CN106878174A - Internet communication node influence power based on Betweenness Centrality finds method - Google Patents
Internet communication node influence power based on Betweenness Centrality finds method Download PDFInfo
- Publication number
- CN106878174A CN106878174A CN201710168140.3A CN201710168140A CN106878174A CN 106878174 A CN106878174 A CN 106878174A CN 201710168140 A CN201710168140 A CN 201710168140A CN 106878174 A CN106878174 A CN 106878174A
- Authority
- CN
- China
- Prior art keywords
- node
- path
- influence power
- shortest path
- dlbc
- 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 18
- 238000004891 communication Methods 0.000 title claims abstract description 15
- 230000004069 differentiation Effects 0.000 claims description 17
- 230000001105 regulatory effect Effects 0.000 claims description 6
- 230000007423 decrease Effects 0.000 claims 1
- 238000005259 measurement Methods 0.000 abstract description 16
- 239000013589 supplement Substances 0.000 abstract description 5
- 238000004422 calculation algorithm Methods 0.000 description 86
- 230000000694 effects Effects 0.000 description 7
- 238000002474 experimental method Methods 0.000 description 7
- 238000011160 research Methods 0.000 description 6
- 230000000644 propagated effect Effects 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 230000007812 deficiency Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006855 networking Effects 0.000 description 2
- 230000001902 propagating effect Effects 0.000 description 2
- 238000005303 weighing Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000006806 disease prevention Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000003012 network analysis Methods 0.000 description 1
- 238000005295 random walk Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000009331 sowing Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/12—Shortest path evaluation
-
- 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
- H04L41/14—Network analysis or design
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L51/00—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
- H04L51/52—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail for supporting social networking services
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computing Systems (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Method is found the present invention is to provide a kind of Internet communication node influence power based on Betweenness Centrality.Shortest path between traversal any two points, obtains set of paths δ ij；DBC (u) ij, and Loc (u) ij are calculated each node u in path P；Each node u in traversing graph is traveled through after all paths are terminated, DLBC is calculated, and the DLBC values of all nodes are arranged from big to small；Export the array after whole DLBC sequences.The present invention has given full play to the advantage that traditional intermediary centrality is capable of node metric global impact power, and by the way that to adding decay factor and the node significance level factor, the precision in measurement in node global impact is improved.By by the number of nodes apart from node shortest path length less than or equal to L and the ratio of overall number of nodes, as node in the measurement of local influence power, as the supplement to node global impact power, foring node combined influence power.
Description
Technical field
Method is found the present invention relates to a kind of Internet communication node influence power.
Background technology
With developing rapidly for Internet technology, the concept of social networks is no longer limited only in actual life, more
It is embodied in internet social platform, internet social platform has been increasingly becoming the main channel of current Information Communication.With
Society and social networks gradually show networking tendency, " social networks revolution " (Social Network of initiation
Revolution), with " mobile revolution " (Mobile Revolution), " Net-volution " (Internet Revolution)
And the referred to as new period influences the three great revolutionary movements of the mankind.Social networks is propagated node influence power and is the discovery that one in social network analysis
Crucial research topic.Social networks propagates the purpose that node influence power finds, is, using existing social network relationships, to excavate
Go out the maximum Top-K node sets of influence power in social networks.In current social networks, the discovery of node influence power by
Step is applied in actual life and online social platform, and it is monitored in public sentiment, the marketing, disease prevention and cure, Information Communication etc.
Field has and is widely applied.
Key and influence power of the Betweenness Centrality algorithm in one node of measurement in whole social networks has more
Accurately effect, but the space that precision of its measurement is still improved, and it is also one of its shortcoming that time complexity is larger.
On the precision of measurement results, shortcoming one：Saved by the shortest path of same node different length and in shortest path
Point position difference all does same treatment；Shortcoming two：Influence power of the node in global network is laid particular stress on, node is have ignored in local area network
Influence power in network.
Traditional Betweenness Centrality is to the shortest path and node by same node different length in shortest path
Position difference is all as same treatment, but information can occur the decay of information content in the propagation at networking, so by same
The shortest path and node of node different length position difference in shortest path shall not do same treatment.
For example in Fig. 1, by node V11 from node V9 to the shortest path of node V15 be P1 (9-11-12-15), warp
Cross node V11 is P2 (2-9-11-12-15) from node V2 to the shortest path of node V15, for path P 1 and P2, P1
Path length is that the path length of 4, P2 is 5, and information is more long from start node arrival path end node distance, and information attenuation is got over
Greatly, therefore the paths of path two are having different effects to V11 node influence powers.Again in such as Fig. 1, by node V11 two
Bar shortest path P3 (2-9-11-12) and P4 (10-11-12-15), its interior joint V11 location in path P 3 are road
The 3rd node in footpath, the position in P3 is second node, has Information Communication attenuation theory to understand, identical information is reached
It is bigger in the decay of same node paths traversed more long message.
Even if influence power and importance of the traditional intermediary centrality algorithm in one node of measurement in global network have
Deficiency, but the method is still one of relatively good algorithm for weighing influence power of the node in global network at present.Due to
Traditional intermediary centrality is to have weighed an influence power for node from global angle, and it will necessarily ignore node in part
Influence power.Three degree of famous influence power principles point out that the neighbor node within three degree of node is influenceed larger by the node, so
The local influence power of node is that traditional intermediary centrality does not embody.
The content of the invention
Can solve the problem that traditional intermediary centrality algorithm in node metric influence power it is an object of the invention to provide one kind
The not enough Internet communication node influence power based on Betweenness Centrality find method.
The object of the present invention is achieved like this：
Shortest path first between traversal any two points, obtains set of paths δ ij；
Belong to δ ij for the path P between any two points, calculate each node u in path P DBC (u) ij, and
Loc (u) ij, wherein DBC (u) ij=((1- β Path [i] [j])/1- β) * ((Path [i] [j]-Pos [u] [i] [j])/Path
[i] [j]), the length of wherein Path [i] [j] delegated path P, Pos [u] [i] [j] represents node u positions in the paths, and β is
Decay factor, Loc (u)=Loc (u) ∪ node is；
Travel through after all paths are terminated each node u in traversing graph, calculate DLBC [u]=α DBC [u]+(1- α) (|
Loc (u) |/n, α are regulatory factor, and the DLBC values of all nodes are arranged from big to small；
Export the array after whole DLBC sequences.
The present invention can also include：
1st, DBC (u) is calculated by below equation：
Wherein：
δ in formulaijIt is the shortest path from node i to node j, | δij| it is the length from node i to the shortest path of node j
Degree, δijU () is by the shortest path from i to j of u, Pu(i, j) is node u in δijU the position on (), S (t) is passed for information
Broadcast decay formula, Wu(i, j) is node in δijU the importance in (), DBC (u) is after adding the shortest path differentiation factor
The influence power size that Betweenness Centrality is measured out.
2nd, Loc (u) is calculated by equation below：
Wherein δuiIt is the shortest path from node u to node, | δij| it is the length from node i to the shortest path of node j
Degree.
3rd, it is 0.7 to take regulatory factor α.
In order to solve deficiency of the traditional intermediary centrality algorithm in node metric influence power, the present invention is proposed based on office
Differentiation Betweenness Centrality algorithm (the Different Betweeness Centrality Algorithm Based of portion's factor
On Local Factors, DLBC), propose differentiation from the path to different length and the diverse location of node and think
Think, and be supplement and the correction to node global impact power with reference to node local influence masterpiece.
(1) basic thought of shortest path differentiation treatment is the shortest path according to where the length and node of shortest path
Position, combining information attenuation theory, the shortest path different to every all do the treatment of differentiation.Algorithm proposed by the present invention
Think, information is often propagated once, due to reasons such as noises, information can decay.When path length is long, the decay of information
Degree can not state the implication of prime information, present invention provide that the path that length exceedes threshold value L is measured to node influence power
Influence can be ignored, therefore shortest path of the length more than L need not be counted, and which reduce the meter of shortest path
Calculation amount, reduces the complexity of time to a certain extent.According to the analysis of above Jie's centrality algorithm advantage and disadvantage, do as follows
Improve：
1. for shortest path δij(shortest path from i to j), | δij| it is the length of shortest path, to the paths,
Learnt by Information Communication decay formula, information from i travel to node j when, information decayed to S (| δij|), shortest path
Length is different, and attenuation degree is also different, when path length is more than L, abandons the paths.
2. for node Vu, it is located at δijU the w of () (from node i to node j and by the shortest path of node u) is jumped
Position, w is smaller, then node location importance is bigger, conversely, location prominence is smaller, the present invention passes through shortest path length
The ratio of result and shortest path length behind the position of node is subtracted, as the mark for weighing the node location importance
Standard, if the ratio is bigger, position is more important.
2 points of the analysis and improvement according to more than, formula proposed by the present invention is shown in formula (1).
Wherein：
δ in formulaijIt is the shortest path from node i to node j, | δij| it is the length from node i to the shortest path of node j
Degree, δijU () is by the shortest path from i to j of u, Pu(i, j) is node u in δijPosition on (u).S (t) is passed for information
Broadcast decay formula, Wu(i, j) is node in δijU the importance in (), DBC (u) is after adding the shortest path differentiation factor
The influence power size that Betweenness Centrality is measured out.
For example in Fig. 1, δV2V15(V11) it is by node V11From node V2To node V15Shortest path, wherein one
Shortest path is P1(2-9-11-12-15), with path P1As a example by, then | δV2V15(V11) | it is 5, P11(1,15) is 3, WV11(V2,
V15) it is 3/5.
(2) in order to supplement missing of the Betweenness Centrality in node local influence power, the present invention proposes node part shadow
The power factor is rung, traditional Betweenness Centrality algorithm is combined with node cube theory, it is proposed that the office based on Betweenness Centrality
The basic thought of portion Effetiveness factor Measure Indexes local center LBC, LBC is that to certain node, shortest path length is less than
All node total numbers and the ratio of network node sum equal to 3, it follows that LBC is more big, represent the node in part
The quantity of the node that can be had influence on is bigger, and local influence power is bigger.Its computing formula is shown in formula (4).
Wherein δuiIt is the shortest path from node u to node, | δij| it is the length from node i to the shortest path of node j
Degree.
With the node V in Fig. 19As a example by, shortest path length less than or equal to 3 collection be combined into 1,2,3,4,5,7,8,10,
, therefore LBC (v 11,12,13 }9) it is 11/17.
In summary formula (1) and (4), draw the Betweenness Centrality innovatory algorithm based on differentiation and local factors
(DLBC), DLBC is in improve traditional intermediary by considering the diverse location of shortest path where different shortest paths and node
Accuracy of the disposition in node global impact dynamics amount, in compensate for tradition further through the local influence dynamics quantity algorithm for proposing
Missing of Jie's centrality in node local influence dynamics amount.DLBC computing formula are shown in formula (5).
DLBC (u)=α DBC (u)+(1- α) LBC (u) (5)
DBC (u) is shown in formula (1) in formula, and LBC (u) is shown in formula (4), and α is regulatory factor, the experiment proved that size be 0.7 when, knot
Fruit is the most accurate.
Beneficial effects of the present invention are：
The method of the present invention has given full play to the advantage that traditional intermediary centrality is capable of node metric global impact power, and
By to adding decay factor and the node significance level factor, enabling precision of the algorithm in measurement in node global impact
Improve.Secondly, the number of nodes and overall number of nodes by the way that 3 will be less than or equal to apart from node shortest path length of the invention
Ratio, as node in the measurement of local influence power, in this, as the supplement to node global impact power, forms node synthesis
Influence power.Finally on Wiki, Google+ and Wiki sample data set, tetra- algorithms of DLBC and CC, LC, BC and DC are carried out
Contrast experiment, it was demonstrated that the validity of the differentiation Betweenness Centrality algorithm based on local factors proposed by the present invention.
Brief description of the drawings
Fig. 1 is social network diagram；
Fig. 2 is the correlation that centrality algorithm and F (t) are spent on Google+ data sets；
Fig. 3 is the correlation of Betweenness Centrality algorithm and F (t) on Google+ data sets；
Fig. 4 is the correlation of each local center algorithm and F (t) on Google+ data sets；
Fig. 5 is the correlation of each close centers algorithm and F (t) on Google+ data sets；
Fig. 6 is that respectively the differentiation Betweenness Centrality algorithm based on local factors is related to F (t) on Google+ data sets
Property；
When Fig. 7 is time t=4 on Google+ data sets algorithms of different Top-K and spread scope relation；
Fig. 8 is the correlation that centrality algorithm and F (t) are spent on Wiki data sets；
Fig. 9 is the correlation of Betweenness Centrality algorithm and F (t) on Wiki data sets；
Figure 10 is the correlation of each local center algorithm and F (t) on Wiki data sets；
Figure 11 is the correlation of each close centers algorithm and F (t) on Wiki data sets；
Figure 12 is the correlation of the differentiation Betweenness Centrality algorithm and F (t) that local factors are respectively based on Wiki data sets；
When Figure 13 is time t=3 on Wiki data sets algorithms of different Top-K and spread scope relation；
Figure 14 is implementing procedure figure of the invention.
Specific embodiment
The invention will be further described for citing below in conjunction with the accompanying drawings.
(1) implementing procedure of the invention
With reference to Figure 14, the shortest path between any two points is traveled through first, set of paths δ ij are obtained, between any two points
Path P belong to δ ij, DBC (u) ij, and Loc (u) ij, wherein DBC (u) ij=are calculated to each node u in path P
((1- β Path [i] [j])/1- β) * ((Path [i] [j]-Pos [u] [i] [j])/Path [i] [j]), wherein Path [i] [j] generation
The length of table path P, Pos [u] [i] [j] represents node u positions in the paths, and β is decay factor, Loc (u)=Loc (u)
∪ node is, travel through after all paths are terminated each node u in traversing graph, calculate DLBC [u]=α DBC [u]+(1- α) (|
Loc (u) |/n, and the DLBC values of all nodes are arranged from big to small, export the array after whole DLBC sequences.
(2) evaluation method
Dependency relation between the influence power and node true impact power mainly measured out by influence power discovery algorithm
Carry out the validity of measure algorithm, the true impact power of its interior joint is the propagation by SIR propagation model analog node influence powers
Obtain.If the influence power that algorithm is measured out has obvious positive correlation with node true impact power, the influence power is illustrated
Effectively, positive correlation is more obvious, and algorithm effect is better for metric algorithm.
By taking different Top-K node sets, the propagation of analog node, under Top-K identical with other method comparisons, no
The spread scope of the Top-K node sets measured out with algorithm, the bigger algorithm of spread scope, node influence power measurement is more smart
It is accurate.Meanwhile, the present invention additionally uses the Top-10 node sets measured out of contrast algorithms of different node shadow over time
The spread scope of power is rung, the precision of algorithms of different, the bigger algorithm of same time spread scope, degree of influence are contrasted with this
Amount is more accurate.
(3) result and analysis
Using two different pieces of information collection, respectively Google+ and Wikipedia complete data sets, calculation proposed by the present invention
Attenuation parameter β is 0.8 in method, it is allowed to which shortest path L most long is 45, and information uploads sowing time in path of the length more than 45, is more than
The information that 45 node is received can be ignored.Respectively by DLBC algorithms proposed by the present invention with close to centrality
(CC), local center (LC), Betweenness Centrality (BC) and degree centrality (DC) have done comparative analysis.
Fig. 2-Fig. 6 illustrates the relation of DC, five algorithms of BC, LC, CC and DLBC and F (t), the tendency of scatter diagram from figure
Can easily show that the result of DLBC and CC is more similar, with the increase that centrality is measured, correlation of nodes is also gradually
Increase, is closer to the real propagating influence of node.DC and BC algorithms are all the global impact power or part from node
Influence power some angle is set out, so point distribution is more at random, big node and the centrality of centrality measurement are measured small
Correlation of nodes relation is not relatively obvious.LC algorithms consider the neighbours of node and neighbours' quantity of neighbours to be influenceed as node
The measurement of power, therefore the more single algorithm effect of relative DC and BC modules is preferably, but not as combined influence power algorithm
DLBC and Algorithms T-cbmplexity CC algorithms higher, the centrality and F (t) positive correlations trend of LC algorithms is not equally obvious.By
Positive correlation experiment can draw on Google+ data sets：The good relationship of DLBC algorithms, measurement precision is higher.
Fig. 7 illustrates DC, five algorithms of BC, LC, CC and DLBC in the different Top-K values of selection, during propagation time t=4
The spread scope of node influence power.It is evident that DLBC algorithms proposed by the present invention are propagated when k takes different value from Fig. 7
Scope is big compared with other algorithms.Compared with CC, due to CC algorithms consider node in network the beeline of other nodes it is total
With, node local influence power has been lacked to a certain extent, therefore in the less gap for propagating initial stage and DLBC of Top-K nodes
It is more obvious, but with the increase of Top-K number of nodes, gap is gradually decreased, from Fig. 7 as can be seen that in t=4, k<=
When 40, the gap with DLBC is more obvious, in K>Gap is gradually decreased when=40.DC algorithms are in t=4, k<When=54, calculated compared with BC
Method is good, and the initial stage is propagated in node influence power, and mainly in the part of network, DC algorithms have relatively good advantage to spread scope.LC is calculated
Method effect effect in 5 algorithms is placed in the middle, with the increase of Top-K nodes, moves closer to DC, more poor than DLBC and CC.Pass through
Fig. 7, can be apparent draw：When different Top-K values are gone, DLBC algorithms are effective with respect to other algorithms.
Be can be seen that in more complicated network by Fig. 8-Figure 12, DC and BC algorithms have become with the relation of F (t)
Very fuzzy, the scatter diagram midpoint distribution of two algorithms is more dispersed without obvious dependency relation, and measurement effect is poor.LC algorithms exist
Local center DC and BC algorithms relative with F (t) have obvious positive correlation, but positive correlation is closed compared with CC and DLBC
System is weaker.CC algorithms are more obvious with the positive relationship of F (t) when close centers are less than 800, but big in close centers
There is no obvious relation in 800 time points distribution relative distribution.DLBC is the most obvious with F (t) positive correlations in five algorithms.It is logical
Analysis more than crossing, DLBC algorithms are maximally efficient on the data set.
When Figure 13 reflects the propagation time t=3 on Wiki data sets, Top-K and the influence power of algorithms of different propagate model
The relation enclosed.As can be seen from the figure spread scopes of the DLBC than other algorithms on the Top-K nodes of 0-200 is most extensive
's.CC algorithms take second place, and are closer to LC algorithms.DC algorithms are shorter due to the propagation time, and its spread scope is better than BC, traditional
BC algorithms show worst in five algorithms, and main cause is that propagation time t is smaller, are propagated the initial stage in node influence power, and
BC propagates the later stage compared with DC advantageously from global angle node metric influence power in node influence power.
(4) conclusion：
The research direction that current social networks node influence power finds is to improve the precision of arithmetic result.Although right at present
Social networks propagates the research that node influence power finds to be had a lot, but can either while guarantee arithmetic result precision,
There can also be the achievement in research of preferable execution efficiency relatively fewer.Proposed based on differentiation for traditional Betweenness Centrality and
2 points of improvement of local factors, substantially increase the result precision of traditional intermediary centrality algorithm.By in Wiki, Google
+ and Wiki sample data sets on carry out the experiment such as correlation, find that algorithms are contrasted with other conventional node influence powers, verify
Betweenness Centrality node influence power based on differentiation and local factors proposed by the present invention finds the validity of algorithm.This hair
Bright main contents are as follows：
The present invention is based on traditional intermediary centrality algorithm, it is proposed that the differentiation Betweenness Centrality algorithm based on local factors
(DLBC).The algorithm has given full play to the advantage that traditional intermediary centrality is capable of node metric global impact power, and by right
Decay factor and the node significance level factor are added, precision of the algorithm in measurement in node global impact is improved.
Secondly, the ratio by the way that 3 number of nodes and overall number of nodes will be less than or equal to apart from node shortest path length of the invention,
As node in the measurement of local influence power, in this, as the supplement to node global impact power, node combined influence is formd
Power.Finally on Wiki, Google+ and Wiki sample data set, it is right that tetra- algorithms of DLBC and CC, LC, BC and DC have been carried out
Than experiment, it was demonstrated that the validity of the differentiation Betweenness Centrality algorithm based on local factors proposed by the present invention.
Although it is effective that the present invention demonstrates algorithm proposed by the present invention by different contrast experiments, do not wait general
Rate random walk sampling of data algorithm and the differentiation Betweenness Centrality algorithm based on local factors also have deficiency and need
Further research and improvement.Particular problem is as follows：
Although the differentiation Betweenness Centrality algorithm based on local factors all does in the global and local influence power of node
Consider, improve the accuracy of influence power measurement, execution efficiency is reduced simultaneously also by experiment is carried out on sample, but its
The too high problem of algorithm time complexity in itself is not still solved.In research and work after, can use for reference
Brande is based on the algorithm of the theoretical quick calculate node Betweenness Centrality for proposing of signal source shortest path, from algorithm drop in itself
Low time complexity.
Claims (5)
1. a kind of Internet communication node influence power based on Betweenness Centrality finds method, it is characterized in that：
Shortest path first between traversal any two points, obtains set of paths δ ij；
Belong to δ ij for the path P between any two points, calculate each node u in path P DBC (u) ij, and Loc
(u) ij, wherein DBC (u) ij=((1- β Path [i] [j])/1- β) * ((Path [i] [j]-Pos [u] [i] [j])/Path [i]
[j]), the length of wherein Path [i] [j] delegated path P, Pos [u] [i] [j] represents node u positions in the paths, and β is to decline
Subtracting coefficient, Loc (u)=Loc (u) ∪ node is；
Travel through after all paths are terminated each node u in traversing graph, calculate DLBC [u]=α DBC [u]+(1- α) (| Loc
(u) |/n, α are regulatory factor, and the DLBC values of all nodes are arranged from big to small；
Export the array after whole DLBC sequences.
2. the Internet communication node influence power based on Betweenness Centrality according to claim 1 finds method, it is characterized in that
DBC (u) is calculated by below equation：
Wherein：
δ in formulaijIt is the shortest path from node i to node j, | δij| it is the length from node i to the shortest path of node j, δij
U () is by the shortest path from i to j of u, Pu(i, j) is node u in δijU the position on (), S (t) decays for Information Communication
Formula, Wu(i, j) is node in δijImportance in (u), DBC (u) be add the shortest path differentiation factor after intermediary in
The influence power size that disposition is measured out.
3. the Internet communication node influence power based on Betweenness Centrality according to claim 1 and 2 finds method, its feature
It is that Loc (u) is calculated by equation below：
Wherein δuiIt is the shortest path from node u to node, | δij| it is the length from node i to the shortest path of node j.
4. the Internet communication node influence power based on Betweenness Centrality according to claim 1 and 2 finds method, its feature
It is to take regulatory factor α for 0.7.
5. the Internet communication node influence power based on Betweenness Centrality according to claim 3 finds method, it is characterized in that
It is 0.7 to take regulatory factor α.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710168140.3A CN106878174A (en) | 2017-03-21 | 2017-03-21 | Internet communication node influence power based on Betweenness Centrality finds method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710168140.3A CN106878174A (en) | 2017-03-21 | 2017-03-21 | Internet communication node influence power based on Betweenness Centrality finds method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106878174A true CN106878174A (en) | 2017-06-20 |
Family
ID=59171642
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710168140.3A Pending CN106878174A (en) | 2017-03-21 | 2017-03-21 | Internet communication node influence power based on Betweenness Centrality finds method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106878174A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108092818A (en) * | 2017-12-26 | 2018-05-29 | 北京理工大学 | A kind of intelligent agent method that can promote node in dynamic network terminal impacts power |
| CN109919459A (en) * | 2019-02-21 | 2019-06-21 | 武汉大学 | Method for measuring influence among social network objects |
| CN112291827A (en) * | 2020-10-29 | 2021-01-29 | 王程 | Social attribute driven delay tolerant network route improvement algorithm |
| CN114090752A (en) * | 2021-11-17 | 2022-02-25 | 中国建设银行股份有限公司 | Problem thread mining method, device, computer equipment and medium |
-
2017
- 2017-03-21 CN CN201710168140.3A patent/CN106878174A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108092818A (en) * | 2017-12-26 | 2018-05-29 | 北京理工大学 | A kind of intelligent agent method that can promote node in dynamic network terminal impacts power |
| CN108092818B (en) * | 2017-12-26 | 2020-06-05 | 北京理工大学 | Intelligent agent method capable of improving influence of node on dynamic network terminal |
| CN109919459A (en) * | 2019-02-21 | 2019-06-21 | 武汉大学 | Method for measuring influence among social network objects |
| CN109919459B (en) * | 2019-02-21 | 2022-05-13 | 武汉大学 | Method for measuring influence among social network objects |
| CN112291827A (en) * | 2020-10-29 | 2021-01-29 | 王程 | Social attribute driven delay tolerant network route improvement algorithm |
| CN114090752A (en) * | 2021-11-17 | 2022-02-25 | 中国建设银行股份有限公司 | Problem thread mining method, device, computer equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106878174A (en) | Internet communication node influence power based on Betweenness Centrality finds method | |
| Lim et al. | Mascot: Memory-efficient and accurate sampling for counting local triangles in graph streams | |
| Yao et al. | The tug of war between local government and enterprises in reducing China's carbon dioxide emissions intensity | |
| Cădariu et al. | Fixed Points and Generalized Hyers‐Ulam Stability | |
| CN102855185B (en) | Pair-wise test method based on priority | |
| Chen et al. | Impact of temporary fork on the evolution of mining pools in blockchain networks: An evolutionary game analysis | |
| CN110213164A (en) | A kind of method and device of the identification network key disseminator based on topology information fusion | |
| Li et al. | Classical and quantum correlative capacities of quantum systems | |
| CN110110529B (en) | Software network key node mining method based on complex network | |
| CN104008163A (en) | Trust based social network maximum influence node calculation method | |
| CN103344971A (en) | Optimization method suitable for GNSS real-time data processing | |
| Velichkov et al. | The additive differential probability of ARX | |
| CN105869054A (en) | Three-degree influence principle-based social network influence maximizing method | |
| Xue et al. | Evaluating and analyzing the operation efficiency of urban rail transit systems in china using an integrated approach of DEA model, Malmquist productivity index, and tobit regression model | |
| Li et al. | Dynamic trustworthiness overlapping community discovery in mobile internet of things | |
| Gupta et al. | Sum rules for the masses of the 2P-and higher levels of charmonium | |
| Geba et al. | Ill-posedness results for generalized Boussinesq equations | |
| Wu | An ecological environmental carrying capacity estimation of tourist attractions based on structural equation | |
| Ghareeb | Weak forms of continuity in I-double gradation fuzzy topological spaces | |
| CN103618702A (en) | Social network information transmission situation awareness method and system | |
| Bayraktar et al. | Weak reflection principle for Lévy processes | |
| Arash | Meson structure functions in the valon model | |
| CN105489014B (en) | A kind of compensation method based on principal component analysis transport information | |
| Guicciardi | Scalability of Encointer-a Proof-Of-Personhood Cryptocurrency | |
| Jiao et al. | Multiobjective traffic signal control model for intersection based on dynamic turning movements estimation |
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: 20170620 |
|
| RJ01 | Rejection of invention patent application after publication |