CN110086670A - Large-scale complex network community discovery method and application based on local neighbor information - Google Patents
Large-scale complex network community discovery method and application based on local neighbor information Download PDFInfo
- Publication number
- CN110086670A CN110086670A CN201910356601.9A CN201910356601A CN110086670A CN 110086670 A CN110086670 A CN 110086670A CN 201910356601 A CN201910356601 A CN 201910356601A CN 110086670 A CN110086670 A CN 110086670A
- Authority
- CN
- China
- Prior art keywords
- node
- corporations
- network
- initialcommunity
- nodes
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 241000700605 Viruses Species 0.000 claims abstract description 6
- 230000007704 transition Effects 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 3
- 238000012217 deletion Methods 0.000 claims description 2
- 230000037430 deletion Effects 0.000 claims description 2
- 238000000151 deposition Methods 0.000 claims description 2
- 230000008901 benefit Effects 0.000 abstract description 8
- 238000010586 diagram Methods 0.000 description 6
- 238000001514 detection method Methods 0.000 description 5
- 102000004169 proteins and genes Human genes 0.000 description 4
- 108090000623 proteins and genes Proteins 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000010295 mobile communication Methods 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000009412 basement excavation Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000005314 correlation function Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000005304 joining Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
Classifications
-
- 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
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/12—Discovery or management of network topologies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a kind of, and the large-scale complex network community based on local neighbor information finds method, comprising the following steps: step A: obtaining network structure, the maximum non-accessed node of selectance and its all neighbor nodes as initialization corporations；Step B: it deletes in corporations and connects untight node；Step C: expand initialization corporations using network local message selection alternate node；Step D: calculate node is subordinate to coefficient and judges whether node should stay, and obtains a corporations and mark all nodes of the corporations to be to have accessed；Return step A exports community division result until non-accessed node is not present.The present invention also provides a kind of Computer Virus Spread control methods, construct computernetwork model, are obtained using the above method and divide corporations, carry out key protection to corporations' core node and boundary node.The present invention has the advantages that having good computational efficiency and arithmetic accuracy in ultra-large problem.
Description
Technical field
It is specially a kind of based on the extensive of local neighbor information the present invention relates to complex network community partitioning technology field
Complex network community finds method and application.
Background technique
Complex network is modeled as according to graph theory comprising node and connects the figure on side, provides a kind of expression real world system
The effective means of relationship between object.Such as social networks, computer network and bio-networks.Complex network research is mainly asked
Topic is the excavation of community structure with difficult point, i.e., by a true network be divided into dense inner connect sparse group between connect
The node set connect.By excavating the community structure of complex network, facilitate the function mould for more preferably understanding and excavating network system
Block, topological structure etc..
For example, establishing network according to the purchase relationship between online commodity and customer group, marks off user and buy commodity
Community structure, help to establish efficient recommender system, preferably guided by inventory user buy commodity, provide excellent
Recommendation service more.In mobile communication network, whole network is made of communication node movable in the same area, because mobile
Communication apparatus is mobile rapidly and active, is used to specify the routing table how a communication node communicates with other nodes and is difficult to tie up
Shield easily facilitates maintenance by communication node grouping and classifying to generate compact routing table just very effectively.Protein is mutual
Effect network can effectively indicate the interactively each other in biosystem between protein, by joining in discovery biosystem
With the protein population of same biological respinse, such biological respinse can be carried out being effectively isolated control, and found unknown
The correlation function of protein.
In recent years, due to the rise of the development of social networks and big data era, for large-scale complex network community
Detection has become one of research hotspot.With complex network scale be in explosive growth, how millions of orders of magnitude even
It is even more to become new challenge that corporations' detection is carried out on above ultra-large complex network.
In order to solve the problems, such as extensive corporations' detection, expand the overlapping corporations detection algorithm quilt of optimization thought based on part
It proposes.Local extending method is generally divided into two processes: firstly, select a seed as initial corporations, this seed can be with
It is some randomly selected node, is also possible to a complete connected subgraph, the quality of seed selection draws final corporations
Point result has very big influence.Secondly, becoming corporations to allow seed constantly to expand, it is necessary to which predefined corporations' benefit function is used
To assess the quality of corporations.Using greedy thought, constantly selection increases benefit function from the neighbor node for expanding corporations
Maximum node is added to be added to current corporations, until benefit function reaches maximum until the corporations.Algorithm is expanded in classical part
There are LFM, GCE scheduling algorithm.
Although above method fast speed, the ultra-large complex network of millions of orders of magnitude still can not be effectively solved
Problem, however it remains following problems:
(1) tradition needs to be repeated continuously the current corporations of calculating based on the Combo discovering method of expansion during expansion
The benefit function value of neighbor node, the speed of service on ultra-large complex network are unsatisfactory；
(2) although some algorithm fast speeds, large scale network community division can rapidly be obtained as a result, accurate
Property is general, and obtained overlapping corporations quality is to be improved.
Summary of the invention
The purpose of the present invention is to provide the complex network community discoveries that one kind can take into account computational efficiency and arithmetic accuracy
Method.
To achieve the above object, the invention provides the following technical scheme:
A kind of large-scale complex network community discovery method based on local neighbor information, comprising the following steps:
Step A: obtaining current network topology structure, and the maximum non-accessed node of selectance and its all neighbor nodes are made
To initialize corporations；
Step B: it deletes in initialization corporations and connects untight node；
Step C: one group of alternate node set is selected to expand initialization corporations using network local message；
Step D: being subordinate to coefficient by local message calculate node and judge to initialize whether the point in corporations should stay,
It obtains a community division and marks all nodes in the corporations to be access attribute；Step A~D is repeated until in network
There is no non-accessed nodes, export community division result.
Preferably, the method that building initializes corporations in step A are as follows: construct target network model G={ V, E }, wherein V=
{vi| i ∈ [1, | V |] it is all node v in target networkiSet, V represents the scale of whole network, i.e. interstitial content, E=
{eij(vi,vj)|vi,vj∈ V } it represents between node and there is even side, | E | represent the number for connecting side in network；
The non-maximum node of accessed node moderate is found in above-mentioned network G, i.e., it is most to connect number of edges mesh with other nodes
Node is expressed as v0=argmax (degree (vi)) and visited (v0)=0, wherein degree (vi) be node degree,
visited(vi) it is access label, visited (vi)=0 indicates node viNot visited mistake, visited (vi)=1 indicates section
Point viFor accessed node；It then initializes corporations and is expressed as InitialCommunit y={ v0∪v0All neighbor nodes }.
Preferably, the method for deletion of node described in step B are as follows: the collection for connecting untight node is combined intoWhereinFor node viWith initial corporations
The number on the company side of the node in InitialCommunity,
Preferably, described in step C expansion initialization corporations method the following steps are included:
NewAdd is initialized as society by step i: the node set newAdd and empty transition set Add that construction newly expands
Node v is removed in group InitialCommunity0Except all nodes；
Step ii: each node v in set Candidate record newAdd is usediEach neighbor node ukAnd its with it is first
All nodes connect the quantity on side in Shi Hua corporations InitialCommunity
Step iii: if the node u in set CandidatekMeetAnd
Munity, by node ukIt is added in transition set Add, wherein v is the empirical value for controlling nodes degree of overlapping；
Step iv: InitialCommunity is added in the node in set Add, enables newAdd=Add, repeats step i
~iv, until set newAdd is sky.
Preferably, judge what whether node should stay described in step D method particularly includes: for set
Node v in InitialCommunityiObtain its all neighbor node set in set InitialCommunity
NeighborTemp=neighbors (vi) ∩ InitialCommunity, wherein neighbors (vi) be and node viIn the presence of
The even set of all nodes on side traverses all node u in set NeighborTempk, coefficient score initialization will be subordinate to
It is 0, enablesThen node viThe coefficient that is subordinate to be expressed as: score (vi)=score/
degree(vi), if score (vi) < α, α are preset threshold, which is removed from InitialCommunity；Traversal
All node v in InitialCommunityi, removal is all not to meet the node v for being subordinate to coefficient requirementsi, finally obtained set
InitialCommunity is the corporations detected.
Preferably, visited (v is enabledi)=1vi∈ InitialCommunity, other nodes in traverses network, if depositing
In non-accessed node, i.e. visited (vi)=0, return step A, otherwise terminates algorithm.
Present invention also provides a kind of Computer Virus Spread control methods, are constructed according to the demand of computer sending and receiving data
Computernetwork model, obtains network community structure using the above method, carries out emphasis to corporations' core node and boundary node
Protection, the core node are to be subordinate to coefficient score (v in each corporationsi) the highest node of value, the boundary node is two
Corporations Ci、CjIntersection point, i.e. Ci∩Cj。
The advantages of large-scale complex network community discovery method and application provided by the invention based on local neighbor information
It is: only needs to check the crucial neighbours in part when expanding corporations, avoid by computing repeatedly certain using local benefit function
A node is to the contribution degree of corporations, and the method proposed will assess less node than existing method, to substantially increase place
The ability for managing large scale network, since the application utilizes the partial structurtes of node, time complexity is lower；Returned by introducing
Belong to coefficient, the node for identifying mistake during expanding corporations can be effectively removed, to improve arithmetic accuracy.
Detailed description of the invention
Fig. 1 is the large-scale complex network community discovery side based on local neighbor information provided by the embodiment of the present invention
The flow chart of method；
Fig. 2 is the large-scale complex network community discovery side based on local neighbor information provided by the embodiment of the present invention
The step A schematic diagram of method；
Fig. 3 is the large-scale complex network community discovery side based on local neighbor information provided by the embodiment of the present invention
The step B schematic diagram of method；
Fig. 4 is the large-scale complex network community discovery side based on local neighbor information provided by the embodiment of the present invention
The step C schematic diagram of method；
Fig. 5 is the large-scale complex network community discovery side based on local neighbor information provided by the embodiment of the present invention
The division result schematic diagram of method；
Fig. 6 is that the large-scale complex network community discovery method using provided in this embodiment based on local neighbor information is true
Determine the schematic diagram of key node；
Fig. 7-9 is the effect of the large-scale complex network community discovery method provided in this embodiment based on local neighbor information
Fruit explanatory diagram.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
As shown in Figure 1, the present embodiment provides a kind of large-scale complex network community discovery side based on local neighbor information
Method, comprising the following steps:
Step A: obtaining current network topology structure, and the maximum non-accessed node of selectance and its all neighbor nodes are made
To initialize corporations；Method particularly includes:
It constructs target network model G={ V, E }, wherein V={ vi| i ∈ [1, | V |] it is all node v in target networki
Set, | V | represent the scale of whole network, i.e. interstitial content, E={ eij(vi,vj)|vi,vj∈ V } it represents between node and exists
Lian Bian, | E | represent the number for connecting side in network；
With reference to Fig. 2, the non-maximum node of accessed node moderate is found in above-mentioned network G, i.e., connects number of edges with other nodes
The most node of mesh, is expressed as v0=argmax (degree (vi)) and visited (v0)=0, wherein degree (vi) it is node
Degree, visited (vi) it is access label, visited (vi)=0 indicates node viNot visited mistake, visited (viThe table of)=1
Show node viFor accessed node；Node 34 known to by analysis in Fig. 2 is degree maximum node, is saved with node 34 and its neighbours
Point then initializes corporations and is expressed as InitialCommunity={ v as initialization corporations0∪v0All neighbor nodes }.
Step B: it deletes in initialization corporations and connects untight node；The set for connecting untight node is defined asWhereinFor node viWith initial corporations
The number on the company side of the node in InitialCommunity,As shown in figure 3, will point 14,20
It is deleted from initial corporations.
Step C: one group of alternate node set is selected to expand initialization corporations using network local message；Specific method includes
Following steps:
NewAdd is initialized as society by step i: the node set newAdd and empty transition set Add that construction newly expands
Node v is removed in group InitialCommunity0Except all nodes；
Step ii: each node v in set Candidate record newAdd is usediEach neighbor node ukAnd its with it is first
All nodes connect the quantity on side in Shi Hua corporations InitialCommunity
Step iii: if the node u in set CandidatekMeetAnd
Munity, by node ukIt is added in transition set Add, wherein v is the empirical value for controlling nodes degree of overlapping；
Step iv: InitialCommunity is added in the node in set Add, enables newAdd=Add, repeats step i
~iv, until set newAdd is sky.
As shown in figure 4, point 3,25,26 is supplemented in corporations InitialCommunity.
Step D: being subordinate to coefficient by local message calculate node and judge to initialize whether the point in corporations should stay,
Judge what whether node should stay method particularly includes: for the node v in set InitialCommunityiObtain its
All neighbor node set NeighborTemp=neighbors (v in set InitialCommunityi)∩
InitialCommunity, wherein neighbors (vi) be and node viIn the presence of the set of even all nodes on side, traversal set
All node u in NeighborTempk, coefficient score will be subordinate to and be initialized as 0, enabledThen node viThe coefficient that is subordinate to be expressed as: score (vi)=score/degree
(vi), if score (vi) < α, α are preset threshold, which is removed from InitialCommunity；Traversal
All node v in InitialCommunityi, removal is all not to meet the node v for being subordinate to coefficient requirementsi, finally obtained set
InitialCommunity is the corporations detected, and corporations InitialCommunity is exported；By judgement point 3,
25,26 coefficient that is subordinate to meets the requirements, therefore all stays in corporations InitialCommunity.
To obtain all vertex ticks in corporations InitialCommunity is access attribute, enables visited (vi)=
1vi∈ InitialCommunity, other nodes in traverses network, if it exists non-accessed node, i.e. visited (vi)=0,
Return step A finds the maximum node of remaining node moderate, repeats step A~D until there is no do not access section in network
Point terminates algorithm；With reference to Fig. 5, network provided in this embodiment is divided into Liang Ge corporations, is distinguished with round and square.
The present embodiment additionally provides a kind of Computer Virus Spread control method, according to the demand structure of computer sending and receiving data
Make computernetwork model, obtain community structure using the above method, to corporations' core node and adjacent corporations' boundary node into
Row key protection, the core node are to be subordinate to coefficient score (v in corporationsi) highest node, the boundary node is two
Corporations Ci、CjIntersection point, i.e. Ci∩Cj.With reference to Fig. 6, core node is point 1 and 33 in the present embodiment, and boundary node is point 3.
It should be noted that the present embodiment only passes through the transmission controe of computer virus to network society provided in this embodiment
The application in group's discovery direction is introduced, and is not meant to that the application can be only applied to the field, for example biological virus is come
Say, after determining circulation way can according to personal relationship tectonic network model, and then using scheme provided by the present application into
Row community division is so that it is determined that key node carries out emphasis prevention and control.In addition to prevention and cure of viruses, method provided by the present application can also be used
It can be by special as carried out community division to network client similar in geographical location according to hobby in other field
Mirror image server provides service for each client cluster, improves the performance of server.In WWW, community structure is carried out
Analysis, it can be found that the page with same or like theme, can help user to search desired information faster.
Illustrate the extending method provided in this embodiment based on local neighbor information by Fig. 7 and network shown in Fig. 8
Validity, it is assumed that the network has 12 nodes, and current corporations include square nodes 1,2,3,4,5, neighbor node in Fig. 7
For hexagon node 6,7,10,11,12, interior joint 4 and 5 is the node increased newly in current corporations, and node 6,7 is current corporations
Local neighbor node, corporations extend when, it is only necessary to consider whether node 6 and 7 meets expansion condition, correspondingly, in the side of Fig. 8
When shape node corporations extend, the newly-increased node of node 6 and 7, extension only needs to consider node 8 and 9, thus the application provides
Method it is less than the node that existing method is assessed, to effectively improve the ability for handling extensive corporations, improve efficiency of algorithm.
Illustrate the large-scale complex network society provided in this embodiment based on local neighbor information by network described in Fig. 9
The accuracy of group's discovery method；For the corporations that the square nodes in Fig. 9 are constituted, node 1 is for corporations { 1,2,3,4,5 }
Be subordinate to coefficient be 0.4, through reasonable settings threshold value enable 0.4 be less than threshold value the node can be deleted from corporations, thus
Guarantee the precision of community division.
The application is also by Stamford Network data set (http://snap.stanford.edu/data/
Index.html community structure is detected on five large-scale live networks disclosed in), algorithm provided in this embodiment is compared and shows
There is the performance of the algorithm in technology, specific data are referring to table 1.
| Networks | Interstitial content | Number of edges | Average degree | Maximum value | True corporations number |
| Amazon | 0.3M | 0.9M | 5.53 | 549 | 271K |
| DBLP | 0.3M | 1M | 6.62 | 343 | 13.5K |
| YouTube | 1.1M | 3M | 5.27 | 28.8K | 16.4K |
| LiveJournal | 4M | 34.7M | 17.35 | 13.8K | 0.66M |
| Orkut | 3M | 117.2M | 76.3 | 33.3K | 6.3M |
1: five, table large-scale live network data
Wherein K indicates that thousand, M indicates million.
Table 2 and table 3 give method and the other five kinds Chong Die corporations' detection methods in the prior art of the application proposition
The comparing result of runing time and NMI precision on this five extensive live networks, NMI value describe algorithm partition result with
Difference between legitimate reading, NMI is between 0-1, and NMI value is closer to 1 illustrated divisions result closer to legitimate reading.
| Networks | SCP | LFM | LC | FOCS | NISE | This method |
| Amazon | 34.513s | 23.1m | 42.1s | 3.76s | 1.16h | 2.6s |
| DBLP | 32.525s | 24.2m | 128.7s | 2.71s | 22.4m | 3.4s |
| YouTube | 43.5m | 10.6h | 4.9h | 1.1m | 2.19h | 38.2m |
| LiveJournal | - | - | - | 12.1m | - | 38.2m |
| Orkut | - | - | - | 1.3h | - | 3.5h |
Table 2: algorithms of different computational efficiency correlation data
| Networks | SCP | LFM | LC | FOCS | NISE | This method |
| Amazon | 0.2208 | 0.2074 | 0.2451 | 0.2236 | 0.0969 | 0.2732 |
| DBLP | 0.2196 | 0.1289 | 0.1797 | 0.2145 | 0.0704 | 0.1307 |
| YouTube | 0.0426 | 0.0212 | 0.0052 | 0.0335 | - | 0.0434 |
| LiveJournal | - | - | - | 0.0307 | - | 0.6079 |
| Orkut | - | - | - | 0.0611 | - | 0.5937 |
Table 3: the NMI accuracy comparison of algorithms of different community division result and truthful data
Figure it is seen that with million ranks that increase to of network size, existing four kinds of algorithms can not be given
Division result is obtained in 24 hours, method provided by the present application only lags behind FOCS algorithm in efficiency, hence it is evident that than other four kinds
Efficiency of algorithm is higher.And from figure 3, it can be seen that although algorithm provided by the present application lags behind FOCS algorithm in efficiency, originally
Apply for that the NMI precision of the algorithm provided is significantly larger than FOCS algorithm and other algorithms, it is especially bright on ultra-large network
It is aobvious.To illustrate that the division quality of method provided by the present application has apparent advantage.
Claims (7)
1. a kind of large-scale complex network community based on local neighbor information finds method, it is characterised in that: including following step
It is rapid:
Step A: current network topology structure, the maximum non-accessed node of selectance and its all neighbor nodes are obtained as just
Shi Hua corporations；
Step B: it deletes in initialization corporations and connects untight node；
Step C: one group of alternate node set is selected to expand initialization corporations using network local message；
Step D: coefficient is subordinate to by local message calculate node and judges to initialize whether the point in corporations should stay, is obtained
One community division and to mark in the corporations all nodes be access attribute；Step A~D is repeated until not depositing in network
In non-accessed node, community division result is exported.
2. a kind of large-scale complex network community based on local neighbor information according to claim 1 finds method,
It is characterized in that: the method for building initialization corporations in step A are as follows: construct target network model G={ V, E }, wherein V={ vi|i
∈ [1, | V |] it is all node v in target networkiSet, | V | represent the scale of whole network, i.e. interstitial content, E=
{eij(vi,vj)|vi,vj∈ V } it represents between node and there is even side, | E | represent the number for connecting side in network；
The non-maximum node of accessed node moderate is found in above-mentioned network G, i.e., connects the most node of number of edges mesh with other nodes,
It is expressed as v0=argmax (degree (vi)) and visited (v0)=0, wherein degree (vi) be node degree, visited
(vi) it is access label, visited (vi)=0 indicates node viNot visited mistake, visited (vi)=1 indicates node viFor
Accessed node；It then initializes corporations and is expressed as InitialCommunit y={ v0∪v0All neighbor nodes }.
3. a kind of large-scale complex network community based on local neighbor information according to claim 2 finds method,
It is characterized in that: the method for deletion of node described in step B are as follows: the collection for connecting untight node is combined intoWhereinFor node viWith initial corporations
The number on the company side of the node in InitialCommunity,
4. a kind of large-scale complex network community based on local neighbor information according to claim 3 finds method,
Be characterized in that: described in step C expansion initialization corporations method the following steps are included:
NewAdd is initialized as corporations by step i: the node set newAdd and empty transition set Add that construction newly expands
Node v is removed in InitialCommunity0Except all nodes；
Step ii: each node v in set Candidate record newAdd is usediEach neighbor node ukAnd its with initialization
All nodes connect the quantity on side in corporations InitialCommunity
Step iii: if the node u in set CandidatekMeetAnd
By node ukIt is added in transition set Add, wherein v is the empirical value for controlling nodes degree of overlapping；
Step iv: being added InitialCommunity for the node in set Add, enable newAdd=Add, repeat step i~iv,
Until set newAdd is sky.
5. a kind of large-scale complex network community based on local neighbor information according to claim 4 finds method,
It is characterized in that: judging what whether node should stay described in step D method particularly includes: for set
Node v in InitialCommunityiObtain its all neighbor node set in set InitialCommunity
NeighborTemp=neighbors (vi) ∩ InitialCommunity, wherein neig hbors (vi) be and node viIn the presence of
The even set of all nodes on side traverses all node u in set NeighborTempk, coefficient score initialization will be subordinate to
It is 0, enablesThen node viThe coefficient that is subordinate to be expressed as: score (vi)=score/
degree(vi), if score (vi) < α, α are preset threshold, which is removed from InitialCommunity；Traversal
All node v in InitialCommunityi, removal is all not to meet the node v for being subordinate to coefficient requirementsi, output set
InitialCommu nity is the corporations detected.
6. a kind of large-scale complex network community based on local neighbor information according to claim 5 finds method,
It is characterized in that: enabling visited (vi)=1vi∈ InitialCommunity, other nodes in traverses network, is not visited if it exists
Ask node, i.e. visited (vi)=0, return step A, otherwise terminates algorithm.
7. a kind of Computer Virus Spread control method, it is characterised in that: constructed and calculated according to the demand of computer sending and receiving data
Machine network model obtains network community structure using method described in any one of claims 1-6, to corporations' core node and side
Boundary's node carries out key protection, and the core node is to be subordinate to coefficient score (v in each corporationsi) the highest node of value, it is described
Boundary node is any two corporations Ci、CjIntersection point, i.e. Ci∩Cj。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910356601.9A CN110086670A (en) | 2019-04-29 | 2019-04-29 | Large-scale complex network community discovery method and application based on local neighbor information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910356601.9A CN110086670A (en) | 2019-04-29 | 2019-04-29 | Large-scale complex network community discovery method and application based on local neighbor information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110086670A true CN110086670A (en) | 2019-08-02 |
Family
ID=67417728
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910356601.9A Pending CN110086670A (en) | 2019-04-29 | 2019-04-29 | Large-scale complex network community discovery method and application based on local neighbor information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110086670A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826003A (en) * | 2019-11-01 | 2020-02-21 | 陕西师范大学 | Illegal or harmful network information propagation control method based on edge deletion cluster |
| CN111382318A (en) * | 2020-03-14 | 2020-07-07 | 平顶山学院 | Dynamic community detection method based on information dynamics |
| CN113411691A (en) * | 2021-06-18 | 2021-09-17 | 东北电力大学 | Power optical fiber network community division method |
| CN114090835A (en) * | 2021-11-24 | 2022-02-25 | 山东大学 | Community detection method based on attribute graph information |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104933624A (en) * | 2015-06-29 | 2015-09-23 | 电子科技大学 | Community discovery method of complex network and important node discovery method of community |
| CN107222334A (en) * | 2017-05-24 | 2017-09-29 | 南京大学 | Suitable for the local Combo discovering method based on core triangle of social networks |
| US10135683B1 (en) * | 2010-12-30 | 2018-11-20 | Juniper Networks, Inc. | Dynamically generating application-layer traffic optimization protocol endpoint attributes |
| CN109376544A (en) * | 2018-09-18 | 2019-02-22 | 浙江工业大学 | A method of prevent the community structure in complex network from being excavated by depth |
| CN109617871A (en) * | 2018-12-06 | 2019-04-12 | 西安电子科技大学 | Network node immunization method based on community structure information and threshold value |
-
2019
- 2019-04-29 CN CN201910356601.9A patent/CN110086670A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10135683B1 (en) * | 2010-12-30 | 2018-11-20 | Juniper Networks, Inc. | Dynamically generating application-layer traffic optimization protocol endpoint attributes |
| CN104933624A (en) * | 2015-06-29 | 2015-09-23 | 电子科技大学 | Community discovery method of complex network and important node discovery method of community |
| CN107222334A (en) * | 2017-05-24 | 2017-09-29 | 南京大学 | Suitable for the local Combo discovering method based on core triangle of social networks |
| CN109376544A (en) * | 2018-09-18 | 2019-02-22 | 浙江工业大学 | A method of prevent the community structure in complex network from being excavated by depth |
| CN109617871A (en) * | 2018-12-06 | 2019-04-12 | 西安电子科技大学 | Network node immunization method based on community structure information and threshold value |
Non-Patent Citations (1)
| Title |
|---|
| 王从涛: "基于大规模复杂网络的重叠社团检测算法研究", 《中国优秀硕士学位论文全文数据库 基础科学辑》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826003A (en) * | 2019-11-01 | 2020-02-21 | 陕西师范大学 | Illegal or harmful network information propagation control method based on edge deletion cluster |
| CN110826003B (en) * | 2019-11-01 | 2022-05-17 | 陕西师范大学 | Illegal or harmful network information propagation control method based on edge deletion cluster |
| CN111382318A (en) * | 2020-03-14 | 2020-07-07 | 平顶山学院 | Dynamic community detection method based on information dynamics |
| CN111382318B (en) * | 2020-03-14 | 2024-02-02 | 平顶山学院 | Dynamic community detection method based on information dynamics |
| CN113411691A (en) * | 2021-06-18 | 2021-09-17 | 东北电力大学 | Power optical fiber network community division method |
| CN113411691B (en) * | 2021-06-18 | 2022-03-11 | 东北电力大学 | Power optical fiber network community division method |
| CN114090835A (en) * | 2021-11-24 | 2022-02-25 | 山东大学 | Community detection method based on attribute graph information |
| CN114090835B (en) * | 2021-11-24 | 2024-07-23 | 山东大学 | Community detection method based on attribute map information |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110086670A (en) | Large-scale complex network community discovery method and application based on local neighbor information | |
| CN109272170B (en) | A kind of traffic zone dividing system based on Louvain algorithm | |
| CN105677648B (en) | A kind of Combo discovering method and system based on label propagation algorithm | |
| CN107133704B (en) | Simulation method for large-scale crowd dynamic emergency evacuation following optimal-forgetting rule | |
| CN104933624A (en) | Community discovery method of complex network and important node discovery method of community | |
| CN106952167A (en) | A kind of catering trade good friend Lian Bian influence force prediction methods based on multiple linear regression | |
| CN109921939A (en) | The choosing method and system of key node in a kind of communication network | |
| CN105872054B (en) | A kind of mobile social networking resource discovery method based on interest hot spot | |
| CN108345661A (en) | A kind of Wi-Fi clustering methods and system based on extensive Embedding technologies | |
| CN107085594A (en) | Subgraph match method based on set similarity in big chart database | |
| Zhu et al. | Shape segmentation and applications in sensor networks | |
| CN108683448A (en) | Influence power node recognition methods suitable for air net and system | |
| CN111159577B (en) | Community dividing method and device, storage medium and electronic device | |
| CN104700311B (en) | A kind of neighborhood in community network follows community discovery method | |
| Liu et al. | Finding densest lasting subgraphs in dynamic graphs: A stochastic approach | |
| CN103345509B (en) | Obtain the level partition tree method and system of the most farthest multiple neighbours on road network | |
| CN108400889A (en) | A kind of community discovery method based on suboptimization | |
| CN110881178B (en) | Data aggregation method for Internet of things based on branch migration | |
| CN108198084A (en) | A kind of complex network is overlapped community discovery method | |
| CN105897463A (en) | Multinode power line channel response generation method based on low-voltage topological network | |
| CN112131486B (en) | E-commerce network platform user community discovery method based on graph convolution neural network | |
| CN107862073A (en) | A kind of Web community division methods based on pitch point importance and separating degree | |
| CN113312364A (en) | Smart cloud service updating method based on block chain and block chain service system | |
| CN102694725B (en) | Method for bi-directionally searching paths based on bandwidth | |
| CN103051476A (en) | Topology analysis-based network community discovery method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190802 |
|
| RJ01 | Rejection of invention patent application after publication |