CN104866781B - The community network data publication method for secret protection of Community-oriented detection application - Google Patents
The community network data publication method for secret protection of Community-oriented detection application Download PDFInfo
- Publication number
- CN104866781B CN104866781B CN201510277513.1A CN201510277513A CN104866781B CN 104866781 B CN104866781 B CN 104866781B CN 201510277513 A CN201510277513 A CN 201510277513A CN 104866781 B CN104866781 B CN 104866781B
- Authority
- CN
- China
- Prior art keywords
- node
- community
- super
- data
- network data
- 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
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
- G06F21/6254—Protecting personal data, e.g. for financial or medical purposes by anonymising data, e.g. decorrelating personal data from the owner's identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2221/00—Indexing scheme relating to security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F2221/21—Indexing scheme relating to G06F21/00 and subgroups addressing additional information or applications relating to security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F2221/2107—File encryption
Abstract
The present invention discloses a kind of community network data publication method for secret protection of Community-oriented detection application, after data are carried out with initialization data and community's detection, the node in network is arranged;Data are formed into super node by the way that K is anonymous, and it is the Super Edge of Weight generally to change side;Super node and Super Edge are split;Community network data after issue anonymity.The present invention is when data are issued, it is ensured that while anonymity is required, preferably saves the community structure of original graph this structural information, is conducive to the correlation analysis of data analysis Zhe Zuo communities.
Description
Technical field
The present invention relates to network data security technical field, and in particular to a kind of community network of Community-oriented detection application
Data publication method for secret protection.
Background technology
With the development of community network application, such as Facebook, MySpace etc. are made friends by social network sites, are joined
The user of system is more and more.This has also attracted the notice of more data research personnel and application developer.By to society
Can network analysis can provide more rich, more accurately data source for data mining and pattern analysis.But, issue social network
Network data will cause the sensitive information leakage of user, and this will cause harm to user to a certain extent.Due to community network
Sensitive information variation in data, such as node is identified, nodal community leakage, and the relation between node is identified etc., then reveal
Mode also have various, corresponding anonymous safeguard procedures also have various.And protect the privacy needs of community network data to be based on
The characteristics of network data, carrys out the corresponding protection technique of research institute.It is the most frequently used and intuitively a kind of anonymous methods be it is simple anonymous, i.e.,
Remove the explicit identification symbol attribute of energy unique mark user (node i.e. in figure), such as name, birthday.However, many previous
Research has confirmed that simple anonymity is not enough to protect privacy of user.
In order to preferably protect the data-privacy in community network, existing community network data anonymization method to have:Node
K- is anonymous, i.e., similar or closely located according to structure, and all node clusterings in community network are formed into super node so that
Each super node comprises at least K primitive network node of graph.Meanwhile, the connected side of super node is generalized as Super Edge.Due to,
Node inside each super node undistinguishable each other, so, the identified probability of this Anonymizing networks node less than etc.
In 1/K;Subgraph K is anonymous, that is, assume attacker know destination node where specific picture information, in making community network after anonymity
At least K-1 subgraph undistinguishable therewith.In addition the methods such as the anonymity of K- degree, randomization, difference privacy also protect
User profile is not compromised.But in current community network secret protection research, more emphasize corresponding to special scenes
The method for secret protection of anonymity model, the purposes of less specified issue data, so during anonymity, changes side and node
Certain original structure attribute will to a certain extent be destroyed.And in actual life, for the community network data issued
It is commonly used in various special-purposes, such as:Data cube computation is excavated, and community center finds etc., and this is accomplished by improving original anonymity
Method.
The content of the invention
Existing community network method for secret protection to be solved by this invention can to a certain extent destroy primitive network
A kind of deficiency of certain structure attribute, there is provided community network data publication method for secret protection of Community-oriented detection application, its
When data are issued, it is ensured that while anonymity is required, the community structure of original graph this structural information is preferably saved, favorably
In the correlation analysis of data analysis Zhe Zuo communities.
To solve the above problems, the present invention is achieved by the following technical solutions:
The community network data publication method for secret protection of Community-oriented detection application, comprises the following steps:
Step 1, first initialization data;Again the data after initialization are carried out with preliminary community's detection, before obtaining anonymity
Initial community divides;The node set of data is arranged by the number of degrees afterwards, is obtained new sequence node;
Step 2, K are anonymous to form super node;I.e.
The maximum node of the number of degrees in step 2.1, selection present node sequence, as the seed node of current super node,
And the node chosen is removed in node set;
Step 2.2, the distance of each node calculated in this super node and node set, select nearest node to close
And be super node, and the node chosen is removed in node set;
Step 2.3, continue to repeat the above steps 2.2, that is, calculate distance, merge node and update the process of node set,
Until the horizontal K of privacy that the node number that current super node is included reaches setting, then start the treatment of next super node;
Step 2.4, repeat the above steps 2.1-2.3, until remaining node number is hidden less than setting in node set
Private horizontal K;
Step 2.5, the distance of node and several super nodes established before remained by difference calculate node set,
And remaining node is incorporated into the minimum super node of distance one by one, until node set is sky, i.e. node set
All nodes be clustered into super node;
Connection two is super during step 3, the Super Edge that the side of data is generalized as Weight, wherein weight are original graph
The number on the side between node;
Step 4, super node and Super Edge are split;
Community network data after step 5, issue anonymity.
In step 1, initialization data is the identity property for removing display, uses the identifier for renumbeing instead and represents.
In step 1, detected using GN algorithm Lai Zuo communities.
The node set of data is pressed into the arrangement of number of degrees descending in step 1, new sequence node is obtained;Now, selected in step 2
The node that the node of number of degrees maximum in present node sequence ranks the first is selected, as the seed node of current super node.
In step 2.2, when have multiple nodes minimum with the distance of current super node and it is identical when, then prioritizing selection with
The seed node of current super node is that the both candidate nodes of same community merge into super node;If do not had in both candidate nodes
During with the node that the seed node of current super node is same community, then prioritizing selection single node community node is merged into super
Level node.
In step 2.3 and 2.4, the span of the horizontal K of privacy for setting is 1 < K≤n, during wherein n is primitive network figure
Node total number.
In step 2.5, when the node in present node set is minimum and identical with the distance of multiple super nodes, then ought
Preceding remaining node is merged into seed node therewith in the super node of same community.
In step 4, super node and Super Edge are split with equiprobability.
Compared with prior art, the present invention detects this specific data application target for community, mainly from following several
Aspect improves original anonymity algorithm:
1st, because to protect community structure, the change being allowed to before and after anonymity is minimum.The present invention is in logarithm at first
Detected according to preliminary community has been carried out, obtain initial community and divide.Additionally, in order to follow-up selection node needs, to node
List V presses the arrangement of number of degrees descending.
2nd, during the minimum node of each super node chosen distance merges into super node, both candidate nodes have multiple
When, preferably select and the node that the seed node of current super node is same community.
When the node of community the 3rd, is not all with the seed node of current super node in both candidate nodes, candidate is selected
The minimum node of the node at end in node listing, the i.e. number of degrees.So can guarantee that prioritizing selection single node community node is (such
Often the number of degrees are zero to node, must be that the number of degrees are minimum, and come the end of candidate node list if there is such node
Tail), so can guarantee that and less destroy other non-single node communities.
4th, when the interstitial content in node listing is less than K, it is necessary to these nodes are added separately into distance minimum therewith
Super node in, it is preferential to add seed node V when such super node is more than oneseedIt is therewith same society
In the super node in area.
Brief description of the drawings
Fig. 1 is the flow chart of the community network data publication method for secret protection of detection application in Community-oriented of the present invention.
Fig. 2 is a kind of original graph of community network data.
Fig. 3 is the community network data publication method for secret protection of preferred embodiment of the present invention Community-oriented detection application
Flow chart.
Fig. 4 is the figure after being split by corresponding probability after the community network data K anonymities of Fig. 2.
Specific embodiment
The community network data publication method for secret protection of Community-oriented detection application, as shown in figure 1, it includes mistake as follows
Journey:
Step one, initialization data, that is, remove the identity property of display, uses the identifier for renumbeing instead and represents.To this
A little data carry out preliminary community's detection, obtain the division of the community before anonymity.In addition, then by node set V dropped by the number of degrees
Sequence is arranged, and obtains new sequence node.Community detection method of the present invention, can using GN algorithms (Girvan and
The community structure detection algorithm that Newman is proposed) detection of Lai Zuo communities, or other known community detection methods.
The anonymous process for forming super node of step 2, K.During this:
First, first seed node of super node is selected, the section of number of degrees maximum in present node list is selected here
Point, that is, the node for ranking the first now removes the node chosen in node set V.
Then, the distance of each node in this super node and node set V is calculated, selects nearest node to be combined into
Super node.When have multiple nodes minimum with the distance of current super node and it is identical when, pay the utmost attention to and current super section
The seed node of point is the both candidate nodes of same community;If in both candidate nodes not with the seed node of current super node
For same community node when, prioritizing selection single node community node.Now the node chosen is removed in node set V.
Then, continue cycling through and calculate distance, merge node and update node set until current super node is included
Node number reaches privacy horizontal K, i.e., >=K.The span of the horizontal K of privacy is 1 < K≤n, during wherein n is primitive network figure
Node total number.
Afterwards, the seed node of next super node is selected.Repetition above step is individual until remaining node set V's
Number is less than K, then calculate the distance of these nodes and above established several super nodes respectively, is then combined with distance minimum
Super node in, when such super node has multiple, prioritizing selection seed node is super for same community therewith
Node merges.Until node set is sky, i.e., all of node is clustered into super node.
Step 3, the Super Edge that the side in original graph is generalized as Weight, wherein weight connect two in being original graph
The number on the side between individual super node.
Step 4, complete the work of cluster after, be the anonymous process of weight, will super node and Super Edge split.
Method for splitting of the present invention, can be equiprobability method for splitting, or other known method for splitting.
Community network data after step 5, issue anonymity.
This application is detected below for the community in community network data publication purposes, with specific example to the present invention
It is further elaborated.
Original graph, as shown in Figure 2.
The n dimension boolean vectors of node:
V1:B1(b 0 1 0 1 1 0 0 0 0 1 0 0)
V2:B2(0 b 1 0 0 0 1 1 1 0 0 0 0)
V3:B3(1 1 b 1 0 0 0 0 0 0 0 0 0)
V4:B4(0 0 1 b 0 0 0 1 0 0 0 1 0)
V5:B5(1 0 0 0 b 1 0 0 0 0 0 0 0)
V6:B6(1 0 0 0 1 b 0 0 0 0 0 0 0)
V7:B7(0 1 0 0 0 0 b 0 0 0 0 0 0)
V8:B8(0 1 0 1 0 0 0 b 0 0 0 0 0)
V9:B9(0 1 0 0 0 0 0 0 b 0 0 0 0)
V10:B10(0 0 0 0 0 0 0 0 0 b 0 0 0)
V11:B11(1 0 0 0 0 0 0 0 0 0 b 0 0)
V12:B12(0 0 0 1 0 0 0 0 0 0 0 b 0)
V13:B13(0 0 0 0 0 0 0 0 0 0 0 0 b)
Range formula between two nodes:
Range formula between node and overtrick:
The community network data publication method for secret protection of the Community-oriented detection application that this application example is used, such as schemes
Shown in 3, it comprises the following steps:
Step 1:Initialization data set, and data are carried out with preliminary community's detection, obtain initial community and divide.Then
Node set V is carried out into descending arrangement by the number of degrees, to facilitate the selection of down-stream interior joint.
Since it is considered that this anonymous methods, it is ensured that the data of issue, can be to greatest extent while level of anonymity is reached
Preserve original community structure information, it is therefore desirable to community's detection first is carried out to data before anonymity, obtain initial community and draw
Point.Then the influence to community structure is reduced as far as possible during anonymity.The present invention can be divided using classical community and calculated
Method is the detection of GN algorithm Lai Zuo communities.The division result of the community that GN algorithms are obtained is:{V1,V11,V5,V6, { V9,V2,V7,
{V8,V12,V3,V4, { V10, { V13}
Node is pressed into number of degrees sequence V:{V1=4, V2=4, V3=3, V4=3, V5=2, V6=2, V8=2, V7=1, V9=
1, V11=1, V12=1, V10=0, V13=0 }
Step 2:Judge whether current node listing V is empty.If it is empty, then step 11 is gone to.Otherwise go to step 3.
Step 3:Judge that whether the number of current node listing interior joint, less than K, if less than K, goes to step 10, it is no
Then go to step 4.It is the several times of the horizontal K of privacy because the community network data volume in reality is than larger.So without examining
Worry program is performed in this step once just there is the situation for jumping out circulation.
Step 4:Select seed node (first node of the super node to be formed) V of super nodei seed;Then when
Preceding super node is cli={ Vi seed};Current node set V=V-cli。
Step 5:Judge whether the node number that current super node is included is equal to K.If equal to K, then i++, then
Step 2 is turned to, step 6 is otherwise turned to.
Step 6:The distance of each node in current super node and node set V is calculated, is obtained and super node distance
Minimum node is used as both candidate nodes (node in now both candidate nodes set C is still to be arranged by number of degrees descending).
Step 7:Judge whether the both candidate nodes number minimum with current super nodal distance be more than one.If so
Step 8 is gone to, step 9 is otherwise gone to.
Step 8:Node in traversal both candidate nodes set successively, when having the seed node with current super node
Vi seedDuring for same community, then terminate traversal, select this node to be merged into current super node cliIn;If traversed
Last node of both candidate nodes set, during without such node occur, then selects last section in candidate collection C
Point is merged into current super node.
In this step, community structure information is protected in terms of following two:
When both candidate nodes composition super node is selected, when the seed node of both candidate nodes and current super node has same society
During area's node, prioritizing selection.
When in both candidate nodes not with the seed node of current super node with community node when, prioritizing selection single node
Community's node, the so less destruction community structure of energy.Such as:In the case of 3- anonymities, the remaining (V of current V lists1, V2,
V3, V4), and { V1, V2, V3It is a community, { V4It is a single node community, if the seed node of current super node is
Vi(i ≠ 1,2,3,4), and present both candidate nodes are { V1, V4, then prioritizing selection V4, (because node is arranged by number of degrees descending
Row, V1The number of degrees be necessarily more than zero, then single node community node V4Naturally last position is come.In algorithm, when both candidate nodes concentration
Not at a community, selection comes the node of list end position for node and seed node).So protect to a certain extent
{V1, V2, V3This community structure.
Step 9:This unique both candidate nodes is selected to be merged into current super node cliIn.
Step 10:Node by remaining number in node set V less than K is incorporated into the minimum super node of distance
In.For the node V in present node set Vj, if the number of the minimum super node of distance therewith is more than one,
First judge to whether there is and V in the seed node in these super nodesjCommunity is all, is preferentially merged into if existing
Such super node.If not existing, last super node in selection candidate list merges.Then removed in V
The V that anonymity is crossedj, above procedure is then circulated, distance of the remaining node with established super node, conjunction in V are calculated successively
And, update V, until V interior joints for sky.
In this step, the mode of the protection community structure information taken is analogous to step 8.
Specific to this application example, the K anonymities process of step 2-10 is:
It is not sky due to current node set V, and node number in gathering is more than K.So selection first
Super node cl1Seed node, in the node for selecting the number of degrees in node set maximum here, that is, current node set
Positioned at the first node.Now select V1, then cl1={ V1 seed}={ V1, V=V-cl1, i.e. V=(V2,V3,V4,V5,V6,V8,V7,
V9,V11,V12,V10,V13) dist (cl are calculated successively1,Vi), obtain min (dist (cl1,Vi))=dist (cl1,V5)=dist
(cl1,V6)=2/11, due to the two nodes and V1All a community, according to algorithm, then select candidate collection first same
Community's node, i.e. V5, cl1=cl1∪{V5, i.e. cl1={ V1,V5}.Then V=V-cl1, V=(V2,V3,V4,V6,V8,V7,V9,
V11,V12,V10,V13).Because | cl1|<4, cycle calculations are until cl successively1Interior joint number is equal to K, obtains cl1={ V1,V5,V6,
V11, current V=(V2,V3,V4,V8,V7,V9,V12,V10,V13)。
Current node set V is not sky, and node number is more than K, then next select super node cl2Seed section
Point, selects V here2, cl2={ V2 seed}={ V2}.Then V=V-cl2, V=(V3,V4,V8,V7,V9,V12,V10,V13).Calculate
dist(cl2,Vi), obtain min (dist (cl2,Vi))=dist (cl2,V4)=dist (cl2,V7)=dist (cl2,V9)=3/
11, in both candidate nodes, V7, V9With V2It is a community, and V4And V2Not in same community.The V in V7It is forward, V is selected here7。
cl2=cl2∪{V7, i.e. cl2={ V2,V7}.Then V=V-cl2, V=(V3,V4,V8,V9,V12,V10,V13).Because | cl2|<4,
Continue to calculate dist (cl2,Vi), now min (dist (cl2,Vi))=dist (cl2,V9)=3/22, cl2=cl2∪{V9, i.e.,
cl2={ V2,V7,V9}.Then V=V-cl2, V=(V3,V4,V8,V12,V10,V13).Because | cl2|<4, next proceed to calculate
dist(cl2,Vi), obtain min (dist (cl2,Vi))=dist (cl2,V10)=dist (cl2,V13)=6/33.Due to the two
Node and V2According to algorithm, then the node of candidate collection most end, i.e. V are not selected a community13.So can guarantee that preferential
Selection single node community node.cl2=cl2∪{V13, i.e. cl2={ V2,V7,V9,V13}.Then V=V-cl2, V=(V3,V4,V8,
V12,V10)。
Current node set V is not sky, and node number is more than K, then next select super node cl3Composition section
Point, similar above step obtains cl3={ V3,V8,V10,V12}.Then V=V-cl3, V=(V4)。
Current node set V is not sky, but node number is less than K, then next calculate V interior joints V4It is each with what is existed
The distance of individual super node.Obtain the min (dist (V of minimum4,cli))=dist (V4,cl3), according to algorithm steps, select here
Select V4Add cl3, cl3=cl3∪{V4, i.e. cl3={ V3,V4,V8,V10,V12}。
To sum up step, draws super node S={ cl1={ V1,V5,V6,V11, cl2={ V2,V7,V9,V13, cl3={ V3,
V4,V8,V10,V12}}。
Step 11:Generalization node and side.The super node of cluster is represented with a pair of ordered pairs (a, b), a represents super section
The number of point internal node, b represents connection side number of the super node internal node in primitive society's network.Between super node
Represented with the side of a Weight, the side number that wherein weight is connected with each other between being super node in original graph.
cl1In have four nodes, this four nodes have 4 sides each other, then represent super node with ordered pair (4,4)
cl1, same cl2Represented with (4,2), cl3Represented with (5,3).
cl1With cl3The mutual side quantity in original graph of super node is 1, then the weight on the side of the two super nodes
It is 1;cl2With cl3The mutual side quantity of super node is 2, then the weight on the side of the two super nodes is 2.
Step 12:Weight is anonymous, i.e., equiprobably split these super nodes and Super Edge.This process is in super node
The process on side is added between internal and super node.In super node cliInside, with P1The not connected sides of a pair of probability selection enter
Row connection, untill side number reaches original SuperNode in 504 internal edges number;In super node clpAnd clqBetween, with P2Probability choosing
Select a pair not connected sides to be attached, untill the weight that side number reaches Super Edge.So, it is whole anonymous front and rear, society
The node and side number of network keep constant.Here
The process for splitting super node and Super Edge is:
cl1Inside selection node is p=1/4* (4-1)/2=33% to the probability for connecting side, until inside connection side number
Untill 4;
cl2Inside selection node is p=1/4* (4-1)/2=33% to the probability for connecting side, until inside connection side number
Untill 2;
cl3Inside selection node is p=1/5* (5-1)/2=10% to the probability for connecting side, until inside connection side number
Untill 3;
cl1With cl3It is p=1/4*5=5% to the probability for connecting side that node is selected between super node, until connection each other
Untill side number is 1;
cl2With cl3It is p=2/4*5=10% to the probability for connecting side that node is selected between super node, until being connected with each other
Untill side number is 2.
Fig. 4 is the figure after being split by corresponding probability after the community network data K anonymities of Fig. 2.
Step 13:Issue data, terminate.
Believe the anonymous front and rear community that preserves of community network diagram data below for the algorithm for designing of the invention in the present embodiment
The situation of breath.
The community network data that the present invention is used are the simple undirected graphs of not tape label, and the background knowledge of attacker can be
Specific picture information where arbitrary node, such as degree information.Community network data need to carry out preliminary anonymity before issue
Treatment, that is, remove the display identity property of unique mark node, such as name, uses the identifier for renumbeing instead and represents.Issue
Figure with G (V, E) represent, wherein V for node combination, represent community network in individual or other entities;E is two on V
The set on side in first relation, i.e. figure, represents personal or inter-entity relation, such as friend or cooperative relationship.The figure of issue is by this
Anonymous methods treatment in invention, can effectively prevent attacker from using background knowledge to weight in the data of the issue where user
New definition, at the same time, can effectively protect the community structure information in original graph.Node can not only so be protected in community
The important information such as status, effect, be also advantageous for the relevant information that data analysis person analyzes community.
Claims (6)
1. the community network data publication method for secret protection that Community-oriented detection is applied, it is characterized in that, comprise the following steps:
Step 1, first initialization data;Again the data after initialization are carried out with preliminary community's detection, obtains initial before anonymity
Community divide;The node set of data is arranged by the number of degrees afterwards, is obtained new sequence node;
Step 2, K are anonymous to form super node;I.e.
The maximum node of the number of degrees in step 2.1, selection present node sequence, as the seed node of current super node, and
The node chosen is removed in node set;
Step 2.2, the distance of each node calculated in this super node and node set, select nearest node to merge into
Super node, and the node chosen is removed in node set;
When have multiple nodes minimum with the distance of current super node and it is identical when, then prioritizing selection and current super node
Seed node is that the both candidate nodes of same community merge into super node;If in both candidate nodes not with current super node
Seed node when being the node of same community, then prioritizing selection single node community node merges into super node;
Step 2.3, continue to repeat the above steps 2.2, that is, calculate distance, merge node and update the process of node set, until
The node number that current super node is included reaches the horizontal K of privacy of setting, then start the treatment of next super node;
Step 2.4, repeat the above steps 2.1-2.3, until privacy water of the remaining node number less than setting in node set
Flat K;
Step 2.5, the distance of node and several super nodes established before remained by difference calculate node set, and by
It is individual that remaining node is incorporated into the minimum super node of distance, until node set is sky, the i.e. institute of node set
There is node to be clustered into super node;
When node in present node set is minimum and identical with the distance of multiple super nodes, then current remaining node is closed
And to seed node therewith in the super node of same community;
Step 3, the Super Edge that the side of data is generalized as Weight, wherein weight connect two super nodes in being original graph
Between side number;
Step 4, super node and Super Edge are split;
Community network data after step 5, issue anonymity.
2. the community network data publication method for secret protection that Community-oriented detection according to claim 1 is applied, it is special
Levying is, in step 1, initialization data is the identity property for removing display, uses the identifier for renumbeing instead and represents.
3. the community network data publication method for secret protection that Community-oriented detection according to claim 1 is applied, it is special
Levying is, in step 1, is detected using GN algorithm Lai Zuo communities.
4. the community network data publication method for secret protection that Community-oriented detection according to claim 1 is applied, it is special
Levying is, in step 2.3 and 2.4, the span of the horizontal K of privacy for setting is 1 < K≤n, and wherein n is section in primitive network figure
Point total number.
5. the community network data publication method for secret protection that Community-oriented detection according to claim 1 is applied, it is special
Levying is, in step 4, super node and Super Edge is split with equiprobability.
6. the community network data publication method for secret protection that Community-oriented detection according to claim 1 is applied, it is special
Levying is, the node set of data is pressed into the arrangement of number of degrees descending in step 1, obtains new sequence node;Now, selected in step 2
The maximum node of the number of degrees is the node for ranking the first in present node sequence, used as the seed node of current super node.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510277513.1A CN104866781B (en) | 2015-05-27 | 2015-05-27 | The community network data publication method for secret protection of Community-oriented detection application |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510277513.1A CN104866781B (en) | 2015-05-27 | 2015-05-27 | The community network data publication method for secret protection of Community-oriented detection application |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104866781A CN104866781A (en) | 2015-08-26 |
CN104866781B true CN104866781B (en) | 2017-07-04 |
Family
ID=53912603
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510277513.1A Active CN104866781B (en) | 2015-05-27 | 2015-05-27 | The community network data publication method for secret protection of Community-oriented detection application |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104866781B (en) |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106599725B (en) * | 2016-12-22 | 2019-05-10 | 广西师范大学 | The randomization method for secret protection of diagram data publication |
CN106709372B (en) * | 2017-01-09 | 2019-04-30 | 广西师范大学 | The method for secret protection that enhanced identity is kept |
CN106778352B (en) * | 2017-01-13 | 2020-04-07 | 广西师范大学 | Multisource privacy protection method for combined release of set value data and social network data |
CN106980795B (en) * | 2017-02-28 | 2019-11-05 | 广西师范大学 | Community network data-privacy guard method |
CN108696418B (en) * | 2017-04-06 | 2020-07-28 | 腾讯科技(深圳)有限公司 | Privacy protection method and device in social network |
CN107070932B (en) * | 2017-04-24 | 2020-02-07 | 广西师范大学 | Anonymous method for preventing label neighbor attack in social network dynamic release |
CN107104962B (en) * | 2017-04-24 | 2020-02-07 | 广西师范大学 | Anonymous method for preventing label neighbor attack in dynamic network multi-release |
CN107742083B (en) * | 2017-10-31 | 2019-10-25 | 华中科技大学 | A kind of method for secret protection and system towards large-scale graph data publication |
CN108809764B (en) * | 2018-06-15 | 2021-11-02 | 东南大学 | GPU-based network access over-point connection number estimation method under sliding window |
CN111046429B (en) * | 2019-12-13 | 2021-06-04 | 支付宝(杭州)信息技术有限公司 | Method and device for establishing relationship network based on privacy protection |
CN113486396A (en) * | 2021-07-02 | 2021-10-08 | 北京工业大学 | Social network-oriented high-availability K-anonymous data processing method and device, electronic equipment and storage medium |
CN113706326B (en) * | 2021-08-31 | 2023-05-30 | 福建师范大学 | Mobile social network diagram modification method based on matrix operation |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101834872A (en) * | 2010-05-19 | 2010-09-15 | 天津大学 | Data processing method of K-Anonymity anonymity algorithm based on degree priority |
JP2011209800A (en) * | 2010-03-29 | 2011-10-20 | Kddi Corp | Privacy protection device for disclosed information, privacy protection method for the disclosed information and program |
KR20120063050A (en) * | 2010-12-07 | 2012-06-15 | 서울대학교산학협력단 | Method and apparatus for protecting information providing k-anonymity |
CN103218397A (en) * | 2013-03-12 | 2013-07-24 | 浙江大学 | Privacy protecting method for social network based on undirected graph modification |
CN103279713A (en) * | 2013-06-08 | 2013-09-04 | 广西师范大学 | Optimized SNS (social network service) graph data publication privacy protection method |
CN104317904A (en) * | 2014-10-24 | 2015-01-28 | 南京信息工程大学 | Generalization method for weighted social network |
-
2015
- 2015-05-27 CN CN201510277513.1A patent/CN104866781B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011209800A (en) * | 2010-03-29 | 2011-10-20 | Kddi Corp | Privacy protection device for disclosed information, privacy protection method for the disclosed information and program |
CN101834872A (en) * | 2010-05-19 | 2010-09-15 | 天津大学 | Data processing method of K-Anonymity anonymity algorithm based on degree priority |
KR20120063050A (en) * | 2010-12-07 | 2012-06-15 | 서울대학교산학협력단 | Method and apparatus for protecting information providing k-anonymity |
CN103218397A (en) * | 2013-03-12 | 2013-07-24 | 浙江大学 | Privacy protecting method for social network based on undirected graph modification |
CN103279713A (en) * | 2013-06-08 | 2013-09-04 | 广西师范大学 | Optimized SNS (social network service) graph data publication privacy protection method |
CN104317904A (en) * | 2014-10-24 | 2015-01-28 | 南京信息工程大学 | Generalization method for weighted social network |
Non-Patent Citations (1)
Title |
---|
"社会网络的数据隐私保护方法研究";王建超;《中国优秀硕士学位论文全文数据库 信息科技辑》;20140415(第4期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN104866781A (en) | 2015-08-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104866781B (en) | The community network data publication method for secret protection of Community-oriented detection application | |
Zhang et al. | PPO-CPQ: a privacy-preserving optimization of clinical pathway query for e-healthcare systems | |
Milli et al. | Model reconstruction from model explanations | |
CN107729767B (en) | Social network data privacy protection method based on graph elements | |
CN106980795B (en) | Community network data-privacy guard method | |
CN103218397B (en) | A kind of social networks method for secret protection based on non-directed graph amendment | |
CN109344326B (en) | Social circle mining method and device | |
CN109862018A (en) | Anti- crawler method and system based on user access activity | |
CN104317904B (en) | A kind of extensive method of Weight community network | |
CN111104968B (en) | Safety SVM training method based on block chain | |
CN106599725B (en) | The randomization method for secret protection of diagram data publication | |
CN107945037A (en) | A kind of social networks based on node structure feature goes de-identification method | |
Zhang et al. | Privacy threats and protection in machine learning | |
CN107070932B (en) | Anonymous method for preventing label neighbor attack in social network dynamic release | |
CN112685772A (en) | Intrinsic-computation-oriented DIKW-mode-crossing relative difference privacy protection method | |
CN109003218A (en) | A method of construction thresholding visual cryptography matrix | |
Qian et al. | Social network de-anonymization: More adversarial knowledge, more users re-identified? | |
CN110247805A (en) | A kind of method and device for propagating key node based on the identification that K shell decomposes | |
CN103973675B (en) | Method for detecting segmented redundancy in cross-domain collaboration firewalls | |
CN106778352B (en) | Multisource privacy protection method for combined release of set value data and social network data | |
Zheng et al. | Analysis of criminal social networks with typed and directed edges | |
CN106685893A (en) | Authority control method based on social networking group | |
CN115630700A (en) | Label inference attack defense method and device based on decentralized training | |
CN107104962B (en) | Anonymous method for preventing label neighbor attack in dynamic network multi-release | |
US20180114027A1 (en) | Attribute determination using secure list matching protocol |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
EXSB | Decision made by sipo to initiate substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |