CN109002523A - Node inquiring method is affected based on time contact network - Google Patents
Node inquiring method is affected based on time contact network Download PDFInfo
- Publication number
- CN109002523A CN109002523A CN201810766666.6A CN201810766666A CN109002523A CN 109002523 A CN109002523 A CN 109002523A CN 201810766666 A CN201810766666 A CN 201810766666A CN 109002523 A CN109002523 A CN 109002523A
- Authority
- CN
- China
- Prior art keywords
- node
- affected
- probability
- queue
- destination
- 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
Abstract
Node inquiring method is affected based on time contact network the present invention provides a kind of, it include: to choose both candidate nodes from being affected in node queue, and the both candidate nodes maximum node of in-degree in h jump range is focused to find out from data and obtains destination node set as destination node;The probability of spreading of destination node is calculated, the maximum candidate point of probability of spreading is therefrom selected.If candidate point is to be affected a little, the path for influencing maximum probability is then found into all paths of candidate point from influence point, it finds on the path and is pointing directly at the node to be verified of candidate point and is verified, it is to be affected node from common node of the point to all paths of the node to be verified is influenced if node to be verified is to be affected a little.So as to which under the premise of given confirmation times N, discovery as much as possible is affected node, the upper limit for breaking through N is made it possible to.
Description
Technical field
The present invention relates to technical field of information processing, and in particular, to the node that is affected based on time contact network is looked into
Inquiry method.
Background technique
Complicated network, such as transportation network, mankind's social network etc. are formd in real world between the individual of interaction,
These may be expressed as the network by putting and side is constituted.The propagation of research information in complex network, be for many years always by
One of hot issue of tracking.The development of human society is so that information spread speed is also getting faster.Some online social networks
Network, such as Facebook, Twitter will support the information of tens users to share activity daily with microblogging, pass through these social activities
Tool, people can exchange mood anywhere or anytime, share the information such as picture and video.
During information is propagated, information needs to be propagated by contact, thus can have the time interval propagated,
Information the information that be abstracted as on time diagram can be propagated to propagate, immediately between message transmission on contact network.When it is indirect
In net-fault network, the range of scatter of research information, it is significant that prediction, which is affected node, for example can help Marketing group
Specific marketing strategy is formulated to the crowd that is affected, found in infectious disease prevention and control infected crowd can prevent epidemic situation into
One step is propagated and diffusion.
Currently, being affected querying node mainly by modes such as time diagram, uncertain figures for time contact network
It carries out.But to be affected probability complex for calculate node on uncertain time diagram, accurate calculate is affected generally by node
Rate is a NP-hard problem.After whether being affected due to prediction egress, it is also necessary to further be carried out with the node true
Recognize, if be affected really, the node which is just really affected at last.Given confirmation times N, it is assumed that predict each time
Node be all successfully affected, under best-case, can only also find out N number of node being affected, the upper limit of N cannot be broken through.
Summary of the invention
For the defects in the prior art, section is affected based on time contact network the object of the present invention is to provide a kind of
Point querying method.
What is provided according to the present invention a kind of is affected node inquiring method based on time contact network, comprising:
Step 1: Candidate Set generator chooses both candidate nodes from being affected in node queue Q, and is focused to find out institute from data
It states both candidate nodes maximum node of in-degree in h jump range and obtains destination node set as destination node;
Step 2: calculating the influence probability of each destination node in the destination node set, obtaining record has destination node
Influence the Candidate Set C of probability;
Step 3: the maximum node V of select probability from the Candidate Set CkIt is verified, if the node VkIt is to be affected
Node, then by the node VkLabeled as node is affected, by the node VkIt is affected in node queue Q described in being added to,
And execute step 4;
If the node VkIt is not the node being affected, then this verifying terminates, and it is default to judge whether verifying number reaches
Upper limit value N, if do not reach preset upper limit value N, the section of unauthenticated maximum probability is selected from the Candidate Set C again
Point is verified, until find one it is new be affected node Vk, be to be affected node by the vertex ticks, be affected node
It is put into the information of the point in queue Q, executes step 4, if verifying number reaches upper limit value N in the process, executes step 7;
If verifying number is not up to upper limit value N, and all nodes are verified in the Candidate Set C, then from the data
It is focused to find out the both candidate nodes second largest node of in-degree in h jump range and obtains updated target as destination node
Node set;Return to step 2;
Step 4: if the node VkIn-degree be greater than 1, then calculate from start node V1To node VkAll Paths it is general
Rate, one paths of select probability maximum in these paths, to being pointing directly at the node VkNode VjIt is verified, if institute
State node VjIt is to be affected node, then by the node VjIt is added to described be affected in node queue Q;
Step 5: by start node V1To node VjAll paths on common node be added to and described be affected node team
It arranges in Q;
Step 6: if verifying number is not up to upper limit value N, will have been determined as the node being affected is start node, in institute
It states start node h and jumps in-degree maximum in range, and not marked node obtains updated target section as destination node
Point set returns to step 2;
Step 7: the node being affected described in output in collection queue Q.
Optionally, before the step 1 further include:
Initial effects node is determined from data set, and is added in the queue Q for being affected node collection described initial
Influence node.
Optionally, the step 2 specifically includes:
The shadow for utilizing inclusion-exclusion principle to solve all destination nodes in the destination node set by influencing probabilities calculator
Ring probability, by the destination node in the destination node set according to influence probability from big to small be sequentially recorded to Candidate Set C
In.
Optionally, the step 4 further include:
If the node VjIt is not to be affected node, then by the node VjLabeled as not being affected node.
Optionally, the step 4 further include:
If the node VkIn-degree be 1, then by start node V1To node VkAll paths on common node it is whole
It is added to described be affected in node queue Q.
Compared with prior art, the present invention have it is following the utility model has the advantages that
It is provided by the invention that node inquiring method is affected based on time contact network, it can be in a time contact net
It is as much as possible to find by the number of the node of a certain message activation in network, i.e., under the premise of given confirmation times N, to the greatest extent may be used
Discovery more than energy is affected node, makes it possible to the upper limit for breaking through N, is more affected node so as to find.
Detailed description of the invention
Upon reading the detailed description of non-limiting embodiments with reference to the following drawings, other feature of the invention,
Objects and advantages will become more apparent upon:
Fig. 1 is the logical construction schematic diagram of time contact network;
Fig. 2 is time propagation path schematic diagram;
Fig. 3 is the storage organization schematic diagram of time contact network;
Fig. 4 is the node choosing principles schematic diagram to be set out within the scope of 2 jumps with A node;
Fig. 5 is the principle signal for being affected node inquiring method for the time contact network that one embodiment of the invention provides
Figure;
Fig. 6 is the common node schematic diagram on propagation path;
Fig. 7 is that the time contact network schematic diagram for being affected node is found by 2 verifyings;
Fig. 8 is that the time contact network schematic diagram for being affected node is found by 5 verifyings.
Specific embodiment
The present invention is described in detail combined with specific embodiments below.Following embodiment will be helpful to the technology of this field
Personnel further understand the present invention, but the invention is not limited in any way.It should be pointed out that the ordinary skill of this field
For personnel, without departing from the inventive concept of the premise, several changes and improvements can also be made.These belong to the present invention
Protection scope.
Fig. 1 is the logical construction schematic diagram of time contact network, as shown in Figure 1,0 moment is in addition to node in the network
A, remaining node are not to be affected node, by node A at 1 moment gradually activate other nodes.By taking the side AB as an example, side
On (1,2,0.8) expression node A be in contact in time 1 and node B, time of contact length is 2, is propagated in current contact
The probability of message is 0.8.Fig. 2 be time propagation path schematic diagram, as shown in Fig. 2, path A → E → F, path A → B → C is can
Reach path.By taking path A → E → F as an example, which is in contact in time 2 and node E, time of contact length
It is 1, the probability to spread news in current contact is 0.9.After node E is activated, it is in contact, connects in time 3 and node F
Touching time span is 3, and the probability that message is propagated is 0.5.The inaccessible propagation path of dotted portion expression, such as A → B → C →
F, A → B → C is reachable on the path, but C → F be it is inaccessible because A → B is (1,2,0.8), illustrate node B
There is 0.8 probability to be activated by node A, and receiving message is on the time 3, thus the node that it activates oneself adjacent also will be
After time 3.B → C is (4,1,0.9), illustrates that node B is in contact in time 4 and node C after receiving the message, contact
Time span is 1, and message probability of spreading is 0.9.C → F is (3,2,0.8), indicates that node C is connect in time 3 and node F
Touching, and obtaining message according to A → B → C, node C is to illustrate that, when node C and node F are in contact, message does not have also in the time 5
It is sent to, thus is an inaccessible message propagation path.
In time contact network, from influencing node and reach not influencing on the path of node there are the probability that message forwards,
Confirm whether node is the node being really affected, and needs to carry out information authentication.For example the number of given information authentication is 10
It is secondary, then having whether the node of 10 chance acknowledging time contact networks is affected, can only be carried out each time with a node true
Recognize.Since the number of information authentication is limited, it is not affected if authenticating to the node, wastes the chance of one-time authentication,
Thus it is also larger to influence a possibility that relatively large node of probability is verified, is proved to be successful for selection.It is contacted in the entire time
The biggish node of probability is selected in network to verify, efficiency is lower, thus can be from node is influenced, and path length, which is arranged, is
H jumps the selection that node is carried out in range in h.
Specifically, as shown in figure 4, from the point within the scope of 2 jump of A point, i.e., node on the left of Fig. 4 dotted portion is exactly
From A point, path length is less than or equal to the point in 2 ranges, there is node B, node C, node D, node E, node F, node G.
Fig. 5 is the principle signal for being affected node inquiring method for the time contact network that one embodiment of the invention provides
Figure;Method in Fig. 5 may include steps of:
Step 1: the initial effects nodal information in data set being added in the queue Q for being affected node collection;
Step 2: Candidate Set generator reads the node being affected in node queue Q, finds these nodes h- in data set
Jump the maximum point set S={ V of in-degree in rangei, Vj... Vm }, S, which is sent into, to be influenced in probabilities calculator;
Step 3: influencing the approximation probability that probabilities calculator solves all the points in S using inclusion-exclusion principle, they are pressed into probability
Journal from big to small is into Candidate Set C;
Step 4: the maximum node V of select probability from Candidate Set CkIt is verified, if the node is the section being affected
Point is then marked as being affected node, and the information of the point is added in being affected node queue Q, and executes step 5;If
No, this verifying terminates, and the label not being affected is done to the node.If verifying number does not reach the upper limit of N, from candidate item
It reselects the maximum point of current probability in collection C and is verified, new be affected node V until finding onek, by the node mark
It is denoted as and is affected node, the information of the point is added in being affected node queue Q, enters step 5.In the process, if it is candidate
The point that can not verified in collection C then jumps the point that selection in-degree time is big in range in h-, constitutes new point set S, return to step
Rapid 3.If verifying number arrived the upper limit of N, 8 are entered step;
Step 5: if node VkIn-degree be greater than 1, then calculate node V1To node VkThe probability of All Paths, at these
One paths of select probability maximum, verifying are pointing directly at V in pathkThe V of nodejNode, if VjNode determination is to be affected section
Point, then by node VjLabeled as node is affected, node V is added in influencing node queue QjInformation, execute step 6;If
VjNode is not to be affected node, then is marked as not being affected node.If node VkIn-degree be 1, then by V1To VkAll
Concurrent on path, which is added, to be influenced in node queue Q;
Step 6: by V1To VjConcurrent on All Paths, which is added, to be influenced in node queue Q.
In Fig. 6, there are two paths, A → B → C → E → F → G and A → B → D → E → F → G from node A to node G,
Their common node is node B, node E, node F.Because if G is the node being affected, node F is uniquely to be directed toward its
Node, message have necessarily passed to G node from F node, and so on, all concurrents are all to be affected node.
Step 7: if verifying number does not reach the upper limit of N, Candidate Set generator is found from the current node for influencing to concentrate
It sets out, h- jumps the point set S that in-degree in range is maximum and is not labeled, and returns to step 3.
Step 8: output is affected the node in collection queue Q.
Example explanation:
It is illustrated with time contact network shown in FIG. 1.Assuming that A point is initial effects node, time that at most can confirm that
Number N is 5 times, and h value is 4.Specifically, it may include steps of:
Step 1: the initial effects node A in data set is added in the queue Q for being affected node collection;
Step 2: Candidate Set generator reads first node A being affected in node queue Q, and the 4- for finding node A is jumped
The in-degree of the maximum point set S={ F, G } of in-degree in range, the two points is all 3, and S={ F, G }, which is sent into, influences probabilities calculator
In;
Step 3: influencing the approximate propagation probability that probabilities calculator solves the midpoint S using inclusion-exclusion principle.For example F is saved
Point, reachable path are A → E → F, A → H → F.
P (F)=0.9*0.5+0.6*0.7-0.9*0.5*0.6*0.7=0.681
It is acquired also with inclusion-exclusion principle:
P (G)=1- (1-0.8*0.9*0.8*0.95) (1-0.681*0.65) (1-0.6*0.9*0.9*0.5)=
0.8089574
P (G) > P (F), thus G, F are recorded in Candidate Set.
Step 4: information authentication being carried out to G node, it is assumed that the node is really to be affected the vertex ticks, thus be
It is affected node, the information of the node is added in influencing collection queue Q.
Step 5: because the in-degree of G node is 3, being greater than 1, so finding general in all paths from node A to node G
The maximum path of rate:
P (A → B → C → D → G)=0.8*0.9*0.8*0.95=0.5472
P (A → E → F → G)=0.9*0.5*0.65=0.2925
P (A → H → F → G)=0.6*0.7*0.65=0.273
P (A → H → I → J → G)=0.6*0.9*0.9*0.5=0.243
In select probability maximum path A → B → C → D → G, the D node for being pointing directly at node G is verified, it is assumed that the section
Point is to be affected node, and being added influences in collection queue Q.
Step 6: there was only a paths from node A to node D, thus the point on entire path is all concurrent, it can be direct
Being added influences in collection queue Q.The node in queue Q has Q={ G, D, C, B, A } at this time.
Step 7: due to only demonstrating 2 times, the maximum number of times of verifying is 5, and Candidate Set generator is calculated from being affected node
{ G, D, C, B, A } sets out, and jumps the maximum point set S={ F } of in-degree in range in 4-.
Step 8: influencing the approximate propagation probability that probabilities calculator utilizes inclusion-exclusion principle solution node F.The probability of node F does not have
It changes, is still 0.681, do not need the probability for updating candidate interior joint F.
Step 9: whether verifying node F is affected.Assuming that F node is not affected, this verifying terminates, and does to node F
On the label that is not affected.
Step 10: due to can verify without node in candidate, thus it is big to look for 4- to jump in-degree time in range
Point constitutes point set S.At this point, S={ J }, because the in-degree of node J is 2.
Step 11: influencing the approximate propagation probability that probabilities calculator utilizes inclusion-exclusion principle solution node J.
P (J)=0.9*0.6*0.85+0.6*0.9*0.9-0.9*0.6*0.85*0.6*0.9*0.9=0.459+ 0.486-
0.223074=0.721926
The information of node J is put into candidate C.
Step 12: information authentication is carried out to J node, it is assumed that the node is really to be affected, thus by the node mark
It is denoted as and is affected node, and the information of node J is added in influencing collection queue Q.
Step 13: because the in-degree of J node is 2, being greater than 1, so from two paths of A → H → I → J and G → L → K → J
In, find the path of maximum probability:
P (A → H → I → J)=0.6*0.9*0.9=0.486
P (G → L → K → J)=0.9*0.6*0.85=0.459
The inode that node J is pointing directly in select probability maximum path A → H → I → J is verified, it is assumed that the node is
It is verified node, being added influences in collection queue Q.
Step 14: there was only a paths from node A to node I, thus the point on entire path is all concurrent, it can be direct
Being added influences in collection queue Q.The node in queue Q has Q={ H, I, J, G, D, C, B, A } at this time.
Specific embodiments of the present invention are described above.It is to be appreciated that the invention is not limited to above-mentioned
Particular implementation, those skilled in the art can make a variety of changes or modify within the scope of the claims, this not shadow
Ring substantive content of the invention.In the absence of conflict, the feature in embodiments herein and embodiment can any phase
Mutually combination.
Claims (5)
1. a kind of be affected node inquiring method based on time contact network characterized by comprising
Step 1: Candidate Set generator chooses both candidate nodes from being affected in node queue Q, and is focused to find out the time from data
It selects node to jump the maximum node of in-degree in range in h and obtains destination node set as destination node;
Step 2: calculating the influence probability of each destination node in the destination node set, obtaining record has destination node influence
The Candidate Set C of probability;
Step 3: the maximum node V of select probability from the Candidate Set CkIt is verified, if the node VkIt is the section being affected
Point, then by the node VkLabeled as node is affected, by the node VkIt is affected in node queue Q, and holds described in being added to
Row step 4;
If the node VkIt is not the node being affected, then this verifying terminates, and judges to verify whether number reaches the preset upper limit
Value N selects the node of unauthenticated maximum probability to carry out from the Candidate Set C again if not reaching preset upper limit value N
Verifying new is affected node V until finding onek, it is to be affected node by the vertex ticks, is being affected node queue Q
In be put into the information of the point, execute step 4, if verifying number reaches upper limit value N in the process, execute step 7;
If verifying number is not up to upper limit value N, and all nodes are verified in the Candidate Set C, then from the data set
It finds the both candidate nodes second largest node of in-degree in h jump range and obtains updated destination node as destination node
Set;Return to step 2;
Step 4: if the node VkIn-degree be greater than 1, then calculate from start node V1To node VkThe probability of All Paths,
One paths of select probability maximum in these paths, to being pointing directly at the node VkNode VjIt is verified, if the node
VjIt is to be affected node, then by the node VjIt is added to described be affected in node queue Q;
Step 5: by start node V1To node VjAll paths on common node be added to and described be affected node queue Q
In;
Step 6: if verifying number is not up to upper limit value N, the node being affected will be had been determined as start node, will described in
It is maximum that start node h jumps in-degree in range, and not marked node obtains updated destination node as destination node
Set, returns to step 2;
Step 7: the node being affected described in output in collection queue Q.
2. according to claim 1 be affected node inquiring method based on time contact network, which is characterized in that in institute
Before stating step 1 further include:
Initial effects node is determined from data set, and the initial effects are added in the queue Q for being affected node collection
Node.
3. according to claim 1 be affected node inquiring method based on time contact network, which is characterized in that described
Step 2 specifically includes:
It is general by influencing the influence that probabilities calculator solves all destination nodes in the destination node set using inclusion-exclusion principle
Rate, by the destination node in the destination node set according to influencing probability being sequentially recorded in Candidate Set C from big to small.
4. according to claim 1 be affected node inquiring method based on time contact network, which is characterized in that described
Step 4 further include:
If the node VjIt is not to be affected node, then by the node VjLabeled as not being affected node.
5. according to claim 1 or 4 be affected node inquiring method based on time contact network, which is characterized in that
The step 4 further include:
If the node VkIn-degree be 1, then by start node V1To node VkAll paths on common node all be added
It is affected in node queue Q to described.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810766666.6A CN109002523A (en) | 2018-07-12 | 2018-07-12 | Node inquiring method is affected based on time contact network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810766666.6A CN109002523A (en) | 2018-07-12 | 2018-07-12 | Node inquiring method is affected based on time contact network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109002523A true CN109002523A (en) | 2018-12-14 |
Family
ID=64599229
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810766666.6A Pending CN109002523A (en) | 2018-07-12 | 2018-07-12 | Node inquiring method is affected based on time contact network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109002523A (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103106280A (en) * | 2013-02-22 | 2013-05-15 | 浙江大学 | Uncertain space-time trajectory data range query method under road network environment |
CN104954263A (en) * | 2015-06-24 | 2015-09-30 | 广州时韵信息科技有限公司 | Method and device for searching target nodes of complex network |
CN105117421A (en) * | 2015-07-31 | 2015-12-02 | 四川长虹电器股份有限公司 | Social network analysis method based on graph structure matching |
CN106557985A (en) * | 2016-11-21 | 2017-04-05 | 云南大学 | A kind of social network information propagating source method for solving based on random walk |
CN106712995A (en) * | 2015-11-16 | 2017-05-24 | 杭州华为数字技术有限公司 | Method and device for obtaining multi-hop neighbor node |
CN107145901A (en) * | 2017-04-24 | 2017-09-08 | 武汉大学 | A kind of method for quickly querying towards rare class data in big data |
-
2018
- 2018-07-12 CN CN201810766666.6A patent/CN109002523A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103106280A (en) * | 2013-02-22 | 2013-05-15 | 浙江大学 | Uncertain space-time trajectory data range query method under road network environment |
CN104954263A (en) * | 2015-06-24 | 2015-09-30 | 广州时韵信息科技有限公司 | Method and device for searching target nodes of complex network |
CN105117421A (en) * | 2015-07-31 | 2015-12-02 | 四川长虹电器股份有限公司 | Social network analysis method based on graph structure matching |
CN106712995A (en) * | 2015-11-16 | 2017-05-24 | 杭州华为数字技术有限公司 | Method and device for obtaining multi-hop neighbor node |
CN106557985A (en) * | 2016-11-21 | 2017-04-05 | 云南大学 | A kind of social network information propagating source method for solving based on random walk |
CN107145901A (en) * | 2017-04-24 | 2017-09-08 | 武汉大学 | A kind of method for quickly querying towards rare class data in big data |
Non-Patent Citations (1)
Title |
---|
张玲玉: ""基于本体和邻居信息的知识图谱查询算法研究"", 《中国优秀硕士学位论文全文数据库工程科技辑》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Yin et al. | Lcars: a location-content-aware recommender system | |
Chen et al. | Goal-driven service composition in mobile and pervasive computing | |
CN103620585B (en) | virtual identity manager | |
CN102750375B (en) | Service and tag recommendation method based on random walk | |
CN110929162A (en) | Recommendation method and device based on interest points, computer equipment and storage medium | |
CN105262794B (en) | Content put-on method and device | |
US9619780B2 (en) | Information exchange in the social network environment | |
Zuo et al. | The power of indirect ties | |
CN104424187B (en) | A kind of method and device to client user's commending friends | |
CN108170842A (en) | Hot microblog topic source tracing method based on tripartite graph model | |
Chang et al. | Protecting mobile crowd sensing against sybil attacks using cloud based trust management system | |
Khedim et al. | A cognitive chronometry strategy associated with a revised cloud model to deal with the dishonest recommendations attacks in wireless sensor networks | |
CN104750718B (en) | The searching method and equipment of a kind of data information | |
Bosunia et al. | Efficient data delivery based on content-centric networking for Internet of Things applications | |
KR101509888B1 (en) | Method and apparatus for message spreading in social network | |
JP5896303B2 (en) | Collaborative cooperation information collection system, collaborative cooperation information collection method, and program | |
CN109002523A (en) | Node inquiring method is affected based on time contact network | |
Hu et al. | Towards context-aware mobile crowdsensing in vehicular social networks | |
Khamayseh et al. | Dynamic framework to mining internet of things for multimedia services | |
CN109284444A (en) | A kind of recommended method of good friend, device, server and storage medium | |
JP2002175405A (en) | Method for providing adaptive network service and its recording medium | |
Zhao et al. | Mobile crowd sensing service platform for social security incidents in edge computing | |
CN107092648A (en) | A kind of micro blog network interior joint influence power computational methods | |
Misra et al. | Delay Tolerant Routing and Applications | |
Wan et al. | Node placement analysis for overlay networks in IoT applications |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181214 |