CN111988129B - Influence maximization data set processing method, device and system - Google Patents
Influence maximization data set processing method, device and system Download PDFInfo
- Publication number
- CN111988129B CN111988129B CN201910425628.9A CN201910425628A CN111988129B CN 111988129 B CN111988129 B CN 111988129B CN 201910425628 A CN201910425628 A CN 201910425628A CN 111988129 B CN111988129 B CN 111988129B
- Authority
- CN
- China
- Prior art keywords
- node
- weight value
- influence
- nodes
- target
- 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.)
- Active
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 18
- 239000013598 vector Substances 0.000 claims abstract description 72
- 238000012545 processing Methods 0.000 claims abstract description 56
- 238000000034 method Methods 0.000 claims abstract description 40
- 230000008569 process Effects 0.000 claims abstract description 12
- 238000004590 computer program Methods 0.000 claims description 17
- 230000003993 interaction Effects 0.000 claims description 7
- 238000007418 data mining Methods 0.000 abstract description 5
- 238000010586 diagram Methods 0.000 description 27
- 238000004422 calculation algorithm Methods 0.000 description 12
- 230000006870 function Effects 0.000 description 7
- 239000011159 matrix material Substances 0.000 description 7
- 230000006855 networking Effects 0.000 description 7
- 238000011160 research Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000003612 virological effect Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000014509 gene expression Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/008—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols involving homomorphic encryption
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/958—Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
-
- 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
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/04—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks
- H04L63/0428—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks wherein the data content is protected, e.g. by encrypting or encapsulating the payload
- H04L63/0435—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks wherein the data content is protected, e.g. by encrypting or encapsulating the payload wherein the sending and receiving network entities apply symmetric encryption, i.e. same key used for encryption and decryption
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Bioethics (AREA)
- Computer Hardware Design (AREA)
- Computing Systems (AREA)
- Business, Economics & Management (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the disclosure discloses a processing method of an influence maximization data set, which is applied to a cloud service platform and comprises the following steps: receiving homomorphic encrypted social data to be analyzed, wherein the social data to be analyzed comprises nodes formed by preference vectors used for representing preference of users to set topics; determining weight values among different nodes based on the preference vectors, wherein the weight values are used for representing the mutual influence among the nodes; calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is determined to be greater than a set threshold value; and forming an influence maximization data set according to the target nodes. The embodiment of the disclosure further discloses an influence maximization data set processing device and system. The embodiment of the disclosure processes the encrypted data to be analyzed, realizes the determination of the influence maximization data set on the premise of ensuring the privacy of the data owner, and improves the value of the social data in the data mining field.
Description
Technical Field
The embodiment of the disclosure relates to the field of data processing, in particular to a method, a device and a system for processing an influence maximization data set.
Background
With the development and popularization of cloud computing, a cloud service platform is favored by more and more people and small and medium-sized enterprises. In the cloud service, a user uses the cloud service by selecting different payment modes provided by a cloud service platform, and the cloud service platform provides various types of services for the user by utilizing a high-performance server of the cloud service platform. Various social networking platforms based on these cloud computing social networking platforms have been widely used in our daily lives, and the social networking platforms constantly collect user data while allowing users to better enjoy the benefits of being connected with friends, thereby driving advanced functions using analysis results. According to research, useful information obtained by data analysis and extraction is far more valuable than data per se. Despite the rapid development of cloud services, users are reluctant to outsource sensitive data to these third party clouds due to privacy concerns. Most of the existing research works are focused on realizing specific functions, and the privacy problem of a data owner in a cloud environment is not fully solved. In their research, data owners outsourced data in clear text to third party cloud service platforms and allow the cloud to fully analyze and process their privacy sensitive data, including identity, gender, residence, health, and friends list, etc.
The impact maximization problem is a function based on social network data mining that aims to select from the data packets an impact maximization set S that has the greatest impact on the entire network through propagation. The technology attracts a great deal of research in the field of data mining, and is also widely applied to viral marketing, in which new products are promoted through relationships among users, gossip control and information monitoring. However, the problem of maximizing the execution influence on the cloud service platform cannot be solved well, and the privacy problem of the data owner in the cloud environment cannot be solved sufficiently.
Disclosure of Invention
In order to solve the existing technical problem, embodiments of the present disclosure provide a data set processing method, apparatus, and system capable of maximizing influence of data owner privacy.
In order to achieve the above purpose, the technical solution of the embodiment of the present disclosure is implemented as follows:
in a first aspect, an embodiment of the present disclosure provides an influence maximization data set processing method, applied to a cloud service platform, including:
receiving homomorphic encrypted social data to be analyzed, wherein the social data to be analyzed comprises nodes formed by preference vectors used for representing preference of users to set topics;
determining a weight value between different nodes based on the preference vector, wherein the weight value is used for representing the interaction force between the nodes; calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is determined to be larger than a set threshold value;
and forming an influence maximization data set according to the target nodes.
Wherein, the calculating a total inbound weight value of each node according to the weight value, and when determining that the total inbound weight value of a node is greater than a set threshold, taking the node as a target node includes:
receiving an influence maximization query request sent by a terminal, wherein the request carries influence target group information of a preselected target node and source group information of the influence node;
calculating a total inbound weight value corresponding to each pre-selected target node in the influence target group information and each influence node in the source group information based on the weight value, and taking the pre-selected target node as a target node when the total inbound weight value of the pre-selected target node is determined to be larger than a set threshold value.
Wherein the determining weight values between different nodes based on the preference vector comprises:
obtaining the preference vector of each node after homomorphic encryption;
and obtaining the weight value between different nodes by calculating the inner product between the preference vectors of different nodes.
Wherein the calculating a total inbound weight value for each node according to the weight value further comprises:
receiving selected information of a selected maximum influence node; adding a set weight value to the total inbound weight value of the corresponding node according to the selected information to obtain an updated total inbound weight value of the node; and setting a weight value as an outbound weight value corresponding to the selected maximum influence node.
Wherein, after obtaining the updated total inbound weight value of the node, the method further includes:
and when the total inbound weight value of the node is greater than the upper limit setting threshold value, updating the total inbound weight value to be the upper limit setting value, and obtaining the updated total inbound weight value of the node.
In a second aspect, an embodiment of the present disclosure provides an influence maximization data set processing method, applied to a social network platform, including:
obtaining social data to be analyzed; the social data to be analyzed comprises nodes formed by preference vectors used for representing preferences of users on set topics;
encrypting the social data to be analyzed through a homomorphic encryption algorithm and sending the encrypted social data to a cloud service platform;
and receiving target node information sent by the cloud service platform according to the received social data to be analyzed, decrypting the target node information, and forming an influence maximization data set according to the target node.
In a third aspect, an embodiment of the present disclosure provides an influence maximization data set processing method, applied to a terminal, including:
acquiring social data and sending the social data to a social network platform; the social data is used for forming social data to be analyzed, which is composed of nodes formed by preference vectors representing preferences of users on set topics after homomorphic encryption, by the social network platform;
receiving a target node which is sent by the social network platform and acquired from a cloud service platform; wherein the target node is determined by the cloud service platform by determining a weight value between different nodes based on the preference vector and calculating a relationship between a total inbound weight value of each node and a set threshold according to the weight value.
Here, after receiving the target node obtained from the cloud service platform and sent by the social network platform, the method further includes:
and sorting the target nodes by the total inbound weight values, and adding the target nodes with the total inbound weight values within a set weight value range to the influence maximization data set as the nodes with the maximum influence.
Here, the adding the target node having the inbound total weight value within a set weight value range to an influence maximization data set as a node having a maximum influence includes:
and after the total inbound weight values are sequenced, when the total inbound weight values comprise two target nodes with the same total inbound weight values in the target nodes in the weight value setting range, selecting the target nodes with relatively smaller index numbers as the nodes with the largest influence to be added into the influence maximization data set.
Here, the adding the target node having the inbound total weight value within a set weight value range to an influence maximization data set as a node having a maximum influence includes:
obtaining a selected instruction for a node with the greatest influence among the target nodes;
and when the number of the nodes with the maximum influence is determined not to be in a set range according to the selected instruction, returning to the step of receiving the target nodes which are sent by the social network platform and acquired from the cloud service platform until the number of the nodes with the maximum influence meets the set range.
In a fourth aspect, an embodiment of the present disclosure provides a cloud service platform, including a first receiving module and a first processing module, wherein,
the first receiving module is used for receiving homomorphic encrypted social data to be analyzed, wherein the social data to be analyzed comprises nodes formed by preference vectors used for representing preferences of users on set topics;
the first processing module is used for determining weight values among different nodes based on the preference vector, wherein the weight values are used for representing the mutual influence among the nodes; calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is determined to be larger than a set threshold value; and forming an influence maximization data set according to the target nodes.
In a fifth aspect, an embodiment of the present disclosure provides a social networking platform, including a first obtaining module, a second receiving module, and a second processing module, wherein,
the first acquisition module is used for acquiring social data to be analyzed; the social data to be analyzed comprises nodes formed by preference vectors used for representing preference of users on set topics;
the second receiving module is used for receiving target node information sent by the cloud service platform according to the received social data to be analyzed;
the second processing module is used for encrypting the social data to be analyzed through a homomorphic encryption algorithm and sending the encrypted social data to the cloud service platform; and decrypting the target node information, and forming an influence maximization data set according to the target node.
In a sixth aspect, embodiments of the present disclosure provide a terminal, including a second obtaining module and a third processing module, wherein,
the second acquisition module is used for acquiring social data and sending the social data to a social network platform; the social data is used for the social network platform to form social data to be analyzed, wherein the social data to be analyzed is composed of nodes formed by preference vectors representing preferences of users on set topics after homomorphic encryption;
the third processing module is configured to receive a target node obtained from a cloud service platform and sent by the social network platform, where the target node is determined by the cloud service platform by determining a weight value between different nodes based on the preference vector and calculating a relationship between a total inbound weight value of each node and a set threshold according to the weight value.
In a seventh aspect, an embodiment of the present disclosure provides an influence maximization data set processing apparatus, including: a processor and a memory for storing a computer program capable of running on the processor; wherein the processor is configured to implement the influence-maximization dataset processing method according to any one of the embodiments of the present disclosure when the computer program is executed.
In an eighth aspect, an embodiment of the present disclosure provides an influence maximization data set processing system, including a cloud service platform according to an embodiment of the present disclosure, a social network platform according to an embodiment of the present disclosure, and a terminal according to an embodiment of the present disclosure.
The embodiment of the disclosure provides a processing method, a processing device and a processing system for an influence maximization data set, wherein a cloud service platform receives homomorphic encrypted social data to be analyzed, wherein the social data to be analyzed comprises nodes formed by preference vectors for representing preference of a user to a set theme; here, the social data to be analyzed is homomorphic encrypted, so that the data security of the social data to be analyzed is ensured; determining a weight value between different nodes based on the preference vector, wherein the weight value is used for representing the interaction force between the nodes; calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is larger than a set threshold value; the encrypted data to be analyzed is processed, the determination of the influence maximization data set is achieved on the premise that the privacy of a data owner is guaranteed, and the value of the social data in the data mining field is improved.
Drawings
FIG. 1 is a schematic diagram of a restaurant marketing scenario provided by an embodiment of the present disclosure;
FIG. 2 is a flow chart illustrating a method for processing an influence maximization data set according to an embodiment of the present disclosure;
FIG. 3 is a schematic flow chart diagram illustrating a method for processing an influence maximization data set according to another embodiment of the present disclosure;
FIG. 4 is a schematic flow chart diagram illustrating a method for processing an influence maximization data set according to another embodiment of the present disclosure;
FIG. 5 is a schematic flow chart diagram illustrating a method for processing an influence maximization data set according to another embodiment of the present disclosure;
FIG. 6 is a schematic flow chart diagram illustrating a method for processing an influence-maximization data set according to another embodiment of the disclosure;
FIG. 7 is a schematic flowchart of a method for processing an influence-maximization data set according to another embodiment of the disclosure;
FIG. 8 is a flowchart illustrating a method for processing an influence maximization data set according to another embodiment of the present disclosure;
fig. 9 is a schematic structural diagram of a cloud service platform according to an embodiment of the present disclosure;
FIG. 10 is a schematic structural diagram of a social networking platform according to another embodiment of the present disclosure;
fig. 11 is a schematic structural diagram of a terminal according to another embodiment of the present disclosure;
fig. 12 is a schematic structural diagram of an influence maximization data set processing apparatus according to an embodiment of the present disclosure;
FIG. 13 is a node of an impact-maximization data-set processing apparatus according to another embodiment of the present disclosure
A schematic diagram;
fig. 14 is a schematic structural diagram of an influence maximization data set processing apparatus according to another embodiment of the present disclosure;
FIG. 15 is a schematic structural diagram of a data set processing system for maximizing influence according to another embodiment of the present disclosure;
FIG. 16 is a schematic diagram of a weight matrix of nodes in a social network according to an embodiment of the present disclosure;
fig. 17 is a flowchart of a method for processing an influence-maximization data set according to another embodiment of the disclosure;
FIG. 18a is a schematic diagram of a1 st iteration process according to an embodiment of the present disclosure;
FIG. 18b is a schematic diagram of a2 nd iteration process according to an embodiment of the present disclosure;
FIG. 18c is a schematic diagram of a3 rd iteration process according to an embodiment of the present disclosure;
fig. 19 is a schematic diagram of a selection process of an impact maximization set node according to an embodiment of the present disclosure.
Detailed Description
The present disclosure will be described in further detail below with reference to the accompanying drawings and examples. It is to be understood that the examples provided herein are merely illustrative of the present disclosure and are not intended to limit the present disclosure. In addition, the embodiments provided below are some embodiments for implementing the disclosure, not all embodiments for implementing the disclosure, and the technical solutions described in the embodiments of the disclosure may be implemented in any combination without conflict.
It should be noted that in the embodiments of the present disclosure, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a method or apparatus that comprises a list of elements does not include only the elements explicitly recited, but also includes other elements not explicitly listed or inherent to the method or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other elements of interest in a method or apparatus comprising the element (e.g., steps in a method or elements in an apparatus, such as units that may be part of a circuit, part of a processor, part of a program or software, etc.).
Before further detailed description of the embodiments of the present disclosure, terms and expressions referred to in the embodiments of the present disclosure are explained, and the terms and expressions referred to in the embodiments of the present disclosure are applied to the following explanations.
(1) Partial homomorphic encryption
Homomorphic Encryption (HE) is a symmetric Encryption algorithm proposed by IBM Craig Gentry in 2009. The purpose of homomorphic encryption is to find an encryption algorithm which can perform addition and multiplication operations on a ciphertext, so that the result obtained by performing certain operation on the encrypted ciphertext is exactly equal to the ciphertext obtained by performing expected operation on the plaintext before encryption and then encrypting the plaintext. The homomorphic encryption effectively ensures that a data processing party can directly carry out corresponding processing on the ciphertext of the data without knowing the plaintext information of the data processed by the data processing party. The characteristic of homomorphic encryption enables data and privacy of a user to be correspondingly secured, and therefore homomorphic encryption is applied to many real-world scenes to ensure the security of the data. For example, a bank needs to authorize a third-party data processing center to analyze transaction data without exposing the transaction data. In order to ensure the confidentiality of the data, the bank encrypts the transaction data and then sends the encrypted transaction data to a third-party data processing center for analysis and processing. And after the third-party data processing center obtains the encrypted data, analyzing the encrypted data to obtain a result, and then returning the result to the bank. In the process, the third-party data processing center obtains only the ciphertext of the transaction data, so that the operation of the third-party data processing center is completed on the basis of the encrypted data, and the data processing center does not know the plaintext of the transaction data.
Homomorphic Encryption is classified into a Fully Homomorphic Encryption scheme (FHE) and a partially Homomorphic Encryption scheme (swe). Although the completely homomorphic encryption scheme has important theoretical significance, the efficiency of the completely homomorphic encryption scheme is far from meeting the requirements of practical application scenarios. The partially homomorphic encryption scheme is a more effective solution to the fully homomorphic encryption scheme, and can effectively perform homomorphic addition of any number of times and homomorphic multiplication of a certain number of times.
(2) Impact maximization set calculation
The impact maximization problem is a function based on social network data mining. Given a social network G and a positive integer k, the influence maximization problem requires finding, under a predefined propagation model, a set S of seed nodes that has the greatest influence on the social network (that is, the most influenced nodes in the social network are propagated through the selected set of seed nodes). This problem is widely used in the field of viral marketing and other services. In viral marketing, companies aim to create a cascade of products by selecting some individuals through a social network. The official would provide these people with a sample of free new products and would like to recommend the products to his/her friends in an iterative fashion through the social relationships of these people.
For convenience of understanding of the embodiment of the present disclosure, a restaurant marketing scenario is taken as an example for explanation, please refer to fig. 1, which is a schematic diagram of a restaurant marketing scenario provided for an embodiment of the present disclosure, and the schematic diagram includes:
in order to promote newly-promoted dishes in restaurants, a restaurant boss wants to select a target group from a client group 01 and provide free dishes for the target group, and wants to recommend the dishes to friends of the client group through social relations of the target group, so as to achieve the purpose of dish promotion, at the moment, the restaurant boss will make a request to a social network platform 03 through a terminal 02 (such as a computer) corresponding to the restaurant, because of the limitation of own resources, the social network platform 03 will send relevant data of the client group to a cloud service platform 04 with very strong data processing capability, the request cloud service platform 04 finds the target group through calculation processing, the cloud service platform 04 will feed results back to the social network platform 03 after finding the target group, the social network platform 03 will feed back information of the target group to the terminal 02 corresponding to the restaurant, the restaurant owner acquires the target group information from the terminal 02 and recommends dishes to the target group.
In a first aspect, please refer to fig. 2, which is a schematic flowchart of a processing method for an influence maximization data set according to an embodiment of the present disclosure, applied to a cloud service platform, including:
step 21, receiving homomorphic encrypted social data to be analyzed, wherein the social data to be analyzed comprises nodes formed by preference vectors for representing preferences of users for set topics;
here, the social data to be analyzed is basic data corresponding to a user, the social data to be analyzed may be obtained from a user terminal in daily operation by the social network platform, and the social data to be analyzed may be preprocessed on the social network platform, for example, field processing and the like; the sending subject of the social data to be analyzed can be a social network platform; the social network platform can adopt a partial Homomorphic Encryption (SWHE) algorithm as an Encryption scheme of the social data to be analyzed, and partial Homomorphic Encryption can efficiently realize Homomorphic addition and Homomorphic multiplication operations. Here, a public and private key pair (pk, sk) may be generated to a data owner, such as a social network platform, with the public key pk used for encryption and homomorphic multiplication and the private key sk used for decryption.
Here, each user may have a preference vector to reflect interest preferences for Z topics. For example, for a node i corresponding to a certain user, the preference vector piComprises the following steps:
for any user i, the social network platform may encrypt p using a partially homomorphic encryption schemeiAfter partial homomorphic encryption is employed, where [ p ]i]Comprises the following steps:
[pi]＝SWHE.Enc(pk,pi) Here, swe.
The social network platform may send the encrypted preference vector to a cloud service platform.
here, the nodes formed by the preference vectors may be constructed as a social network topology.
Here, for any two nodes i and j, the preference vector is [ p ]i]And [ p ]j]The cloud service provider calculates the inner product of the two vectors by the following formulaAs edge eijWeight w ofijI.e. wijIs the edge e from node i to node jijThe weight of (c):
wij＝[＜pi,pj＞]
here, if i ═ j, it is possible to set:
wij＝[θ]
the entire set of weights may form a weight matrix [ W ] for all nodes in the social network.
Here, s may be selected at the node receiving the greatest influence from the end userA]Then, the cloud service platform selects other nodes as total inbound weights [ W ] of the nodes in the influence maximization set based on the constructed social network topology and the testing of each nodea]Calculate the total inbound weight [ w ] for each node jp]Here, [ W ]a]Is the output of the propagation estimation algorithm in the last iteration. In the first iteration, the total inbound weight [ w ]a]Using an encrypted weight matrix W]Is initialized, i.e.:
the total inbound weight is calculated as follows:
[wp]＝[sA]·[Wa]
after calculating the total inbound weight of all nodes, we determine whether the influence of the node is greater than a threshold value, that is:
∑j neighbor of iwij≥θ
to determine whether a node is activated by its neighbors and becomes the target node.
Here, the threshold θ for encryption and the total inbound weight for encryption are required
A comparison operation is performed therebetween. A node may be represented by the integer "1" as being activated by its neighbors, and vice versa by the integer "0". If it is not
When the test selects node i as the node in the impact maximization set, node j can be activated by the current impact maximization set. Otherwise, node j cannot be activated. The overall function of this step can be described as:
however, SWHE only provides addition and multiplication operations and cannot perform a comparison operation of ciphertext data, and here, an extension of Wilson's Theorem (Wilson's Theorem) can be employed. In Wilson's theorem, (p-1)! ≡ -1(mod p) where p is a prime number greater than 2. The Wilson's theorem may be iteratively invoked to determine whether the total inbound weight is greater than a threshold θ, 0<θ<δ, δ being the total inbound weight
The upper limit of (3). To satisfy the requirement of theorem, p>δ. Given input x, the calculation output f (x):
however, here, the partially homomorphic encryption scheme swe hardly supports a large number of consecutive homomorphic multiplication operations. Thus, the above formula can be modified as follows:
cloud service platform calculates evaluation influence for each node
And the result [ sigma ]]And sending the data to a data owner for decryption, wherein the evaluated influence comprises the target node meeting the alternative condition. Here, the cloud service provider is simultaneously responsible for storing [ W ]a]And the method is used for calculating the propagation state in the next iteration.
And step 23, forming an influence maximization data set according to the target nodes.
Here, it may be that the social networking platform decrypts [ ∑ s]Then obtain
And sends the impact result Σ to the user terminal. Through a pair of sigmaiE Σ, in the target nodes, the user adds the target node i with the largest influence to the influence maximization set S.
Referring to fig. 3, a flow chart of a method for processing an influence maximization data set according to another embodiment of the present disclosure is schematically illustrated, where the calculating a total inbound weight value of each node according to the weight value, and when determining that the total inbound weight value of a node is greater than a set threshold, regarding the node as a target node includes:
here, in order to allow a user to flexibly select a node source and an influence target group, influence target group information and source group information may be carried in the influence maximizing query request, which may be used as c1To represent the number of nodes in the source group of impact-maximized rendezvous nodes, c2To indicate the number of nodes in the affected target group. The source and target groups of nodes of the impact-maximization set may be arbitrarily overlapping. To achieve this function, two matrices P may be introduced1And P2，P1Is c1X n dimensional matrix, P2Is c2X n-dimensional matrix. Assuming that in a social network with a total node number n of 3, the impact maximization aggregation node source group includes node 1 and node 2, then P1Form (2)The following were used:
P2form (A) and (B)1Similarly. Thus, given [ P1]、[P2]And a weight matrix [ W]After modification [ W ]]Can be calculated by the following formula:
[W]＝[P1]·([W]·[P2]T)
the transformed [ W ] may specify an impact maximization set node source set and an impact goal set.
As a specific embodiment, the input and output of the algorithm may be:
inputting: [ w ]ij]∈[W]: a weight of the encryption;
[ theta ]: a threshold value of encryption;
[ delta ]: an upper bound for the encrypted inbound weight;
c1，c2: the size of the source group of the maximized set node is influenced, and the size of the target group is influenced;
and (3) outputting:
encrypted node impact;
the corresponding implementation procedure may include:
Referring to fig. 4, a flow chart of a method for processing an influence maximization data set according to another embodiment of the present disclosure is shown, where determining weight values between different nodes based on the preference vector includes:
and step 42, obtaining the weight values among different nodes by calculating the inner products among the preference vectors of the different nodes.
Here, for any two nodes i and j, its preference vector is [ p ]i]And [ p ]j]The cloud service provider calculates the inner product of the two vectors as an edge e by the following formulaijWeight w ofijI.e. wijIs the edge e from node i to node jijThe weight of (c):
wij＝[＜pi,pj＞]
referring to fig. 5, a flow chart of a method for processing an influence maximization data set according to another embodiment of the present disclosure is shown, where the calculating a total inbound weight value for each node according to the weight value further includes:
here, the selected information for receiving the selected maximum influence node may be selected information that is sent by the user terminal to the cloud service platform after the maximum influence node is selected.
Here, a vector [ w ] is definedp]，
Representing current impact maximization set impactNext, node j's encrypted total inbound weight. Next, the algorithm traverses each node in the impact maximization set node source group, adding its outbound weight to all other nodes in the impact target group, respectively. For any other node j in the target group, the total weight after the inbound weight of node i is added
Comprises the following steps:
wherein the content of the first and second substances,
representing the total inbound weight for node j under the influence of the current influence-maximization set. [ w ]ij]Is the edge e from node i to node jijThe weight of (c). The homomorphic addition referred to in the above equation can be provided by swe.
Wherein, after obtaining the updated total inbound weight value of the node, the method further includes:
and when the total inbound weight value of the node is greater than the upper limit setting threshold value, updating the total inbound weight value to be the upper limit setting value, and obtaining the updated total inbound weight value of the node.
Here, in each iteration of the round, [ w ] will beij]To add to [ wp]This results in a total inbound weight
And
is continuously increasing, which would be a heavy computational overhead in the propagation estimation algorithm of the disclosed embodiments. To be able to
Control within a certain range, can be right
The following g (-) function was implemented:
here, the user sends the upper bound [ δ ] of the encrypted total inbound weight to the cloud service provider. All weights above the threshold theta are reduced to theta, and values below theta remain unchanged. To achieve the above object, the following operations are performed:
g(x)＝f(x)·θ+(1-f(x))·x
in a second aspect, please refer to fig. 6, which is a flowchart illustrating a method for processing an influence maximization data set according to another embodiment of the disclosure, applied to a social network platform, including:
here, the social data to be analyzed is basic data corresponding to a user, the social data to be analyzed may be obtained from a user terminal in daily operation by the social network platform, and the social data to be analyzed may be preprocessed on the social network platform, for example, field processing and the like; the sending subject of the social data to be analyzed can be a social network platform;
here, the social network platform may adopt a partial Homomorphic Encryption (swe) algorithm as an Encryption scheme of the social data to be analyzed, and the partial Homomorphic Encryption may efficiently implement Homomorphic addition and Homomorphic multiplication operations. Here, a public and private key pair (pk, sk) may be generated to a data owner, such as a social network platform, with the public key pk used for encryption and homomorphic multiplication and the private key sk used for decryption.
Here, each user may have a preference vector to reflect interest preferences for Z topics. For example, for a node i corresponding to a certain user, the preference vector piComprises the following steps:
for any user i, the social network platform may encrypt p using a partially homomorphic encryption schemeiAfter partial homomorphic encryption is employed, where [ p ]i]Comprises the following steps:
[pi]＝SWHE.Enc(pk,pi)，
the social network platform may send the encrypted preference vector to a cloud service platform.
And 63, receiving target node information sent by the cloud service platform according to the received social data to be analyzed, decrypting the target node information, and forming a data set with the largest influence according to the target node.
In a third aspect, please refer to fig. 7, which is a flowchart illustrating a method for processing an influence maximization data set according to another embodiment of the present disclosure, applied to a terminal, and including:
Referring to fig. 8, a schematic flow chart of a method for processing an influence maximization data set according to another embodiment of the present disclosure is shown, where after receiving a target node obtained from a cloud service platform and sent by a social network platform, the method further includes:
Here, initially, the user has no preference for selection of nodes in the impact maximization set, so initially, s A0. Starting from the second round (if any, meaning k)>1) Only one node per round is selected and added to the influence maximization set, so
Here, the adding the target node having the inbound total weight value within a set weight value range to an influence maximization data set as a node having a maximum influence includes:
and after the total inbound weight values are sequenced, when the total inbound weight values comprise two target nodes with the same total inbound weight values in the target nodes in the weight value setting range, selecting the target nodes with relatively smaller index numbers as the nodes with the largest influence to be added into the influence maximization data set.
Here, it may be for σiE, after sorting by Σ, the user adds the target node i with the largest influence to the influence maximization set S. When an impact value conflict is encountered, the user may select the node with the smaller index number. User settings
Here, the adding the target node having the inbound total weight value within a set weight value range to an influence maximization data set as a node having a maximum influence includes:
obtaining a selected instruction for a node with the greatest influence among the target nodes;
here, the selected instruction may be a user trigger.
And when the number of the nodes with the maximum influence is determined not to be in a set range according to the selected instruction, returning to the step of receiving the target nodes which are sent by the social network platform and acquired from the cloud service platform until the number of the nodes with the maximum influence meets the set range.
If | S |<k, the user constantly sends a selection of nodes s containing the most influential nodeA]And giving the cloud service platform until the size of the influence maximization set reaches k.
In a fourth aspect, please refer to fig. 9, which is a schematic structural diagram of a cloud service platform provided in an embodiment of the present disclosure, including a first receiving module 91 and a first processing module 92, wherein,
the first receiving module 91 is configured to receive homomorphic encrypted social data to be analyzed, where the social data to be analyzed includes nodes formed by preference vectors used for characterizing preferences of users for set topics;
the first processing module 92 is configured to determine a weight value between different nodes based on the preference vector, where the weight value is used to characterize an interaction between the nodes; calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is determined to be larger than a set threshold value; and forming an influence maximization data set according to the target nodes.
In a fifth aspect, please refer to fig. 10, which is a schematic structural diagram of a social networking platform according to another embodiment of the present disclosure, including a first obtaining module 101, a second receiving module 102, and a second processing module 103, wherein,
the first obtaining module 101 is configured to obtain social data to be analyzed; the social data to be analyzed comprises nodes formed by preference vectors used for representing preference of users on set topics;
the second receiving module 102 is configured to receive target node information sent by the cloud service platform according to the received social data to be analyzed;
the second processing module 103 is configured to encrypt the social data to be analyzed through a homomorphic encryption algorithm and send the encrypted social data to the cloud service platform; and decrypting the target node information, and forming an influence maximization data set according to the target node.
In a sixth aspect, please refer to fig. 11, which is a schematic structural diagram of a terminal according to another embodiment of the present disclosure, including a second obtaining module 111 and a third processing module 112, wherein,
the second obtaining module 111 is configured to obtain social data and send the social data to a social network platform; the social data is used for forming social data to be analyzed which is composed of nodes formed by preference vectors representing preferences of users to set subjects after homomorphic encryption of the social network platform;
the third processing module 112 is configured to receive a target node obtained from a cloud service platform and sent by the social network platform, where the target node is determined by the cloud service platform by determining a weight value between different nodes based on the preference vector and calculating a relationship between a total inbound weight value of each node and a set threshold according to the weight value.
In a seventh aspect, please refer to fig. 12, which is a schematic structural diagram of an influence maximization data set processing apparatus provided in an embodiment of the present disclosure, applied to a cloud service platform, including: a first processor 121 and a memory 122 for storing computer programs capable of running on the first processor 121; wherein the first processor 121 is configured to execute the following steps when running the computer program: receiving homomorphic encrypted social data to be analyzed, wherein the social data to be analyzed comprises nodes formed by preference vectors used for representing the preference of a user on a set theme; determining a weight value between different nodes based on the preference vector, wherein the weight value is used for representing the interaction force between the nodes; calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is determined to be larger than a set threshold value; and forming an influence maximization data set according to the target nodes.
The first processor 121 is further configured to execute the following steps when the computer program is executed: receiving an influence maximization query request sent by a terminal, wherein the request carries influence target group information of a preselected target node and source group information of the influence node; calculating a total inbound weight value corresponding to each preselected target node in the influence target group information and each influence node in the source group information based on the weight values, and taking the preselected target node as a target node when the total inbound weight value of the preselected target node is determined to be larger than a set threshold value.
The first processor 121 is further configured to execute the following steps when the computer program is executed: acquiring the preference vector of each node after homomorphic encryption; and obtaining weight values between different nodes by calculating inner products between the preference vectors of the different nodes.
The first processor 121 is further configured to execute the following steps when the computer program is executed: receiving selected information of a selected maximum influence node; increasing a set weight value of a total inbound weight value of a corresponding node according to the selected information to obtain an updated total inbound weight value of the node; and setting a weight value as an outbound weight value corresponding to the selected maximum influence node.
The first processor 121 is further configured to execute the following steps when the computer program is executed: and when the total inbound weight value of the node is greater than the upper limit setting threshold value, updating the total inbound weight value to the upper limit setting value, and obtaining the updated total inbound weight value of the node.
In an eighth aspect, please refer to fig. 13, which is a schematic structural diagram of an influence maximization data set processing apparatus provided in another embodiment of the present disclosure, applied to a social network platform, including: a second processor 131 and a memory 132 for storing computer programs capable of running on the second processor 131; wherein, the second processor 131 executes the following steps when running the computer program: acquiring social data and sending the social data to a social network platform; the social data is used for the social network platform to form social data to be analyzed, wherein the social data to be analyzed is composed of nodes formed by preference vectors representing preferences of users on set topics after homomorphic encryption;
receiving a target node which is sent by the social network platform and acquired from a cloud service platform; wherein the target node is determined by the cloud service platform by determining a weight value between different nodes based on the preference vector and calculating a relationship between a total inbound weight value of each node and a set threshold according to the weight value.
In a ninth aspect, please refer to fig. 14, which is a schematic structural diagram of an influence maximization data set processing apparatus provided in another embodiment of the present disclosure, applied to a terminal, including: a third processor 141 and a memory 142 for storing computer programs capable of running on the third processor 141; wherein the third processor 141 is configured to execute the following steps when running the computer program: obtaining social data and sending the social data to a social network platform; the social data is used for forming social data to be analyzed, which is composed of nodes formed by preference vectors representing preferences of users on set topics after homomorphic encryption is carried out on the social network platform; receiving a target node which is sent by the social network platform and acquired from a cloud service platform; the target node is determined by the cloud service platform through determining weight values among different nodes based on the preference vector and calculating the relation between the total inbound weight value of each node and a set threshold according to the weight values.
The third processor 141 is further configured to execute the following steps when the computer program is executed: and sorting the target nodes by the total inbound weight values, and adding the target nodes with the total inbound weight values within a set weight value range to the influence maximization data set as the nodes with the maximum influence.
The third processor 141 is further configured to execute the following steps when the computer program is executed: and after the total inbound weight values are sequenced, when the total inbound weight values comprise two target nodes with the same total inbound weight values in the target nodes in the weight value setting range, selecting the target node with a relatively smaller index number as the node with the largest influence to be added into the influence maximization data set.
The third processor 141 is further configured to execute the following steps when the computer program is executed: obtaining a selected instruction for a node with the greatest influence among the target nodes; and when the number of the nodes with the maximum influence is determined not to be in a set range according to the selected instruction, returning to the step of receiving the target nodes which are sent by the social network platform and acquired from the cloud service platform until the number of the nodes with the maximum influence meets the set range.
In a tenth aspect, please refer to fig. 15, which is a schematic structural diagram of a data set processing system for maximizing influence according to another embodiment of the present disclosure, including a cloud service platform 153 according to the embodiment of the present disclosure, a social network platform 152 according to the embodiment of the present disclosure, and a terminal 151 according to the embodiment of the present disclosure.
The system is used for realizing the following steps:
the method comprises the steps that 1, a cloud service platform receives social data to be analyzed, which are homomorphically encrypted by a social network platform, wherein the social data to be analyzed comprise nodes formed by preference vectors used for representing preference of a user to a set theme;
and 7, updating the influence maximization set according to the target node in the influence result.
To facilitate an understanding of the embodiments of the disclosure, the disclosure is exemplified by the following embodiments:
examples
Referring to fig. 16, a flow chart of a method for processing an influence maximization data set according to another embodiment of the present disclosure is shown, where the embodiment simulates an influence maximization query process on a social network with 1000 nodes. We set the size of the impact maximization set S to 3, i.e. k is 3, so the selection of the impact maximization set needs to go through 3 iterations.
A1, the cloud service platform receives homomorphic encrypted social data to be analyzed, which is sent by a social network platform, wherein the social data to be analyzed comprises nodes formed by preference vectors for representing preferences of users on set topics;
step a2, determining weight values among different nodes by the cloud service platform based on the preference vectors, wherein the weight values are used for representing the interaction force among the nodes;
step a3, the cloud service platform receives an influence maximization query request sent by a terminal, wherein the influence maximization query request includes information about an influence maximization set node source group and an influence target group selected by a user, the number of the source groups c1 is 10, and the number of the influence target groups c2 is 10; wherein the request carries the influence target group information of the preselected target node and the source group information of the influence node; aggregating node source and impact target groups for selected impact maximization (c)1＝c210), please refer to fig. 17, which is a schematic diagram of a weight matrix of a node in a social network according to an embodiment of the present disclosure.
Step a4, the cloud service platform calculates a total inbound weight value corresponding to each preselected target node in the influence target group information and each influence node in the source group information based on the weight value, determines that the total inbound weight value of the preselected target node is greater than a set threshold value, takes the preselected target node as a target node, sends an influence result corresponding to the target node to the social network platform, decrypts the influence result, and forwards the influence result to the user terminal.
Specifically, the process is as shown in fig. 18a, 18b and 18 c. Referring to FIG. 18a, in the 1 st iteration, [ s ]A]＝{[0],[0],[0],[0],[0],[0],[0],[0],[0],[0]The cloud service platform evaluates the influence of each node [ Sigma ]]＝{[3],[3],[2],[2],[5],[2],[4],[3],[3],[3]And the influence result is returned to the end user through the social network platform.
See FIG. 18b, following iteration 2, [ sA]＝{[0],[0],[0],[0],[1],[0],[0],[0],[0],[0]The cloud service platform is used for calculating to obtain influence [ Sigma ]]＝{[8],[6],[7],[7],[6],[6],[7],[6],[6],[6]And the influence result is returned to the end user through the social network platform.
Referring to FIG. 18c, in iteration 3, [ s ]A]＝{[1],[0],[0],[0],[0],[0],[0],[0],[0],[0]The cloud service provider evaluates the impact of each node as [ sigma ]]＝{[8],[8],[8],[9],[8],[8],[10],[9],[9],[8]And returning the influence result to the end user through the social network platform.
Step a5, the user terminal selects the node with the largest influence from the target nodes, adds the node with the largest influence into the data set with the largest influence and sends the selection result to the cloud service platform.
In the first iteration, the user terminal selects the node 5 with the largest influence from the target nodes corresponding to the influence result, adds the node into the influence maximization set S, and sends the selection result SA]＝{[0],[0],[0],[0],[1],[0],[0],[0],[0],[0]And giving the cloud service platform.
In the second iteration, the user adds node 1 to the influence maximization set S and sends the selection result SA]＝{[1],[0],[0],[0],[0],[0],[0],[0],[0],[0]And giving the cloud service platform. The user selects node 7 to add to the impact maximization set S.
Here, the selection process of the influence maximization set node is as shown in fig. 19.
The above description is only for the preferred embodiment of the present disclosure, and is not intended to limit the scope of the present disclosure. Any modification, equivalent replacement, and improvement made within the spirit and scope of the present disclosure are included in the protection scope of the present disclosure.
Claims (11)
1. A processing method of an influence maximization data set is applied to a cloud service platform and is characterized by comprising the following steps:
receiving homomorphic encrypted social data to be analyzed, wherein the social data to be analyzed comprises nodes formed by preference vectors used for representing preference of users to set topics;
determining a weight value between different nodes based on the preference vector, wherein the weight value is used for representing the interaction force between the nodes; the determining weight values between different nodes based on the preference vector comprises: acquiring the preference vector of each node after homomorphic encryption; obtaining weight values between different nodes by calculating inner products between the preference vectors of the different nodes;
calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is larger than a set threshold value; wherein the calculating a total inbound weight value for each node according to the weight value comprises: receiving selected information of the selected maximum influence node; adding a set weight value to the total inbound weight value of the corresponding node according to the selected information to obtain an updated total inbound weight value of the node; wherein the set weight value is an outbound weight value corresponding to the selected maximum influence node;
and forming an influence maximization data set according to the target nodes.
2. The impact maximization data set processing method of claim 1, wherein the calculating a total inbound weight value for each node according to the weight value, and determining that the total inbound weight value for a node is greater than a set threshold, regarding the node as a target node comprises:
receiving an influence maximization query request sent by a terminal, wherein the request carries influence target group information of a preselected target node and source group information of the influence node;
calculating a total inbound weight value corresponding to each preselected target node in the influence target group information and each influence node in the source group information based on the weight values, and taking the preselected target node as a target node when the total inbound weight value of the preselected target node is determined to be larger than a set threshold value.
3. The impact maximization data set processing method of claim 1, wherein after obtaining the updated total inbound weight value for the node, further comprising:
and when the total inbound weight value of the node is greater than an upper limit set threshold value, updating the total inbound weight value to a set upper limit value, and obtaining the updated total inbound weight value of the node.
4. A processing method of an influence maximization data set is applied to a terminal, and is characterized by comprising the following steps:
acquiring social data and sending the social data to a social network platform; the social data is used for the social network platform to form social data to be analyzed, wherein the social data to be analyzed is composed of nodes formed by preference vectors representing preferences of users on set topics after homomorphic encryption;
receiving a target node which is sent by the social network platform and acquired from a cloud service platform; wherein the target node is determined by the cloud service platform by determining weight values between different nodes based on the preference vector and calculating a relationship between a total inbound weight value of each node and a set threshold according to the weight values;
wherein the process of the cloud service platform determining the target node comprises:
obtaining the preference vector of each node after homomorphic encryption; obtaining weight values between different nodes by calculating inner products between the preference vectors of the different nodes; calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is larger than a set threshold value; wherein said calculating a total inbound weight value for each node according to said weight values comprises: receiving selected information of a selected maximum influence node; adding a set weight value to the total inbound weight value of the corresponding node according to the selected information to obtain an updated total inbound weight value of the node; and setting a weight value as an outbound weight value corresponding to the selected maximum influence node.
5. The influence maximization data set processing method according to claim 4, wherein after receiving the target node obtained from the cloud service platform and sent by the social network platform, the method further comprises:
and sorting the target nodes by the total inbound weight values, and adding the target nodes with the total inbound weight values within a set weight value range to the influence maximization data set as the nodes with the maximum influence.
6. The influence maximization data set processing method according to claim 5, wherein said adding the target node having the inbound total weight value within a set weight value range as the node having the greatest influence to the influence maximization data set comprises:
and after sorting the total inbound weight values, selecting the target node with a relatively smaller index number as the node with the maximum influence to be added into the influence maximization data set when the total inbound weight values comprise two target nodes with the same total inbound weight values in the target nodes in the weight value setting range.
7. The impact maximization data set processing method of claim 5, wherein adding the target node having the inbound total weight value within a set weight value range to an impact maximization data set as the node having the greatest impact comprises:
obtaining a selected instruction for a node with the greatest influence among the target nodes;
and when the number of the nodes with the maximum influence is determined not to be in a set range according to the selected instruction, returning to the step of receiving the target nodes which are sent by the social network platform and acquired from the cloud service platform until the number of the nodes with the maximum influence meets the set range.
8. The cloud service platform is characterized by comprising a first receiving module and a first processing module, wherein,
the first receiving module is used for receiving homomorphic encrypted social data to be analyzed, wherein the social data to be analyzed comprises nodes formed by preference vectors used for representing preferences of users on set topics;
the first processing module is configured to:
determining a weight value between different nodes based on the preference vector, wherein the weight value is used for representing the interaction force between the nodes; the determining weight values between different nodes based on the preference vector comprises: acquiring the preference vector of each node after homomorphic encryption; obtaining weight values between different nodes by calculating inner products between the preference vectors of the different nodes;
calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is determined to be larger than a set threshold value; forming an influence maximization data set according to the target nodes; wherein said calculating a total inbound weight value for each node according to said weight values comprises: receiving selected information of the selected maximum influence node; adding a set weight value to the total inbound weight value of the corresponding node according to the selected information to obtain an updated total inbound weight value of the node; and setting a weight value as an outbound weight value corresponding to the selected maximum influence node.
9. A terminal, comprising a second obtaining module and a third processing module, wherein,
the second acquisition module is used for acquiring social data and sending the social data to a social network platform; the social data is used for the social network platform to form social data to be analyzed, wherein the social data to be analyzed is composed of nodes formed by preference vectors representing preferences of users on set topics after homomorphic encryption;
the third processing module is configured to receive a target node obtained from a cloud service platform and sent by the social network platform, where the target node is determined by the cloud service platform by determining a weight value between different nodes based on the preference vector and calculating a relationship between a total inbound weight value of each node and a set threshold according to the weight value; wherein the process of the cloud service platform determining the target node comprises: obtaining the preference vector of each node after homomorphic encryption; obtaining weight values between different nodes by calculating inner products between the preference vectors of the different nodes; calculating a total inbound weight value of each node according to the weight value, and taking the node as a target node when the total inbound weight value of the node is determined to be larger than a set threshold value; wherein said calculating a total inbound weight value for each node according to said weight values comprises: receiving selected information of the selected maximum influence node; adding a set weight value to the total inbound weight value of the corresponding node according to the selected information to obtain an updated total inbound weight value of the node; and setting a weight value as an outbound weight value corresponding to the selected maximum influence node.
10. An influence maximization data set processing apparatus, comprising: a processor and a memory for storing a computer program capable of running on the processor; wherein the processor is configured to implement the impact maximization data set processing method of any one of claims 1 to 3 or 4 to 7 when running the computer program.
11. An impact maximization data set processing system, comprising the cloud service platform of claim 8 or the terminal of claim 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910425628.9A CN111988129B (en) | 2019-05-21 | 2019-05-21 | Influence maximization data set processing method, device and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910425628.9A CN111988129B (en) | 2019-05-21 | 2019-05-21 | Influence maximization data set processing method, device and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111988129A CN111988129A (en) | 2020-11-24 |
| CN111988129B true CN111988129B (en) | 2022-07-01 |
Family
ID=73436193
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910425628.9A Active CN111988129B (en) | 2019-05-21 | 2019-05-21 | Influence maximization data set processing method, device and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111988129B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114676324A (en) * | 2022-03-28 | 2022-06-28 | 网易（杭州）网络有限公司 | Data processing method, device and equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104092567A (en) * | 2014-06-26 | 2014-10-08 | 华为技术有限公司 | Method and device for confirming influence sequencing of users |
| CN105117422A (en) * | 2015-07-30 | 2015-12-02 | 中国传媒大学 | Intelligent social network recommender system |
| CN107123056A (en) * | 2017-03-03 | 2017-09-01 | 华南理工大学 | A kind of location-based social big data information maximization method |
| CN108492201A (en) * | 2018-03-29 | 2018-09-04 | 山东科技大学 | A kind of social network influence power maximization approach based on community structure |
-
2019
- 2019-05-21 CN CN201910425628.9A patent/CN111988129B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104092567A (en) * | 2014-06-26 | 2014-10-08 | 华为技术有限公司 | Method and device for confirming influence sequencing of users |
| CN105117422A (en) * | 2015-07-30 | 2015-12-02 | 中国传媒大学 | Intelligent social network recommender system |
| CN107123056A (en) * | 2017-03-03 | 2017-09-01 | 华南理工大学 | A kind of location-based social big data information maximization method |
| CN108492201A (en) * | 2018-03-29 | 2018-09-04 | 山东科技大学 | A kind of social network influence power maximization approach based on community structure |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111988129A (en) | 2020-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Riazi et al. | Chameleon: A hybrid secure computation framework for machine learning applications | |
| Xu et al. | Am I eclipsed? A smart detector of eclipse attacks for Ethereum | |
| US11196541B2 (en) | Secure machine learning analytics using homomorphic encryption | |
| EP3198904B1 (en) | Privacy-preserving cookies for personalization without user tracking | |
| CN109891424B (en) | Establishing links between identifiers without revealing specific identifying information | |
| Jung et al. | Privacy oracle: a system for finding application leaks with black box differential testing | |
| AU2018222992B2 (en) | System and method for secure two-party evaluation of utility of sharing data | |
| CN111428887B (en) | Model training control method, device and system based on multiple computing nodes | |
| US20150188941A1 (en) | Method and system for predicting victim users and detecting fake user accounts in online social networks | |
| JP2016509268A (en) | Counting method and system to protect privacy | |
| CN112534431B (en) | Improving security of cryptographically protected resources based on publicly available data | |
| US10599871B2 (en) | System and method for privacy aware information extraction and validation | |
| WO2022237175A1 (en) | Graph data processing method and apparatus, device, storage medium, and program product | |
| Badsha et al. | Privacy preserving user based web service recommendations | |
| CN117390657A (en) | Data encryption method, device, computer equipment and storage medium | |
| CN111988129B (en) | Influence maximization data set processing method, device and system | |
| Medvet et al. | Exploring the usage of topic modeling for android malware static analysis | |
| Goulet et al. | Hidden-service statistics reported by relays | |
| Saputra et al. | Federated learning framework with straggling mitigation and privacy-awareness for AI-based mobile application services | |
| David et al. | Poly-logarithmic side channel rank estimation via exponential sampling | |
| CN110874481A (en) | GBDT model-based prediction method and device | |
| Lin et al. | Opinion leaders discovering in social networks based on complex network and dbscan cluster | |
| CN113157938B (en) | Method and device for jointly processing multiple knowledge graphs for protecting privacy data | |
| CN115033916A (en) | Multi-party combined data based push model training and information push method and device | |
| Barona et al. | Optimal cryptography scheme and efficient neutrosophic C-means clustering for anomaly detection in cloud environment |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |