CN114880709A - E-commerce data protection method and server applying artificial intelligence - Google Patents

E-commerce data protection method and server applying artificial intelligence Download PDF

Info

Publication number
CN114880709A
CN114880709A CN202210561265.3A CN202210561265A CN114880709A CN 114880709 A CN114880709 A CN 114880709A CN 202210561265 A CN202210561265 A CN 202210561265A CN 114880709 A CN114880709 A CN 114880709A
Authority
CN
China
Prior art keywords
big data
user
cross
border
behavior
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202210561265.3A
Other languages
Chinese (zh)
Other versions
CN114880709B (en
Inventor
蔡忠铸
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Yanqi Huawei Information System Technology Co ltd
Original Assignee
Tongren Yingdan Network Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tongren Yingdan Network Technology Co ltd filed Critical Tongren Yingdan Network Technology Co ltd
Priority to CN202210561265.3A priority Critical patent/CN114880709B/en
Publication of CN114880709A publication Critical patent/CN114880709A/en
Application granted granted Critical
Publication of CN114880709B publication Critical patent/CN114880709B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/60—Protecting data
    • G06F21/62—Protecting access to data via a platform, e.g. using keys or access control rules
    • G06F21/6218—Protecting access to data via a platform, e.g. using keys or access control rules to a system of files or objects, e.g. local or distributed file system or database
    • G06F21/6245—Protecting personal data, e.g. for financial or medical purposes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2474—Sequence data queries, e.g. querying versioned data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

The invention provides an artificial intelligence applied E-commerce data protection method and a server, an algorithm variable of a first AI data processing thread is obtained through interactive behavior change information of at least two sets of cross-border E-commerce user big data with continuous time sequence relation, the first AI data processing thread is positioned in an image mining node group of a second AI data processing thread, when the individual privacy protection processing is carried out on one set of first cross-border E-commerce user big data in at least two sets of cross-border E-commerce user big data with continuous time sequence relation based on the second AI data processing thread, the behavior change content between different sets of cross-border E-commerce user big data is concerned, the precision and the credibility of positioning the individual privacy images can be ensured, the probability of misjudging the group privacy images into individual privacy is reduced, and the pertinence of the individual privacy protection processing on the cross-border E-commerce user big data is improved, ensuring maximum availability of cross-border e-commerce user big data after anonymous processing.

Description

E-commerce data protection method and server applying artificial intelligence
Technical Field
The invention relates to the technical field of cross-border e-commerce, in particular to an e-commerce data protection method and a server applying artificial intelligence.
Background
The e-commerce, also called e-commerce, refers to the business interaction activities performed by multiple parties in different regions, including online ordering, remote payment, etc., and cross-border logistics, distribution, etc. In recent years, with the rapid development of artificial intelligence, cross-border e-commerce has a close relationship with the internet, for example, user data can be analyzed through an artificial intelligence technology to realize product push, service upgrade and the like. In the service operation process of cross-border e-commerce, protection for user data is very important, and user privacy protection is taken as one of protection key points, and the related technology still has difficulty in realizing targeted privacy protection processing.
Disclosure of Invention
The invention provides an e-commerce data protection method and a server applying artificial intelligence, and adopts the following technical scheme in order to achieve the technical purpose.
The first aspect is an e-commerce data protection method applying artificial intelligence, which is applied to an artificial intelligence server, and the method at least comprises the following steps: determining no less than two groups of interaction behavior change information of cross-border electric commercial user big data with continuous time sequence relation; determining an algorithm variable of a first AI data processing thread according to the at least two groups of interaction behavior change information of the cross-border electric power business user big data with continuous time sequence relation; and carrying out individual privacy protection processing on first cross-border electric commercial user big data in the at least two sets of cross-border electric commercial user big data with continuous time sequence relation based on a second AI data processing thread to obtain second cross-border electric commercial user big data corresponding to the first cross-border electric commercial user big data, wherein the first cross-border electric commercial user big data is one of the at least two sets of cross-border electric commercial user big data with continuous time sequence relation, the data anonymization degree of the second cross-border electric commercial user big data is greater than that of the first cross-border electric commercial user big data, and an image mining node group of the second AI data processing thread comprises at least one first AI data processing thread.
By the design, algorithm variables of the first AI data processing thread are obtained through interactive behavior change information of at least two groups of cross-border electric commercial user big data with continuous time sequence relation, wherein at least one first AI data processing thread is positioned in the portrait mining node group of the second AI data processing thread, when the individual privacy protection processing is carried out on one group of first cross-border electric commercial user big data in at least two groups of cross-border electric commercial user big data with continuous time sequence relation based on the second AI data processing thread, the behavior change content between different groups of cross-border electric commercial user big data is concerned, the accuracy and the credibility of individual privacy portrait positioning can be ensured, thereby the probability of misjudging the group privacy portrait into individual privacy portrait can be reduced, and the pertinence of individual privacy protection processing on the cross-border electric commercial user big data is improved, ensuring maximum availability of cross-border e-commerce user big data after anonymous processing.
In some possible embodiments, the interactive behavior change information covers at least one scale of user behavior representation fields between the at least two sets of cross-border power consumer big data with continuous time sequence relation; the determination of the interactive behavior change information of at least two groups of cross-border power provider user big data with continuous time sequence relation comprises the following steps: carrying out big data mining on the big data of at least two sets of cross-border electric business users with continuous time sequence relation to obtain a set of electric business activity descriptions, wherein the set of electric business activity descriptions comprises electric business activity descriptions of a plurality of scales; determining a first user behavioral representation field from the first electrical business activity description based on the electrical business activity description being a first electrical business activity description of lowest scale in the set of electrical business activity descriptions; determining a second user behavioral profile field by the second electrical business activity description, a third electrical business activity description associated with the size of the second electrical business activity description, and a user behavioral profile field that is consistent with the size of the second electrical business activity description, the interactive behavior change information encompassing the first user behavioral profile field and the second user behavioral profile field, on the basis that the electrical business activity description is not the second lowest-sized electrical business activity description of the set of electrical business activity descriptions.
By the design, the interactive behavior change information of at least two groups of cross-border electric business user big data with continuous time sequence relation comprises the behavior change content between the continuous groups of cross-border electric business user big data, and the algorithm variable of the first AI data processing thread is determined based on the behavior change content between the continuous two groups of cross-border electric business user big data, so that the second AI data processing thread can accurately position the image field needing anonymization processing.
Under some possible embodiments, the determining no less than two sets of interaction behavior change information of cross-border electric business user big data with continuous time sequence relation comprises: and determining the interactive behavior change information of at least two groups of cross-border electricity business user big data with continuous time sequence relation through a user behavior change analysis thread.
By the design, the artificial intelligence server determines the interactive behavior change information between at least two groups of cross-border electric commercial user big data with continuous time sequence relation based on the user behavior change analysis thread, so that group user portrait and individual user portrait can be accurately distinguished, and the accuracy and pertinence of anonymization processing are improved.
Under some possible embodiments, the determining, based on the information about the change of the interaction behavior of the at least two sets of cross-border electricity business user big data with continuous time sequence relationship, an algorithm variable of the first AI data processing thread includes: vector processing is carried out on the second electric business activity description, a third electric business activity description which is associated with the scale of the second electric business activity description and a user behavior portrait field which is consistent with the scale of the second electric business activity description to obtain an algorithm variable of the first AI data processing thread.
In some possible embodiments, the performing, by the second AI data processing thread, individual privacy protection processing on a first cross-border e-commerce user big data of the at least two sets of cross-border e-commerce user big data having a continuous time sequence relationship to obtain a second cross-border e-commerce user big data corresponding to the first cross-border e-commerce user big data includes: anonymizing the first cross-border electric business user big data through at least one first AI data processing thread to obtain an anonymized behavior portrait field of the first cross-border electric business user big data; and performing big data translation processing on the anonymized behavior portrait field of the big data of the first cross-border e-commerce user to obtain big data of a second cross-border e-commerce user.
By the design, the first AI data processing thread is configured to the image mining node group in the second AI data processing thread, the thread configuration quality of the AI data processing thread for translating the anonymized behavior image field into the cross-border electric commercial user big data can be improved, the false anonymization processing of the group image characteristics in the second cross-border electric commercial user big data is avoided, and the usability of the second cross-border electric commercial user big data in the subsequent use process is improved.
Under some possible embodiments, the determining, from the first e-commerce activity description, a first user behavioral representation field comprises: detecting the interactive behavior change information of the first e-commerce business activity description to obtain first behavior change content; performing description detail expansion on the first electric business activity description to obtain a first expanded activity description; composing the first behavioral change content, the first expanded activity description, and a fourth electrical business activity description associated with a size presence of the first electrical business activity description into the first user behavioral representation field.
Under some possible embodiments, the determining a second user behavioral representation field from the second electrical business activity description, a third electrical business activity description associated with a size of the second electrical business activity description, and a user behavioral representation field that is consistent with the size of the second electrical business activity description, includes: detecting the change information of the interaction behavior of the second electric business activity description to obtain a user behavior portrait field with the same scale as the second electric business activity description; performing description detail expansion on the second electric business activity description to obtain a second expanded activity description; determining the second expanded activity description, the third electrical business activity description, and the user behavioral representation field that is consistent in size with the second electrical business activity description as the second user behavioral representation field.
Under some possible embodiments, the big data translation processing of the anonymized behavioral representation field of the first cross-border e-commerce user big data to obtain second cross-border e-commerce user big data includes: loading a first designated anonymization behavior representation field in the anonymization behavior representation field of the first cross-border power provider user big data into a field translation node of the second AI data processing thread to obtain a first translated behavior representation field; splicing the first translated behavior portrait field and the first appointed anonymization behavior portrait field to obtain a first fused behavior portrait field; loading the first fusion behavior portrait field into a field translation node of the second AI data processing thread to obtain a second translated behavior portrait field; constructing a second fused behavioral representation field by using the second translated behavioral representation field and a next group of anonymized behavioral representation fields in the anonymized behavioral representation field of the first cross-border electric business user big data, wherein the next group of anonymized behavioral representation fields are vector fields which are related to the first designated anonymized behavioral representation field and have larger dimensionality than the first designated anonymized behavioral representation field in the anonymized behavioral representation field of the first cross-border electric business user big data; and loading the third fused behavior image field into a field translation node of the second AI data processing thread until a third fused behavior image field corresponding to a second designated anonymous behavior image field in the anonymous behavior image field of the first cross-border electric business user big data is obtained, and obtaining the second cross-border electric business user big data.
Under some possible embodiments, the first AI data processing thread is an LSTM model.
By the design, the artificial intelligence server obtains the algorithm variable of the LSTM through the change information of the interaction behavior of at least two groups of cross-border E-commerce user big data with continuous time sequence relation, and adds the LSTM into the portrait mining node group of the second AI data processing thread, so that the individual privacy protection processing precision of the cross-border E-commerce user big data can be improved.
In some possible embodiments, before determining the information about the change of the interaction behavior of not less than two groups of cross-border electric business user big data with continuous time sequence relation, the method further comprises: determining a first cross-border electric company user big data template set, wherein the first cross-border electric company user big data template set comprises a non-anonymous user big data template and an anonymous user big data template corresponding to the non-anonymous user big data template; loading the big data template of the non-anonymous user into the second AI data processing thread to obtain a big data template of privacy protection corresponding to the big data template of the non-anonymous user; and determining second thread quality evaluation of the second AI data processing thread through the privacy protection big data template and the anonymous user big data template, and improving the second AI data processing thread through the second thread quality evaluation to obtain the debugged second AI data processing thread.
By the design, the second AI data processing thread is debugged based on the big data template of the non-anonymous user and the big data template of the anonymous user, the individual privacy protection processing of the big data of the cross-border e-commerce user is performed based on the debugged second AI data processing thread, and the accuracy of the thread debugging and the individual privacy protection processing can be guaranteed.
In some possible embodiments, before determining the information about the change of the interaction behavior of not less than two groups of cross-border electric business user big data with continuous time sequence relation, the method further comprises: determining at least two groups of user big data template sets with continuous time sequence relation, wherein the at least two groups of user big data template sets with continuous time sequence relation comprise at least two groups of associated non-anonymous user big data templates and at least two groups of associated non-anonymous user big data templates corresponding to the at least two groups of associated non-anonymous user big data templates; loading the at least two groups of associated non-anonymous user big data templates into the user behavior change analysis thread and the second AI data processing thread for processing to obtain at least two groups of associated privacy protection big data templates corresponding to the at least two groups of associated non-anonymous user big data templates; determining first thread quality evaluation of the user behavior change analysis thread and the second AI data processing thread through the at least two groups of associated privacy protection big data templates and the at least two groups of associated anonymous user big data templates, and improving the user behavior change analysis thread and the second AI data processing thread through the first thread quality evaluation to obtain the debugged user behavior change analysis thread and the second AI data processing thread.
By the design, the user behavior change analysis thread and the second AI data processing thread are simultaneously debugged based on not less than two groups of associated non-anonymous user big data templates and not less than two groups of associated anonymous user big data templates, the algorithm variable of the first AI data processing thread is determined based on the debugged user behavior change analysis thread, and the individual privacy protection processing of the cross-border electric appliance user big data is performed based on the debugged second AI data processing thread, so that the thread debugging efficiency can be improved, and the individual privacy protection processing of the cross-border electric appliance user big data can be accurately and reliably realized.
A second aspect is an artificial intelligence server comprising a memory and a processor; the memory and the processor are coupled; the memory for storing computer program code, the computer program code comprising computer instructions; wherein the computer instructions, when executed by the processor, cause the artificial intelligence server to perform the method of the first aspect.
A third aspect is a computer-readable storage medium having stored thereon a computer program which, when executed, performs the method of the first aspect.
According to one embodiment of the invention, algorithm variables of the first AI data processing thread are obtained through the interactive behavior change information of at least two groups of cross-border electric business user big data with continuous time sequence relation, at least one first AI data processing thread is positioned in an image mining node group of the second AI data processing thread, when the individual privacy protection processing is performed on one group of first cross-border electric company user big data in the cross-border electric company user big data based on the second AI data processing thread, the behavior change content between different groups of cross-border electric company user big data is concerned, thereby ensuring the accuracy and reliability of positioning the individual privacy images, reducing the probability of misjudging the group privacy images into individual privacy images, the pertinence of individual privacy protection processing on the cross-border electric business user big data is further improved, and the maximum availability of the cross-border electric business user big data after anonymous processing is guaranteed.
Drawings
Fig. 1 is a schematic flow chart of an e-commerce data protection method using artificial intelligence according to an embodiment of the present invention.
Fig. 2 is a block diagram of an apparatus for protecting e-commerce data using artificial intelligence according to an embodiment of the present invention.
Detailed Description
In the following, the terms "first", "second" and "third", etc. are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first," "second," or "third," etc., may explicitly or implicitly include one or more of that feature.
Fig. 1 is a schematic flowchart illustrating an e-commerce data protection method using artificial intelligence according to an embodiment of the present invention, where the e-commerce data protection method using artificial intelligence may be implemented by an artificial intelligence server, and the artificial intelligence server may include a memory and a processor; the memory and the processor are coupled; the memory for storing computer program code, the computer program code comprising computer instructions; wherein the computer instructions, when executed by the processor, cause the artificial intelligence server to perform the aspects described in the following steps.
Step 101: the artificial intelligence server determines at least two groups of interaction behavior change information of the cross-border electric business user big data with continuous time sequence relation.
It can be understood that the e-commerce data protection method applying artificial intelligence, which is shown in the embodiment of the invention, is suitable for the application field of the artificial intelligence server in individual privacy protection processing of cross-border e-commerce projects in different states.
In the embodiment of the invention, the real-time cross-border electricity commercial user big data collected in the cross-border electricity commercial projects in different states is composed of a plurality of groups of cross-border electricity commercial user big data with continuous time sequence relations, the real-time cross-border electricity commercial user big data (such as dynamic and streaming cross-border electricity commercial user big data) is loaded into the artificial intelligence server, the artificial intelligence server determines at least two groups of cross-border electricity commercial user big data with continuous time sequence relations from the real-time cross-border electricity commercial user big data, and the cross-border electricity commercial user big data with continuous time sequence relations can be understood as the adjacent groups of cross-border electricity commercial user big data with continuous time sequence relations.
In some demonstrative embodiments, the artificial intelligence server may determine, one by one, two sets of cross-border e-commerce user big data having a continuous timing relationship from the real-time cross-border e-commerce user big data.
In some embodiments of the possibilities, before the artificial intelligence server determines that no less than two groups of cross-border electric commercial user big data with continuous time sequence relationship exist, the artificial intelligence server may perform preprocessing on the determined cross-border electric commercial user big data to be subjected to individual privacy protection processing, and then obtain no less than two groups of cross-border electric commercial user big data with continuous time sequence relationship, where the preprocessing may include processing such as adjusting to a data format adapted to the artificial intelligence server.
Further, the cross-border e-commerce user big data can be understood as some columns of data information generated by the cross-border e-commerce client when interacting with the cross-border e-commerce business. Correspondingly, the interactive behavior change information can be understood as interactive operation change and/or operation habit change and the like of the corresponding user in the service interaction process, and is used for positioning and analyzing the individual user portrait.
For an independently implementable solution, the interactive behavior change information includes at least one scale of user behavior representation fields (such as dynamic user operation characteristics) between at least two sets of cross-border e-commerce user big data in a continuous time sequence relationship. Based on this, the process of the artificial intelligence server determining the interactive behavior variation information of not less than two groups of cross-border electric business user big data with continuous time sequence relation can comprise the following recorded contents of steps 1011-1013.
Step 1011: and the artificial intelligence server performs big data mining on at least two groups of big data of cross-border electric business users with continuous time sequence relation to obtain a group of electric business activity descriptions.
Further, the group of electric business activity descriptions comprises electric business activity descriptions of several scales. In addition, the description of the electric business activity can be understood as electric business activity features, such as feature vectors or feature maps, obtained by performing feature extraction on at least two groups of cross-border electric business user big data with continuous time sequence relations by the artificial intelligence server.
Step 1012: based on the description of the electrical business activity being the first electrical business activity description of the lowest scale in the set of electrical business activity descriptions, the artificial intelligence server determines a first user behavioral representation field from the first electrical business activity description.
In the implementation of the present invention, the scale of the description of the electric business activity may be understood as the size of the description of the electric business activity, such as two-dimensional size 3 × 3, 5 × 5 or 10 × 10, etc., and may also be three-dimensional size, etc., and is not limited herein.
Step 1013: on the basis that the electric business activity description is not the second electric business activity description with the lowest scale in the group of electric business activity descriptions, the artificial intelligence server determines a second user behavior representation field through the second electric business activity description, a third electric business activity description associated with the scale of the second electric business activity description and a user behavior representation field with the scale consistent with the second electric business activity description,
further, the interactive behavior change information includes a first user behavior representation field and a second user behavior representation field.
It can be understood that the artificial intelligence server performs big data mining on at least two groups of cross-border electric business user big data with continuous time sequence relation, the obtained group of electric business activity description comprises a plurality of vector fields (wherein, the vector fields can be understood as feature vectors) with different sizes, and the number of the vector fields of each size is not less than one. For example, a set of electric business activity descriptions includes 12 vector fields of different sizes.
For a separately implementable solution, the artificial intelligence server determining the first user behavior portrayal field by the first e-commerce activity description may include steps 10121-10123 as follows.
Step 10121: and the artificial intelligence server detects the interactive behavior change information of the first e-commerce business activity description to obtain the first behavior change content.
In the embodiment of the present invention, the content of the first behavior change may be understood as a result of predicting the interaction behavior change of the first electric business activity description, including but not limited to a user touch operation behavior change, a voice operation behavior change, a text operation behavior change, and the like.
Step 10122: and the artificial intelligence server expands the description details of the first electric business activity description to obtain a first expanded activity description.
In the embodiment of the present invention, the description details of the first electric business activity description may be expanded, for example, by performing upsampling processing on the first electric business activity description.
Step 10123: the artificial intelligence server constructs the first behavioral change content, the first extended activity description, and a fourth electrical business activity description associated with a size presence of the first electrical business activity description into a first user behavioral representation field.
For an independently implementable solution, the artificial intelligence server determines the second user behavioral representation field by the second electric business activity description, a third electric business activity description associated with the scale of the second electric business activity description, and the user behavioral representation field with the scale consistent with the second electric business activity description, including step 10131-step 10133.
Step 10131: and the artificial intelligence server detects the interactive behavior change information of the second electric business activity description to obtain a user behavior portrait field with the same scale as the second electric business activity description.
Step 10132: and the artificial intelligence server expands the description details of the second electric business activity description to obtain a second expanded activity description.
Step 10133: the artificial intelligence server determines the second expanded activity description, the third electrical business activity description and the user behavior representation field with the size consistent with the second electrical business activity description as a second user behavior representation field.
In some possible embodiments, the artificial intelligence server determines, through the user behavior change analysis thread, the interaction behavior change information of no less than two groups of cross-border power grid user big data with continuous time sequence relationship.
It can be understood that the artificial intelligence server loads at least two sets of cross-border electricity business user big data with continuous time sequence relation to the user behavior change analysis thread to obtain at least two sets of interaction behavior change information of the cross-border electricity business user big data with continuous time sequence relation, and the user behavior change analysis thread may be, for example, a user behavior analysis neural network, such as different types of neural networks.
In some illustrative embodiments, the artificial intelligence server determines two groups of cross-border electricity commercial user big data with continuous time sequence relationship in the real-time cross-border electricity commercial user big data one by one, loads the two groups of cross-border electricity commercial user big data with continuous time sequence relationship into the user behavior change analysis thread, and determines the interaction behavior change information of the two groups of cross-border electricity commercial user big data with continuous time sequence relationship.
In some possible embodiments, the user behavior change analysis thread includes a user behavior activity mining node and a user behavior field translation node, wherein the user behavior activity mining node is used for determining at least two sets of cross-border electric business user big data with continuous time sequence relation, and performing big data mining on at least two sets of cross-border electric business user big data with continuous time sequence relation to obtain at least one set of electric business activity description of the at least two sets of cross-border electric business user big data with continuous time sequence relation; and the user behavior field translation node is used for determining at least two groups of interaction behavior change information of the cross-border electric business user big data with continuous time sequence relation.
Step 102: the artificial intelligence server determines the algorithm variable of the first AI data processing thread through at least two groups of interactive behavior change information of cross-border electric business user big data with continuous time sequence relation.
In the embodiment of the invention, after the artificial intelligence server determines the interactive behavior change information of at least two groups of cross-border electric business user big data with continuous time sequence relation, the artificial intelligence server determines the algorithm variable of the first AI data processing thread through the interactive behavior change information of at least two groups of cross-border electric business user big data with continuous time sequence relation. The AI data processing threads may be, for example, AI neural networks, and further, the algorithm variables of the first AI data processing thread may be understood as the weight of the first AI data processing thread.
In some embodiments of the possibilities, the artificial intelligence server obtains the algorithm variables of the first AI data processing thread by vector processing the second electrical business activity description, a third electrical business activity description associated with a size of the second electrical business activity description, and the user behavior representation field of a size consistent with the second electrical business activity description.
In some possible embodiments, the first AI data processing thread is a long short term memory neural network (LSTM) or the like, which is not limited by the embodiments of the present invention.
Therefore, the artificial intelligence server obtains the algorithm variable of the LSTM through the interactive behavior change information of at least two sets of cross-border E-commerce user big data with continuous time sequence relation, and adds the LSTM to the portrait mining node group (such as a feature extraction layer) of the second AI data processing thread, so that the individual privacy protection processing precision of the cross-border E-commerce user big data can be improved.
In some possible embodiments, the artificial intelligence server deploys the first AI data processing thread configured with the corresponding algorithm variable to the image mining node group of the second AI data processing thread, so that the efficiency of anonymous processing of big data of cross-border e-commerce users can be improved.
Step 103: and the artificial intelligence server carries out individual privacy protection processing on the first cross-border electric commercial user big data in at least two groups of cross-border electric commercial user big data with continuous time sequence relation based on the second AI data processing thread to obtain second cross-border electric commercial user big data corresponding to the first cross-border electric commercial user big data.
In the embodiment of the invention, the first cross-border electricity business user big data is one of at least two sets of cross-border electricity business user big data with continuous time sequence relation, the data anonymization degree of the second cross-border electricity business user big data is larger than that of the first cross-border electricity business user big data, and the image mining node group of the second AI data processing thread comprises at least one first AI data processing thread. Furthermore, the individual privacy protection processing can be realized based on a K anonymization technology, for example, related feature codes are replaced or hidden, so that anonymization processing of individual portrait features is realized, and it can be further ensured that the second cross-border E-commerce user big data after anonymization processing can reflect the group portrait to a certain extent, so that analysis processing of subsequent related group portrait is not influenced to a certain extent.
It can be understood that after the artificial intelligence server determines the algorithm variable of the first AI data processing thread through the interactive behavior change information of at least two sets of cross-border electric business user big data with continuous time sequence relation, the artificial intelligence server performs individual privacy protection processing on the first cross-border electric business user big data in at least two sets of cross-border electric business user big data with continuous time sequence relation based on the second AI data processing thread to obtain second cross-border electric business user big data corresponding to the first cross-border electric business user big data, wherein the portrait mining node group of the second AI data processing thread comprises at least one first AI data processing thread.
In some possible embodiments, the artificial intelligence server performs anonymization processing on the first cross-border electric business user big data through not less than one first AI data processing thread to obtain an anonymized behavior portrait field of the first cross-border electric business user big data; and then, the artificial intelligence server carries out big data translation processing on the anonymized behavior portrait field of the big data of the first cross-border e-commerce user to obtain the big data of the second cross-border e-commerce user.
For an independently implementable technical solution, the artificial intelligence server performs big data translation processing on the anonymized behavioral representation field of the big data of the first cross-border e-commerce user to obtain the big data of the second cross-border e-commerce user, and the technical solution recorded in the step 201 to the step 205 may be included.
Step 201: and the artificial intelligence server loads a first designated anonymization behavior portrait field in the anonymization behavior portrait field of the first cross-border electric business user big data into a field translation node of a second AI data processing thread to obtain a first translated behavior portrait field.
In an embodiment of the present invention, the first designated anonymized behavioral representation field may be understood as a minimum encoding characteristic. The field translation node of the second AI data processing thread may be understood as a decoder of the second AI data processing thread. The translated behavioral representation field may be understood as a behavioral representation feature translated from the first specified anonymized behavioral representation field by a field translation node of the second AI data processing thread. In some examples, portrait mining/extraction may be understood as feature encoding and portrait translation/transcoding may be understood as feature transcoding.
Step 202, the first translated behavioral portrait field and the first designated anonymized behavioral portrait field are spliced to obtain a first fused behavioral portrait field.
Step 203, the artificial intelligence server inputs the first fusion behavior portrait field into a field translation node of the second AI data processing thread to obtain a second translated behavior portrait field.
Step 204, the second translated behavior image field and the next group of anonymized behavior image fields in the anonymized behavior image field of the first cross-border electric commerce user big data form a second fused behavior image field.
In an embodiment of the invention, the next set of anonymized behavioral representation fields is a vector field associated with the first designated anonymized behavioral representation field and having a dimension greater than the first designated anonymized behavioral representation field in the anonymized behavioral representation fields of the first cross-border e-commerce user big data.
And step 205, when the artificial intelligence server obtains a third fusion behavior image field corresponding to a second designated anonymization behavior image field in the anonymization behavior image field of the first cross-border e-commerce user big data, inputting the third fusion behavior image field into a field translation node of a second AI data processing thread by the artificial intelligence server, and obtaining second cross-border e-commerce user big data. The artificial intelligence server can carry out individual privacy protection processing on at least two groups of cross-border E-commerce user big data with continuous time sequence relation one by one so as to complete the process of implementing the individual privacy protection processing of the cross-border E-commerce user big data on cross-border E-commerce projects in different states.
For an independently implementable technical solution, an embodiment of the present invention shows an idea of an artificial intelligence server performing individual privacy protection processing on cross-border e-commerce user big data in cross-border e-commerce projects of different states, the artificial intelligence server including a user behavior change analysis thread and a second AI data processing thread, wherein the user behavior change analysis thread includes a behavior portrait mining node group and a behavior field translation node group, the behavior portrait mining node group is composed of a set of feature extraction units, the behavior field translation node group includes a description detail extension unit, an interaction behavior change information detection unit and an LSTM algorithm variable mining unit, the second AI data processing thread includes a first privacy portrait processing node group and a second privacy portrait processing node group, the first privacy processing node group includes at least one set of scale feature extraction sub-threads, each scale feature extraction sub-thread comprises a feature extraction unit and an LSTM, and the second privacy portrait node processing node group is composed of the feature extraction unit.
It can be understood that, in actual operation, the non-anonymous user big data1 and the non-anonymous user big data2 are associated with two groups of non-anonymous user big data, the non-anonymous user big data1 and the non-anonymous user big data2 are loaded to the behavior sketch mining node group, the non-anonymous user big data1 and the non-anonymous user big data2 are adjusted to a group of multi-modal electric business Activity descriptions based on a group of feature extraction units, the group of multi-modal electric business Activity descriptions are sorted in ascending order according to scale, then, according to the concept of ascending scale, a group of interaction behavior change information of the multi-modal electric business Activity descriptions is determined for the field translation node group one by one, in actual implementation, the lowest-scale electric business Activity description _1 in the group of multi-modal electric business Activity descriptions is first imported into the interaction behavior change information detection unit to obtain behavior change content _1, importing the electric business Activity description Activity _1 into a description detail expansion unit, obtaining an expanded Activity description Extended Activity description _1, and forming a user behavior image field _1 by the behavior change content _1, the expanded Activity description Extended Activity description _1 and the electric business Activity description Activity _2 associated with the size of the electric business Activity description Activity _ 1.
Then, the electric business Activity description Activity _2 is imported into the interactive behavior change information detecting unit, the behavior change content _2 of which the size is consistent with that of the electric business Activity description Activity _2 is obtained, the electric business Activity description Activity _2 is imported into the description detail expanding unit, the expanded Activity description Extended Activity description _2 is obtained, the behavior change content _2, the expanded Activity description Extended Activity description _2 and the user behavior representation field _1 associated with the size of the electric business Activity description Activity _2 are formed into the user behavior representation field _2, and the user behavior representation field _2 is calculated one by one until a set of multimodal electric business Activity descriptions are all imported into the behavior field translation node group, and a set of user behavior representation fields of anonymous user large data1 and anonymous user large data2 is obtained. And then, importing a group of user behavior image fields into an LSTM algorithm variable mining unit to obtain at least one LSTM algorithm variable corresponding to at least one LSTM, wherein one LSTM corresponds to one LSTM algorithm variable.
It is understood that after obtaining at least one LSTM algorithm variable corresponding to at least one LSTM, at least one LSTM algorithm variable may be input to at least one LSTM corresponding to the first privacy portrait processing node group, the non-anonymous user big data1 may be loaded to at least one set of scale feature extraction sub-threads, a dimension description may be obtained in each set of scale feature extraction sub-threads based on a feature extraction unit and one LSTM in sequence until obtaining a set of anonymized behavior portrait fields corresponding to the non-anonymized user big data1 that have been subject to privacy protection, the set of anonymized behavior portrait fields that have been subject to privacy protection being sorted in an ascending order, and then, the lowest anonymized behavior portrait field/feature _ a _ 1/feature extraction single that has been subject to privacy protection being loaded to the second privacy portrait processing node group of anonymized behavior portrait fields that have been subject to privacy protection being processed In the element, translated behavior portrait field/feature _ B _1/, the translated behavior portrait field/feature _ B _1/, and anonymized behavior portrait field/feature _ A _ 2/having completed individual privacy protection processing form a fused behavior portrait field/feature _ C _1/, and the fused behavior portrait field/feature _ C _ 1/are input into a feature extraction unit of a second privacy portrait node processing node group, the translated behavior portrait field/feature _ B _2/, are obtained, are calculated one by one until the fused behavior portrait field corresponding to the anonymized behavior field having completed individual privacy protection processing in the anonymized behavior portrait field having completed individual privacy protection processing is the largest, and are loaded into a feature extraction unit of a second privacy node processing group, and obtaining the anonymized user big data/target _ data _1/, which corresponds to the non-anonymized user big data.
In some possible embodiments, the AI data processing thread may be debugged before applying the AI data processing thread.
For an independently implementable solution, the process of debugging the second AI data processing thread by the artificial intelligence server may include the following recorded contents of steps 301 to 303.
Step 301: the artificial intelligence server determines a first cross-border utility user big data template set.
In the embodiment of the invention, the first cross-border e-commerce user big data template set comprises a non-anonymous user big data template and an anonymous user big data template corresponding to the non-anonymous user big data template.
Step 302: and the artificial intelligence server loads the big data template of the non-anonymous user into the second AI data processing thread to obtain the privacy protection big data template corresponding to the big data template of the non-anonymous user.
Step 303: and the artificial intelligence server determines second thread quality evaluation of the second AI data processing thread through the privacy protection big data template and the anonymous user big data template, improves the second AI data processing thread through the second thread quality evaluation, and obtains the debugged second AI data processing thread.
In the embodiment of the present invention, the thread quality evaluation may be understood as a thread loss variable of the AI data processing thread. For example, during debugging, an undeniated user big data template in a first cross-border power provider user big data template set is randomly selected, the undeniated user big data template is loaded to an image mining node group of a second AI data processing thread, a group of anonymization behavior image fields of the undeniated user big data template is obtained sequentially on the basis of at least one feature extraction unit and at least one LSTM model, then big data translation processing is carried out on the group of anonymization behavior image fields, a privacy protection big data template corresponding to the undeniated user big data template is obtained, and the privacy protection big data template and the anonymized user big data template are loaded to corresponding algorithms to determine a second cost function, for example, the second cost function can be realized by using a cross entropy algorithm or a hinge loss algorithm.
For an independently implementable technical solution, the process of debugging the user behavior change analysis thread and the second AI data processing thread by the artificial intelligence server may include the following steps 401 to 403.
Step 401: the artificial intelligence server determines that at least two groups of user big data template sets with continuous time sequence relation exist.
In the embodiment of the invention, the at least two groups of user big data templates with continuous time sequence relations comprise at least two groups of associated non-anonymous user big data templates and at least two groups of associated non-anonymous user big data templates corresponding to the at least two groups of associated non-anonymous user big data templates.
Step 402: and the artificial intelligence server loads at least two groups of associated non-anonymous user big data templates to the user behavior change analysis thread and the second AI data processing thread for processing to obtain at least two groups of associated privacy protection big data templates corresponding to the at least two groups of associated non-anonymous user big data templates.
Step 403: the artificial intelligence server determines first thread quality evaluation of a user behavior change analysis thread and a second AI data processing thread through at least two groups of associated privacy protection big data templates and at least two groups of associated anonymous user big data templates, improves the user behavior change analysis thread and the second AI data processing thread through the first thread quality evaluation, and obtains the debugged user behavior change analysis thread and the second AI data processing thread. And the artificial intelligent server carries out individual privacy protection processing on the cross-border E-commerce user big data of the cross-border E-commerce project in different states based on the debugged user behavior change analysis thread and the second AI data processing thread.
For example, during debugging, randomly selecting two groups of first non-anonymous user big data and second non-anonymous user big data which are associated in a first cross-border power grid user big data template set, loading the first non-anonymous user big data and the second non-anonymous user big data into an interaction behavior change information activity mining node, respectively performing windowing processing on the first non-anonymous user big data and the second non-anonymous user big data to obtain Q scale power grid business activity descriptions corresponding to the first non-anonymous user big data and Q scale power grid business activity descriptions corresponding to the second non-anonymous user big data, translating the relevant activity description behavior fields into a node group, obtaining interaction behavior change information of the first non-anonymous user big data and the second non-anonymous user big data based on a description detail expansion unit and an interaction behavior change information detection unit, and loading the interaction behavior change information of the first non-anonymous user big data and the second non-anonymous user big data to an LSTM algorithm variable mining unit to obtain at least one LSTM algorithm variable in the portrait mining node group of the second AI data processing thread.
Further, a first non-anonymous user big data template is loaded to an image mining node group of a second AI data processing thread, a group of anonymized behavior image fields of the first non-anonymous user big data template is obtained sequentially based on at least one feature extraction unit and at least one LSTM, then big data translation processing is carried out on the group of anonymized behavior image fields, a first privacy protection big data template corresponding to the first non-anonymous user big data template is obtained, and a first cost function is determined based on the first privacy protection big data template and the first anonymized user big data template.
It can be understood that the first thread quality evaluation is determined not only based on the tested privacy protection big data template but also based on the interactive behavior change information, so that the interactive behavior change information can be defined through the idea of 0 tag.
In some possible embodiments, after obtaining the second cross-border e-commerce user big data corresponding to the first cross-border e-commerce user big data, the method may further include: responding to a group portrait analysis request aiming at second cross-border e-commerce user big data, and performing group portrait mining on the second cross-border e-commerce user big data to obtain a group demand knowledge set of the second cross-border e-commerce user big data; determining a target service upgrading strategy by utilizing the group demand knowledge set; and performing cross-border electricity business upgrading based on the target business upgrading strategy.
In the embodiment of the invention, the group portrait analysis request can be sent by the E-commerce platform server and is used for requesting the artificial intelligence server to determine the corresponding group portrait, and based on the group portrait analysis request, the E-commerce platform server can obtain a target service upgrading strategy issued by the artificial intelligence server, so that upgrading optimization of corresponding cross-border E-commerce services is realized to conduct service drainage and user retention processing.
In some possible embodiments, performing group portrait mining on the second cross-border electric utility consumer big data to obtain a group demand knowledge set of the second cross-border electric utility consumer big data may include the following contents: determining at least one sample group requirement knowledge based on the second cross-border utility user big data; determining a global demand description value of a corresponding demand knowledge unit in each sample group demand knowledge to obtain target group demand knowledge; carrying out requirement text mining on the required knowledge of each sample group to obtain requirement text data corresponding to the required knowledge of the sample group; determining global text characteristics of the required text data corresponding to the required knowledge of each sample group to obtain reference data; the requirement text mining is carried out on the requirement knowledge of the target group to obtain requirement text data of big data of a second cross-border electric company user; optimizing the target group demand knowledge based on the quantitative comparison result of the demand text data of the second cross-border power provider user big data and the reference data to obtain a group demand knowledge set. Therefore, the group demand knowledge is optimized through the sample group demand knowledge and the demand text data, the precision and the credibility of the group demand knowledge set can be guaranteed, the group demand knowledge set is prevented from being lost to a certain extent, and meanwhile transitional mining aiming at individual demands can be avoided as far as possible, so that the privacy of users is protected.
In some possible embodiments, the performing demand text mining on each sample group required knowledge to obtain demand text data corresponding to the sample group required knowledge includes: inputting the required knowledge of each sample group into a required text mining algorithm to perform required text mining processing, so as to obtain required text data corresponding to the required knowledge of the sample group; the step of mining the requirement text of the target group requirement knowledge to obtain requirement text data of second cross-border power company user big data comprises the following steps: inputting the target group demand knowledge into the demand text mining algorithm to perform demand text mining processing to obtain demand text data of the second cross-border power company user big data; optimizing the target group demand knowledge based on the quantitative comparison result of the demand text data of the second cross-border power provider user big data and the reference data to obtain a group demand knowledge set, wherein the group demand knowledge set comprises: determining a cost function according to a quantitative comparison result of the demand text data of the second cross-border power provider user big data and the reference data; and reversely adjusting the variables of the target group demand knowledge based on the cost function until a preset reverse adjustment requirement is met.
In some possible embodiments, the inversely adjusting the variables of the target group demand knowledge based on the cost function until a preset inversely adjusting requirement is reached includes: improving the cost function based on the demand text mining algorithm to obtain an updated parameter of the demand knowledge of the target group; adjusting variables of the target group demand knowledge based on the updated parameters; if the reverse adjustment requirement is not met, inputting the adjusted target group requirement knowledge into the requirement text mining algorithm again for requirement text mining, and adjusting the variable of the target group requirement knowledge secondarily based on the obtained requirement text data of the second cross-border electric commercial user big data.
In some possible embodiments, the method further comprises: determining an importance coefficient corresponding to the required knowledge of each sample group; the determining the global demand description value of the demand knowledge unit corresponding to each sample group demand knowledge to obtain the target group demand knowledge comprises: determining a requirement description value weighting global text feature of a requirement knowledge unit corresponding to each sample group requirement knowledge according to an importance coefficient corresponding to each sample group requirement knowledge to obtain the target group requirement knowledge; the determining global text features of the required text data corresponding to the required knowledge of each sample group to obtain reference data includes: and determining the weighted global text features of the required text data corresponding to the sample group required knowledge according to the importance coefficients corresponding to the sample group required knowledge to obtain the reference data.
In summary, the artificial intelligence server obtains the algorithm variable of the first AI data processing thread through the interactive behavior change information of at least two sets of cross-border electric commercial user big data with continuous time sequence relation, wherein, at least one first AI data processing thread is in the image mining node group of the second AI data processing thread, when the individual privacy protection processing is carried out on one set of the first cross-border electric commercial user big data in at least two sets of cross-border electric commercial user big data with continuous time sequence relation based on the second AI data processing thread, the behavior change content between different sets of cross-border electric commercial user big data is concerned, and then the precision and the credibility of the individual privacy image positioning can be ensured, thereby the probability of misjudging the group privacy image into the individual privacy can be reduced, and the pertinence of the individual privacy protection processing on the cross-border electric commercial user big data can be improved, ensuring maximum availability of cross-border e-commerce user big data after anonymous processing.
Based on the same inventive concept, fig. 2 shows a block diagram of an artificial intelligence application e-commerce data protection apparatus provided by an embodiment of the present invention, and the artificial intelligence application e-commerce data protection apparatus may include an information determination module 21 and a data anonymity module 22 for implementing the relevant method steps shown in fig. 1.
The related embodiment applied to the invention can achieve the following technical effects: the algorithm variable of the first AI data processing thread is obtained through the interactive behavior change information of at least two groups of cross-border electric commercial user big data with continuous time sequence relation, wherein at least one first AI data processing thread is positioned in the portrait mining node group of the second AI data processing thread, when the individual privacy protection processing is carried out on one group of first cross-border electric commercial user big data in at least two groups of cross-border electric commercial user big data with continuous time sequence relation based on the second AI data processing thread, the behavior change content between different groups of cross-border electric commercial user big data is concerned, and the accuracy and the reliability of individual privacy image positioning can be further ensured, so that the probability of misjudging the group privacy images into individual privacy images can be reduced, and the pertinence of the individual privacy protection processing on the cross-border electric commercial user big data can be further improved, ensuring maximum availability of cross-border e-commerce user big data after anonymous processing.
The foregoing is only illustrative of the present invention. Those skilled in the art can conceive of changes or substitutions based on the specific embodiments provided by the present invention, and all such changes or substitutions are intended to be included within the scope of the present invention.

Claims (10)

1. An e-commerce data protection method applying artificial intelligence is characterized by being applied to an artificial intelligence server, and the method at least comprises the following steps:
determining at least two groups of interaction behavior change information of cross-border electricity business user big data with continuous time sequence relation; determining an algorithm variable of a first AI data processing thread according to the at least two groups of interaction behavior change information of the cross-border electric power business user big data with continuous time sequence relation; performing individual privacy protection processing on first cross-border electric commercial user big data in at least two groups of cross-border electric commercial user big data with continuous time sequence relation based on a second AI data processing thread to obtain second cross-border electric commercial user big data corresponding to the first cross-border electric commercial user big data;
wherein: the first cross-border electricity commercial user big data is one of the at least two groups of cross-border electricity commercial user big data with continuous time sequence relation, the data anonymization degree of the second cross-border electricity commercial user big data is larger than that of the first cross-border electricity commercial user big data, and the portrait mining node group of the second AI data processing thread comprises at least one first AI data processing thread.
2. The method according to claim 1, wherein the interactive behavior change information covers at least one scale of user behavior representation fields between the at least two sets of cross-border e-commerce user big data with continuous time sequence relationship; the determination of the interactive behavior change information of at least two groups of cross-border power provider user big data with continuous time sequence relation comprises the following steps:
carrying out big data mining on the big data of at least two sets of cross-border electric business users with continuous time sequence relation to obtain a set of electric business activity descriptions, wherein the set of electric business activity descriptions comprises electric business activity descriptions of a plurality of scales;
determining a first user behavioral representation field from the first electrical business activity description based on the electrical business activity description being a first electrical business activity description of lowest scale in the set of electrical business activity descriptions;
determining a second user behavioral representation field by the second electrical business activity description, a third electrical business activity description associated with the scale of the second electrical business activity description, and a user behavioral representation field consistent with the scale of the second electrical business activity description, on the basis that the electrical business activity description is not the second electrical business activity description of the lowest scale in the set of electrical business activity descriptions, the interaction behavior change information encompassing the first user behavioral representation field and the second user behavioral representation field.
3. The method of claim 1, wherein the determining no less than two sets of cross-border utility consumer big data interaction behavior change information with continuous time sequence relationship comprises: and determining the interactive behavior change information of at least two groups of cross-border electricity business user big data with continuous time sequence relation through a user behavior change analysis thread.
4. The method according to claim 1, wherein the determining algorithm variables of the first AI data processing thread through the information about the change of the interaction behavior of the at least two sets of the cross-border electric business user big data with continuous time sequence relation comprises:
vector processing is carried out through a second electric business activity description, a third electric business activity description which is associated with the size of the second electric business activity description, and a user behavior portrait field which is consistent with the size of the second electric business activity description to obtain an algorithm variable of the first AI data processing thread.
5. The method according to claim 1, wherein the performing individual privacy protection processing on a first cross-border e-commerce user big data in the at least two sets of cross-border e-commerce user big data with continuous time sequence relation based on a second AI data processing thread to obtain a second cross-border e-commerce user big data corresponding to the first cross-border e-commerce user big data comprises:
anonymizing the first cross-border electric business user big data through at least one first AI data processing thread to obtain an anonymized behavior portrait field of the first cross-border electric business user big data;
and performing big data translation processing on the anonymized behavior portrait field of the big data of the first cross-border e-commerce user to obtain big data of a second cross-border e-commerce user.
6. The method of claim 2, wherein determining a first user behavioral representation field from the first e-commerce business activity description comprises:
detecting the interactive behavior change information of the first e-commerce business activity description to obtain first behavior change content;
performing description detail expansion on the first electric business activity description to obtain a first expanded activity description;
composing the first behavioral change content, the first expanded activity description, and a fourth electrical business activity description associated with a size presence of the first electrical business activity description into the first user behavioral representation field.
7. The method of claim 2, wherein determining a second user behavioral representation field from the second electrical business activity description, a third electrical business activity description associated with a size of the second electrical business activity description, and a user behavioral representation field that is consistent with the size of the second electrical business activity description comprises:
detecting the change information of the interaction behavior of the second electric business activity description to obtain a user behavior portrait field with the same scale as the second electric business activity description;
performing description detail expansion on the second electric business activity description to obtain a second expanded activity description;
determining the second expanded activity description, the third electrical business activity description, and the user behavioral representation field that is consistent in size with the second electrical business activity description as the second user behavioral representation field.
8. The method as claimed in claim 5, wherein the big data translation processing of the anonymized behavioral representation field of the big data of the first cross-border e-commerce user to obtain big data of a second cross-border e-commerce user comprises:
loading a first designated anonymization behavior representation field in the anonymization behavior representation field of the first cross-border power provider user big data into a field translation node of the second AI data processing thread to obtain a first translated behavior representation field;
splicing the first translated behavior portrait field and the first appointed anonymization behavior portrait field to obtain a first fused behavior portrait field;
loading the first fusion behavior portrait field into a field translation node of the second AI data processing thread to obtain a second translated behavior portrait field;
constructing a second fused behavioral representation field by using the second translated behavioral representation field and a next group of anonymized behavioral representation fields in the anonymized behavioral representation field of the first cross-border electric business user big data, wherein the next group of anonymized behavioral representation fields are vector fields which are related to the first designated anonymized behavioral representation field and have larger dimensionality than the first designated anonymized behavioral representation field in the anonymized behavioral representation field of the first cross-border electric business user big data;
and loading the third fused behavior image field into a field translation node of the second AI data processing thread until a third fused behavior image field corresponding to a second designated anonymous behavior image field in the anonymous behavior image field of the first cross-border electric business user big data is obtained, and obtaining the second cross-border electric business user big data.
9. The method of claim 1, wherein the first AI data processing thread is an LSTM model;
before determining the interaction behavior change information of at least two groups of cross-border electric business user big data with continuous time sequence relation, the method further comprises the following steps: determining a first cross-border electric company user big data template set, wherein the first cross-border electric company user big data template set comprises a non-anonymous user big data template and an anonymous user big data template corresponding to the non-anonymous user big data template; loading the big data template of the non-anonymous user into the second AI data processing thread to obtain a big data template of privacy protection corresponding to the big data template of the non-anonymous user; determining second thread quality evaluation of the second AI data processing thread through the privacy protection big data template and the anonymous user big data template, and improving the second AI data processing thread through the second thread quality evaluation to obtain the debugged second AI data processing thread;
before determining the interaction behavior change information of at least two groups of cross-border electric business user big data with continuous time sequence relation, the method further comprises the following steps: determining at least two groups of user big data template sets with continuous time sequence relation, wherein the at least two groups of user big data template sets with continuous time sequence relation comprise at least two groups of associated non-anonymous user big data templates and at least two groups of associated non-anonymous user big data templates corresponding to the at least two groups of associated non-anonymous user big data templates; loading the at least two groups of associated non-anonymous user big data templates into the user behavior change analysis thread and the second AI data processing thread for processing to obtain at least two groups of associated privacy protection big data templates corresponding to the at least two groups of associated non-anonymous user big data templates; determining first thread quality evaluation of the user behavior change analysis thread and the second AI data processing thread through the at least two groups of associated privacy protection big data templates and the at least two groups of associated anonymous user big data templates, and improving the user behavior change analysis thread and the second AI data processing thread through the first thread quality evaluation to obtain the debugged user behavior change analysis thread and the second AI data processing thread.
10. An artificial intelligence server, comprising: a memory and a processor; the memory and the processor are coupled; the memory for storing computer program code, the computer program code comprising computer instructions; wherein the computer instructions, when executed by the processor, cause the artificial intelligence server to perform the method of any of claims 1-9.
CN202210561265.3A 2022-05-23 2022-05-23 E-commerce data protection method and server applying artificial intelligence Active CN114880709B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210561265.3A CN114880709B (en) 2022-05-23 2022-05-23 E-commerce data protection method and server applying artificial intelligence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210561265.3A CN114880709B (en) 2022-05-23 2022-05-23 E-commerce data protection method and server applying artificial intelligence

Publications (2)

Publication Number Publication Date
CN114880709A true CN114880709A (en) 2022-08-09
CN114880709B CN114880709B (en) 2023-04-07

Family

ID=82678464

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210561265.3A Active CN114880709B (en) 2022-05-23 2022-05-23 E-commerce data protection method and server applying artificial intelligence

Country Status (1)

Country Link
CN (1) CN114880709B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115422593A (en) * 2022-09-14 2022-12-02 戴丽 Information optimization processing method and server based on Internet and digital technology

Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111723396A (en) * 2020-05-20 2020-09-29 华南理工大学 SaaS-based general cloud data privacy protection platform and method
CN111967521A (en) * 2020-08-18 2020-11-20 中国银行股份有限公司 Cross-border active user identification method and device
CN112052365A (en) * 2020-09-03 2020-12-08 中国银行股份有限公司 Cross-border scene portrait construction method and device
CN112818040A (en) * 2021-03-08 2021-05-18 裴炳坤 Big data combined user behavior analysis method and information processing server
CN112989065A (en) * 2021-03-23 2021-06-18 汪威 Information processing method and cloud computing platform applied to big data user portrait analysis
CN113094724A (en) * 2021-03-30 2021-07-09 看见未来科技发展(深圳)有限公司 Personal data management method and device
CN113468603A (en) * 2021-08-02 2021-10-01 东莞市慧学慧玩教育科技有限公司 Big data privacy security protection method and system based on artificial intelligence
CN113946611A (en) * 2021-10-20 2022-01-18 广州敏捷大数据科技有限公司 Information mining method based on AI network and intelligent operation server
CN114220548A (en) * 2021-12-13 2022-03-22 山东畅想大数据服务有限公司 Big data anonymous protection method and system serving digital medical treatment
CN114417397A (en) * 2021-12-16 2022-04-29 杭州薮猫科技有限公司 Behavior portrait construction method and device, storage medium and computer equipment
CN114417405A (en) * 2022-01-11 2022-04-29 山东泽钜大数据技术有限公司 Privacy service data analysis method based on artificial intelligence and server
CN114463086A (en) * 2021-12-29 2022-05-10 杨金光 E-commerce information security method combining big data and readable storage medium

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111723396A (en) * 2020-05-20 2020-09-29 华南理工大学 SaaS-based general cloud data privacy protection platform and method
CN111967521A (en) * 2020-08-18 2020-11-20 中国银行股份有限公司 Cross-border active user identification method and device
CN112052365A (en) * 2020-09-03 2020-12-08 中国银行股份有限公司 Cross-border scene portrait construction method and device
CN112818040A (en) * 2021-03-08 2021-05-18 裴炳坤 Big data combined user behavior analysis method and information processing server
CN112989065A (en) * 2021-03-23 2021-06-18 汪威 Information processing method and cloud computing platform applied to big data user portrait analysis
CN113094724A (en) * 2021-03-30 2021-07-09 看见未来科技发展(深圳)有限公司 Personal data management method and device
CN113468603A (en) * 2021-08-02 2021-10-01 东莞市慧学慧玩教育科技有限公司 Big data privacy security protection method and system based on artificial intelligence
CN113946611A (en) * 2021-10-20 2022-01-18 广州敏捷大数据科技有限公司 Information mining method based on AI network and intelligent operation server
CN114220548A (en) * 2021-12-13 2022-03-22 山东畅想大数据服务有限公司 Big data anonymous protection method and system serving digital medical treatment
CN114417397A (en) * 2021-12-16 2022-04-29 杭州薮猫科技有限公司 Behavior portrait construction method and device, storage medium and computer equipment
CN114463086A (en) * 2021-12-29 2022-05-10 杨金光 E-commerce information security method combining big data and readable storage medium
CN114417405A (en) * 2022-01-11 2022-04-29 山东泽钜大数据技术有限公司 Privacy service data analysis method based on artificial intelligence and server

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
李家华: "基于大数据的人工智能跨境电商导购平台信息个性化推荐算法", 《科学技术与工程》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115422593A (en) * 2022-09-14 2022-12-02 戴丽 Information optimization processing method and server based on Internet and digital technology

Also Published As

Publication number Publication date
CN114880709B (en) 2023-04-07

Similar Documents

Publication Publication Date Title
CN112000819B (en) Multimedia resource recommendation method and device, electronic equipment and storage medium
CN110909182B (en) Multimedia resource searching method, device, computer equipment and storage medium
CN109299327A (en) Video recommendation method, device, equipment and storage medium
CN112801215A (en) Image processing model search, image processing method, image processing apparatus, and storage medium
CN113379449B (en) Multimedia resource recall method and device, electronic equipment and storage medium
CN112668608B (en) Image recognition method and device, electronic equipment and storage medium
CN111259256B (en) Content processing method, content processing device, computer readable storage medium and computer equipment
CN111651668B (en) User portrait label generation method and device, storage medium and terminal
CN114880709B (en) E-commerce data protection method and server applying artificial intelligence
CN114780753A (en) Dialogue recommendation method, device and equipment based on knowledge graph and storage medium
CN114461906A (en) Sequence recommendation method and device focusing on user core interests
CN113641835A (en) Multimedia resource recommendation method and device, electronic equipment and medium
CN114549849A (en) Image recognition method and device, computer equipment and storage medium
CN111597361B (en) Multimedia data processing method, device, storage medium and equipment
CN117149996A (en) Man-machine interface digital conversation mining method and AI system for artificial intelligence application
CN112183303A (en) Transformer equipment image classification method and device, computer equipment and medium
CN116401522A (en) Financial service dynamic recommendation method and device
CN115756821A (en) Online task processing model training and task processing method and device
CN113516182B (en) Visual question-answering model training and visual question-answering method and device
CN115222112A (en) Behavior prediction method, behavior prediction model generation method and electronic equipment
CN114510627A (en) Object pushing method and device, electronic equipment and storage medium
CN112862002A (en) Training method of multi-scale target detection model, target detection method and device
JP2021152941A (en) Object recommendation method, neural network and training method thereof, device and medium
CN114048392B (en) Multimedia resource pushing method and device, electronic equipment and storage medium
Lee et al. An enhanced memory-based collaborative filtering approach for context-aware recommendation

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20221019

Address after: No. 012, Bohai 12 road, Bincheng District, Binzhou City, Shandong Province 256600

Applicant after: Cai Zhongzhu

Address before: 554300 front of No. 28, Yiwu Commodity City, Wanshan District, Tongren City, Guizhou Province

Applicant before: Tongren Yingdan Network Technology Co.,Ltd.

TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20230116

Address after: No. 1, Lane 200, Longcao Road, Xuhui District, Shanghai, 200000

Applicant after: Shanghai Yanqi Huawei Information System Technology Co.,Ltd.

Address before: No. 012, Bohai 12 road, Bincheng District, Binzhou City, Shandong Province 256600

Applicant before: Cai Zhongzhu

GR01 Patent grant
GR01 Patent grant