and reducing the accuracy of the classification results.

[0075] Therefore, the present application targets the region-boundary integral relationship of the high-dimensional feature manifold of the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, and approximates the simply connected region representation of the high-dimensional manifold composed of eigenvalue pairs of the feature set of the spatiotemporal anchored signature image-user behavior semantic interaction feature vector by defining the overall probability distribution representation of the feature set of the spatiotemporal anchored signature image-user behavior semantic interaction feature vector as a manifold constraint boundary, so as to avoid the ambiguity in the mapping of the diversified feature representation of the spatiotemporal anchored signature image-user behavior semantic interaction feature vector to the local manifold representation in the high-dimensional feature class convergence space, improve the iterative consistency of the execution of each local feature distribution in the mapping task, so as to improve the feature classification convergence effect, and improve the accuracy of the recognition result obtained by the classifier-based recognizer for the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0076] In summary, the above scheme uses intelligent algorithms to conduct in-depth analysis of key data in the electronic signature process to identify potential security threats. Specifically, collect and analyze data related to the electronic signature process, such as signature images, timestamps, IP addresses, and user behavior information. By performing multimodal fusion and co-space embedding comparison on these data, a comprehensive feature representation of the electronic signature process is constructed. While anchoring the spatiotemporal information, the signature image and user behavior are correlated and analyzed and fused, thereby realizing the abnormal identification of electronic signature behavior. In this way, not only can the security and reliability of the electronic signature system be improved, but also various security threats and fraudulent behaviors can be effectively responded to, providing a more reliable electronic signature authentication service for the digital office environment.

[0077] Figure 2 FIG. 1 is a block diagram of an electronic signature security management system based on artificial intelligence according to an exemplary embodiment. Figure 2 As shown, the system 200 includes:

[0078] The electronic signature process related data collection module 201 is used to collect the electronic signature process related data, wherein the electronic signature process related data includes signature image, timestamp, IP address and user behavior;

[0079] A co-space embedding mapping module 202 is used to perform co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp one-hot encoding feature vector, an embedded IP address one-hot encoding feature vector, and an embedded user behavior semantic encoding feature vector;

[0080] The multi-modal feature embedding interaction module 203 is used to perform multi-modal feature embedding interaction modeling on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector;

[0081] The signature behavior recognition result determination module 204 is used to determine the signature behavior recognition result based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0082] Reference below Figure 3 , which shows a schematic diagram of the structure of an electronic device 600 suitable for implementing the embodiment of the present disclosure. The terminal device in the embodiment of the present disclosure may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0083] like Figure 3 As shown, the electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0084] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 3 The electronic device 600 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0085] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0086] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0087] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0088] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0089] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0090] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0091] The modules involved in the embodiments described in the present disclosure may be implemented by software or hardware. The name of a module does not limit the module itself in some cases. For example, a test parameter acquisition module may also be described as a "module for acquiring device test parameters corresponding to a target device".

[0092] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0093] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0094] According to one or more embodiments of the present disclosure, Example 1 provides an artificial intelligence-based electronic signature security management method, the method comprising:

[0095] Collecting data related to the electronic signature process, wherein the data related to the electronic signature process includes signature image, timestamp, IP address and user behavior;

[0096] Performing co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp one-hot encoding feature vector, an embedded IP address one-hot encoding feature vector, and an embedded user behavior semantic encoding feature vector;

[0097] Performing multi-modal feature embedding interaction modeling on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector, and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector;

[0098] The recognition result of the signature behavior is determined based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0099] According to one or more embodiments of the present disclosure, Example 2 provides the method of Example 1, performing a co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp unique hot encoding feature vector, an embedded IP address unique hot encoding feature vector, and an embedded user behavior semantic encoding feature vector, including:

[0100] Encode each data item in the electronic signature process related data respectively to obtain a signature image semantic encoding feature vector, a timestamp one-hot encoding feature vector, an IP address one-hot encoding feature vector, and a user behavior semantic encoding feature vector;

[0101] The signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector are passed through a co-space embedder based on a multi-layer perceptron to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector.

[0102] According to one or more embodiments of the present disclosure, Example 3 provides the method of Example 2, wherein the signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector are passed through a co-space embedder based on a multi-layer perceptron to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector, including: fully connecting the signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector with multiple fully connected layers of the co-space embedder based on the multi-layer perceptron to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector.

[0103] According to one or more embodiments of the present disclosure, Example 4 provides the method of Example 3, performing multi-modal feature embedding interaction modeling on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector, and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including:

[0104] Constructing a spatiotemporal anchored signature image semantic encoding feature vector and a spatiotemporal anchored user behavior semantic encoding feature vector based on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector;

[0105] A feature embedding interaction module is used to perform feature interaction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0106] According to one or more embodiments of the present disclosure, Example 5 provides the method of Example 4, constructing a spatiotemporal anchored signature image semantic encoding feature vector and a spatiotemporal anchored user behavior semantic encoding feature vector based on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector, and the embedded user behavior semantic encoding feature vector, including:

[0107] Concatenate the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, and the embedded IP address one-hot encoding feature vector to obtain the spatiotemporal anchored signature image semantic encoding feature vector;

[0108] The embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector are concatenated to obtain the spatiotemporal anchored user behavior semantic encoding feature vector.

[0109] According to one or more embodiments of the present disclosure, Example 6 provides the method of Example 5, using a feature embedding interaction module to perform feature interaction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including:

[0110] The spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector are processed using the following feature embedding interaction formula to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector; wherein the feature embedding interaction formula is:

[0111]

[0112] in, is the semantic encoding feature vector of the spatiotemporal anchored signature image, is the spatiotemporal anchored user behavior semantic encoding feature vector, vg is the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, t∈[0,1], t is the threshold value, represents the concatenation of the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector, W g is the transformation vector, b is the bias parameter, and sigmoid represents the activation function.

[0113] According to one or more embodiments of the present disclosure, Example 7 provides the method of Example 6, which determines the recognition result of the signature behavior based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including:

[0114] Performing feature distribution correction on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector;

[0115] The spatiotemporal anchored signature image-user behavior semantic interaction feature vector is passed through a classifier-based recognizer to obtain the recognition result, and the recognition result is used to indicate whether the signature behavior is abnormal.

[0116] According to one or more embodiments of the present disclosure, Example 8 provides the method of Example 7, performing feature distribution correction on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including: performing fusion correction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain a corrected feature vector; fusing the corrected feature vector and the spatiotemporal anchored signature image-user behavior semantic interaction feature vector to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0117] According to one or more embodiments of the present disclosure, Example 9 provides the method of Example 8, wherein the spatiotemporal anchored signature image-user behavior semantic interaction feature vector is passed through a classifier-based recognizer to obtain the recognition result, and the recognition result is used to indicate whether the signature behavior is abnormal, including:

[0118] Using multiple fully connected layers of the classifier to fully connect encode the spatiotemporal anchor signature image-user behavior semantic interaction feature vector to obtain an encoded classification feature vector; and

[0119] The encoded classification feature vector is passed through the Softmax classification function of the classifier to obtain the classification result.

[0120] According to one or more embodiments of the present disclosure, Example 10 provides an artificial intelligence-based electronic signature security management system, the system comprising:

[0121] An electronic signature process related data collection module, used to collect electronic signature process related data, wherein the electronic signature process related data includes signature image, timestamp, IP address and user behavior;

[0122] A co-space embedding mapping module is used to perform co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp one-hot encoding feature vector, an embedded IP address one-hot encoding feature vector, and an embedded user behavior semantic encoding feature vector;

[0123] A multi-modal feature embedding interaction module, used for performing multi-modal feature embedding interaction modeling on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector;

[0124] The signature behavior recognition result determination module is used to determine the signature behavior recognition result based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

[0125] Figure 4 FIG. 1 is an application scenario diagram of an artificial intelligence-based electronic signature security management method according to an exemplary embodiment. Figure 4 As shown, in this application scenario, first, collect the data related to the electronic signature process (for example, Figure 4 Then, the acquired electronic signature process related data is input into a server (for example, Figure 4 In the S) shown in , the server is capable of processing the electronic signature process related data based on an artificial intelligence-based electronic signature security management algorithm to determine the recognition result of the signature behavior.

[0126] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

[0127] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0128] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims. Regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.

Claims

1. An artificial intelligence-based electronic signature security management method, characterized in that: include: Collecting data related to the electronic signature process, wherein the data related to the electronic signature process includes signature image, timestamp, IP address and user behavior; Performing co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp one-hot encoding feature vector, an embedded IP address one-hot encoding feature vector, and an embedded user behavior semantic encoding feature vector; Performing multi-modal feature embedding interaction modeling on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector, and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector; The recognition result of the signature behavior is determined based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

2. The method for electronic signature security management based on artificial intelligence according to claim 1 is characterized in that: Performing co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp unique hot encoding feature vector, an embedded IP address unique hot encoding feature vector, and an embedded user behavior semantic encoding feature vector, including: Encode each data item in the electronic signature process related data respectively to obtain a signature image semantic encoding feature vector, a timestamp one-hot encoding feature vector, an IP address one-hot encoding feature vector, and a user behavior semantic encoding feature vector; The signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector are passed through a co-space embedder based on a multi-layer perceptron to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector.

3. The method for electronic signature security management based on artificial intelligence according to claim 2 is characterized in that: The signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector are passed through a co-space embedder based on a multi-layer perceptron to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector, including: using multiple fully connected layers of the co-space embedder based on the multi-layer perceptron to fully connect encode the signature image semantic encoding feature vector, the timestamp one-hot encoding feature vector, the IP address one-hot encoding feature vector and the user behavior semantic encoding feature vector to obtain the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector.

4. The method for electronic signature security management based on artificial intelligence according to claim 3 is characterized in that: The embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector, and the embedded user behavior semantic encoding feature vector are subjected to multimodal feature embedding interaction modeling to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including: Constructing a spatiotemporal anchored signature image semantic encoding feature vector and a spatiotemporal anchored user behavior semantic encoding feature vector based on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector; A feature embedding interaction module is used to perform feature interaction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

5. The method for electronic signature security management based on artificial intelligence according to claim 4 is characterized in that: Constructing a spatiotemporal anchored signature image semantic encoding feature vector and a spatiotemporal anchored user behavior semantic encoding feature vector based on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector, and the embedded user behavior semantic encoding feature vector, including: Concatenate the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, and the embedded IP address one-hot encoding feature vector to obtain the spatiotemporal anchored signature image semantic encoding feature vector; The embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector are concatenated to obtain the spatiotemporal anchored user behavior semantic encoding feature vector.

6. The method for electronic signature security management based on artificial intelligence according to claim 5 is characterized in that: Using a feature embedding interaction module to perform feature interaction on the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, including: The spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector are processed using the following feature embedding interaction formula to obtain the spatiotemporal anchored signature image-user behavior semantic interaction feature vector; wherein the feature embedding interaction formula is: in, is the semantic encoding feature vector of the spatiotemporal anchored signature image, is the spatiotemporal anchored user behavior semantic encoding feature vector, v g is the spatiotemporal anchored signature image-user behavior semantic interaction feature vector, t∈[0,1], t is the threshold value, represents the concatenation of the spatiotemporal anchored signature image semantic encoding feature vector and the spatiotemporal anchored user behavior semantic encoding feature vector, W g is the transformation vector, b is the bias parameter, and sigmoid represents the activation function.

7. The method for electronic signature security management based on artificial intelligence according to claim 6 is characterized in that: Determining the recognition result of the signature behavior based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector includes: The spatiotemporal anchored signature image-user behavior semantic interaction feature vector is passed through a classifier-based recognizer to obtain the recognition result, and the recognition result is used to indicate whether the signature behavior is abnormal.

8. The method for electronic signature security management based on artificial intelligence according to claim 7 is characterized in that: The spatiotemporal anchored signature image-user behavior semantic interaction feature vector is passed through a classifier-based recognizer to obtain the recognition result, which is used to indicate whether the signature behavior is abnormal, including: Using multiple fully connected layers of the classifier to fully connect encode the spatiotemporal anchor signature image-user behavior semantic interaction feature vector to obtain an encoded classification feature vector; and The encoded classification feature vector is passed through the Softmax classification function of the classifier to obtain the classification result.

9. An artificial intelligence-based electronic signature security management system, characterized in that: include: An electronic signature process related data collection module, used to collect electronic signature process related data, wherein the electronic signature process related data includes signature image, timestamp, IP address and user behavior; A co-space embedding mapping module is used to perform co-space embedding mapping on each data item in the electronic signature process related data to obtain an embedded signature image semantic encoding feature vector, an embedded timestamp one-hot encoding feature vector, an embedded IP address one-hot encoding feature vector, and an embedded user behavior semantic encoding feature vector; A multi-modal feature embedding interaction module, used for performing multi-modal feature embedding interaction modeling on the embedded signature image semantic encoding feature vector, the embedded timestamp one-hot encoding feature vector, the embedded IP address one-hot encoding feature vector and the embedded user behavior semantic encoding feature vector to obtain a spatiotemporal anchored signature image-user behavior semantic interaction feature vector; The signature behavior recognition result determination module is used to determine the signature behavior recognition result based on the spatiotemporal anchored signature image-user behavior semantic interaction feature vector.

Images