CN107529656A - The division methods and server of a kind of myspace - Google Patents
The division methods and server of a kind of myspace Download PDFInfo
- Publication number
- CN107529656A CN107529656A CN201610458069.8A CN201610458069A CN107529656A CN 107529656 A CN107529656 A CN 107529656A CN 201610458069 A CN201610458069 A CN 201610458069A CN 107529656 A CN107529656 A CN 107529656A
- Authority
- CN
- China
- Prior art keywords
- label
- node
- neighbors
- weights
- social networks
- 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.)
- Granted
Links
Landscapes
- Information Transfer Between Computers (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention discloses a kind of division methods of myspace, including:Social networks model is built based on the social networks in social networks, assigns weights to the side being connected with each other the node in the social networks model, each node carries label;The label for updating node all in the social networks model is iterated according to the weights on the side between the label of each node and the node;When iteration update times meet predetermined threshold value, user's mark corresponding to the node of same label is carried in the social networks model after renewal is handled and is divided into same community.The embodiment of the invention also discloses a kind of server.Using the present invention, the stability of the label of server update node can be improved, so as to improve the stability of server division community.
Description
Technical field
The present invention relates to field of social network, more particularly to a kind of division methods of myspace and service
Device.
Background technology
With the fast development of internet, the social networks of user's composition in network is no longer only single use
Family and the relation of unique user, but single pair is more and the relation of multi-to-multi.At this point it is possible to according to user's
Social networks in social networks carry out community's division to the user in social networks, so as to enter to user
Row more effectively management and maintenance, and community's push targetedly business can be directed to.
At present, it is to carry out community using LPA (label propagation algorithm) algorithm that community's division of social networks is mostly
Excavate division.But when carrying out tag update using the algorithm, only saved according to the neighbour of present node
Point label obtains optimal label and carries out tag update, and the neighbors label as obtained present node occurs most multiple
Label for optimal label carry out tag update, this allows to that multiple optimal labels occur.Work as appearance
During multiple optimal labels, simply random one of those optimal label of selection is entered row label by the algorithm
Renewal, there is randomness in this in selection course due to label so that and unstability be present in algorithm, therefore,
Identical result can not be obtained when carrying out and repeatedly dividing using the algorithm, so as to which divided community can not be ensured
Stability.
The content of the invention
Technical problem to be solved of the embodiment of the present invention is, there is provided a kind of division side of myspace
Method and server.The stability of the label of server update node can be improved, so as to improve server division society
The stability in area.
In order to solve the above-mentioned technical problem, the embodiments of the invention provide the division methods of kind of myspace,
Including:
Social networks model is built based on the social networks in social networks, in the social networks model
The side being connected with each other between node assigns weights, wherein, the social networks identifies including at least one user,
The social networks model includes at least one node, at least one user's mark with it is described at least one
Node corresponds, and each node carries label;
The renewal society is iterated according to the weights on the side between the label of each node and the node
Hand over the label of node all in network model;
When iteration update times meet predetermined threshold value, phase is carried in the social networks model after renewal is handled
Same community is divided into user's mark corresponding to the node of label.
Correspondingly, the embodiment of the present invention additionally provides a kind of server, including:
Construction unit, for building social networks model based on the social networks in social networks, to the society
The side being connected with each other between the node in network model is handed over to assign weights, wherein, the social networks is included at least
One user's mark, the social networks model include at least one node, at least one user's mark
Corresponded with least one node, each node carries label;
Iteration updating block, the weights for the side between the label according to each node and the node
It is iterated the label for updating node all in the social networks model;
Divide neighbourhood unit, for when iteration update times meet predetermined threshold value, the society after renewal is handled
Hand over user corresponding to carrying the node of same label in network model to identify and be divided into same community.
In embodiments of the present invention, social networks model is built based on the social networks in social networks, to institute
To state the side being connected with each other between the node in social networks model and assign weights, each node carries label, according to
The weights on the side between the label and node of each node, which are iterated in the renewal social networks model, to be owned
Node label, when iteration update times meet predetermined threshold value, the social networks mould after renewal is handled
User's mark corresponding to the node of same label is carried in type and is divided into same community.This effectively improves service
The stability of the label of device more new node, so as to effectively improve the stability of server division community.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to implementing
The required accompanying drawing used is briefly described in example or description of the prior art, it should be apparent that, describe below
In accompanying drawing be only some embodiments of the present invention, for those of ordinary skill in the art, do not paying
On the premise of going out creative work, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 is a kind of a kind of flow signal of the division methods of myspace provided in an embodiment of the present invention
Figure;
Fig. 2 is a kind of part of nodes topological diagram in a kind of social networks model provided in an embodiment of the present invention;
Fig. 3 is that a kind of another flow of the division methods of myspace provided in an embodiment of the present invention is shown
It is intended to;
Fig. 4 is another part of nodes topological diagram in a kind of social networks model provided in an embodiment of the present invention;
Fig. 5 is that a kind of another flow of the division methods of myspace provided in an embodiment of the present invention is shown
It is intended to;
Fig. 6 is another part of nodes topological diagram in a kind of social networks model provided in an embodiment of the present invention;
Fig. 7 is a kind of structural representation of server provided in an embodiment of the present invention;
Fig. 8 is the structural representation of another server provided in an embodiment of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear
Chu, it is fully described by, it is clear that described embodiment is only part of the embodiment of the present invention, rather than
Whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making creation
Property work under the premise of the every other embodiment that is obtained, belong to the scope of protection of the invention.
A kind of division methods of myspace provided in an embodiment of the present invention can apply to server progress
Divide myspace scene, wherein, server can including computer etc. processing equipment.
Below in conjunction with accompanying drawing 1- accompanying drawings 6, to the division methods of myspace provided in an embodiment of the present invention
Describe in detail, wherein, the executive agent of embodiment is server.
It is that a kind of division methods method of myspace provided in an embodiment of the present invention is a kind of real referring to Fig. 1
Apply a schematic flow sheet.A kind of division methods of myspace of the embodiment of the present invention comprise the following steps:
S100, social networks model is built based on the social networks in social networks, to the social networks mould
The side being connected with each other between node in type assigns weights.
In embodiments of the present invention, social networks can be that the various social tools for user's exchange are built
Social networks, social networks may include the social networks of at least one user's mark and each user mark.Its
In, user's mark can be user account, and social networks can be that each user identifies foundation in social networks
Relation chain composition, such as each user mark friend relation chain composition social networks social networks.
In embodiments of the present invention, server can be closed based on the social of social networks in specified social tool
System's structure social networks model, wherein, social networks model can be a kind of topology being made up of node and side
Figure or other complex networks.Specifically, server can open up user's identity map in social networks
The node flutterred in figure, and node identification is assigned to the node in topological diagram according to user's mark, so as to which user marks
Know and corresponded with node, the social networks that each user in social networks identifies are mapped as topology by server
Side in figure, so that with connection side between node corresponding to the mark of the user with friend relation.
In embodiments of the present invention, server can also be being connected with each other each node in social networks model
Side assign weights, wherein, server can be according to the relationship strength being connected with each other between each node to connection
Side assigns different weights.Further, server and to the unique label of each node distribution.
In a particular application, the social networks model of server construction can be:Network G=(V, E, F),
Wherein V is the set of node, and E is the set on side, and F is the set of the weights on side.Wherein social networks model
In the topological diagram of part of nodes can be referring specifically to Fig. 2 topological diagram, the A in node A (A) can be
The identifier of node, (A) can be for the labels of node, the side connected between node A (A) and B (B)
Weights be 0.9.
S101, renewal is iterated according to the weights on the side between the label of each node and the node
The label of all nodes in the social networks model.
In embodiments of the present invention, server is carrying out often taking turns section all in iteration renewal social networks model
During the label of point, server iteration can update social networks mould in the way of the label of one node of following renewal
The label of all nodes in type:Server can be in random alignment social networks model node, obtain with
The order of machine arrangement, server can determine that need to currently be updated label works as prosthomere by the order of random alignment
Point, neighbors is determined according to present node, between the label and neighbors and present node that obtain neighbors
The weights on side are carried out to obtain target labels to update the label of present node.
S102, when iteration update times meet predetermined threshold value, in the social networks model after renewal is handled
Carry user's mark corresponding to the node of same label and be divided into same community.
In embodiments of the present invention, identical mark is carried in the social networks model after server can be handled renewal
User's mark corresponding to the node of label is divided into same community, and the user in the community can be identified and carried out
It is uniformly processed, such as launches identical advertisement.
In embodiments of the present invention, social networks model is built based on the social networks in social networks, to institute
To state the side being connected with each other between the node in social networks model and assign weights, each node carries label, according to
The weights on the side between the label and node of each node, which are iterated in the renewal social networks model, to be owned
Node label, when iteration update times meet predetermined threshold value, the social networks mould after renewal is handled
User's mark corresponding to the node of same label is carried in type and is divided into same community.This effectively improves service
The stability of the label of device more new node, so as to effectively improve the stability of server division community.
It is that a kind of division methods another kind of myspace provided in an embodiment of the present invention is implemented referring to Fig. 3
Example schematic flow sheet.A kind of division methods of myspace of the embodiment of the present invention comprise the following steps:
S200, social networks model is built based on the social networks in social networks, to the social networks mould
The side being connected with each other between node in type assigns weights.
In embodiments of the present invention, server can calculate the node for the node being connected with each other in social networks model
Between cohesion, imparting weights are carried out to the side the node of the node of interconnection according to cohesion, wherein,
Cohesion positive correlation between the weights on the side between the node of the node of interconnection and its node, weights can also
For its corresponding cohesion.
In embodiments of the present invention, social networks may include user mark interactive information, as chat message,
Comment information and information etc. is thumbed up, server calculates the node for the node being connected with each other in social networks model
Between cohesion can be specifically:The corresponding relation that server identifies according to node and user, it is determined that mutually interconnecting
Two users' marks, server are obtained in social networks between two users corresponding to the node difference connect
Interactive information, the cohesion between two user's marks is calculated according to interactive information, two users are identified it
Between cohesion be set to be connected with each other node node between cohesion.Therefore, cohesion can be used for retouching
The interactive frequent degree of two user's marks is stated, cohesion is higher, and the interaction of two user's marks is more frequent,
So that interactive frequently user mark can be divided into a community by server, thrown with improving to community
Put the accuracy of miscellaneous service.
S201, the node in social networks model described in random alignment.
In embodiments of the present invention, server can carry out random alignment to all nodes in social networks model,
The order of random alignment is obtained, so that server carries out more new node one by one according to the order of the random alignment
Label.
S202, iteration perform the order according to random alignment, start successively according between present node and neighbors
The weights on side and the label of the neighbors update the label of the present node, and count iteration more
New number.
In embodiments of the present invention, server can determine that needs are updated successively according to the order of random alignment
Label present node, and according to the weights on the side between present node and neighbors and the neighbors
Label updates the label of the present node.
In embodiments of the present invention, server is according to the weights on the side between present node and neighbors and described
The label of neighbors can be specifically to update the label of the present node:Determined by the label of neighbors every
The neighbors of class label, wherein, the neighbors per class label carries identical label, works as prosthomere according to described
Point and the weights on the side between the neighbors, calculate neighbors per class label weights and, from described per class
The weights of the neighbors of label and it is middle obtain maximum weights and, determine maximum weights and corresponding
Label, and the label for updating the present node is the maximum weights and corresponding label.
In a particular application, reference can be made to Fig. 4 topological diagram, as seen from Figure 4, it is assumed that present node B, then
Neighbors includes node A, node C, node D and node E.Node A label is A, node C's
Label is A, and node D label is D, and node E label is D, and therefore, server can be by neighbors
Label node is classified, wherein, the neighbors with identical label is divided into one kind, then type
Include node A and node C for the neighbors of label A, the neighbors that type is label D include node D and
E.After server determines the neighbors per class label, server can calculate the weights of the neighbors of every class label
With as shown in Figure 2, the weights on the side between node A and current node B are 0.9, node C and work as prosthomere
The weights on the side between point B are 0.6, and the weights on the side between node D and current node B are 0.4, node E with
The weights on the side between current node B are 0.4, then tag types are the weights of the neighbors of label A and are
0.9+0.6=1.5, tag types are the weights of label D neighbors and are 0.4+0.4=0.8, therefore, service
Device can obtain the weights of maximum and be 1.5, be label A so as to obtain the label corresponding to weights and 1.5,
And by node B tag update it is label A.
In embodiments of the present invention, server can also obtain label to update present node according to below equation
Label, wherein, formula is as follows:
In the formula, wijRepresent present node i and neighbors j weights, δ (Lj)=1, and δ (Lj) represent current
The neighbors j of node i label, the neighbors for belonging to same label are added with the weights on the side of node i,
Find out weights and that maximum label passes to i, i.e. node i label L as new labeli.With Fig. 4
Exemplified by, in more new node B label, the label of neighbors is A and D, the power of the neighbors of label A
Value and be 0.9+0.6=1.5, weights of label D neighbors and be 0.8, so selecting label A as B
Label be updated.
In embodiments of the present invention, because weights can be the cohesion between node, therefore, server can be with
Tend to the label of the node higher with present node cohesion to pass to present node and be updated.
In embodiments of the present invention, when server according to upper type to all nodes in social networks model
Label be iterated renewal, and server is updating node all in a wheel social networks model
During label, server will carry out counting current iteration update times, when iteration update times are more than predetermined threshold value
When, server can determine that iteration update times meet predetermined threshold value, wherein, predetermined threshold value can be by user voluntarily
It is configured, does not limit herein.Further, when server detects section all in social networks model
When the label of point all no longer changes, server can determine that iteration update times meet predetermined threshold value.
S203, when iteration update times meet predetermined threshold value, in the social networks model after renewal is handled
Carry user's mark corresponding to the node of same label and be divided into same community.
In embodiments of the present invention, step S200 and step S203 embodiment can be found in above-mentioned reality
The embodiment of the step S100 and step S102 in example are applied, will not be described here.
In embodiments of the present invention, social networks model is built based on the social networks in social networks, to institute
To state the side being connected with each other between the node in social networks model and assign weights, each node carries label, according to
The weights on the side between the label and node of each node, which are iterated in the renewal social networks model, to be owned
Node label, when iteration update times meet predetermined threshold value, the social networks mould after renewal is handled
User's mark corresponding to the node of same label is carried in type and is divided into same community.This effectively improves service
The stability of the label of device more new node, so as to effectively improve the stability of server division community.
It is a kind of another implementation of the division methods of myspace provided in an embodiment of the present invention referring to Fig. 5
Example schematic flow sheet.A kind of division methods of myspace of the embodiment of the present invention comprise the following steps:
S300, social networks model is built based on the social networks in social networks, to the social networks mould
The side being connected with each other between node in type assigns weights.
S301, the node in social networks model described in random alignment.
S302, according to the order of random alignment, start successively according between the present node and the neighbors
The weights on side, the geodesic distance of the label of the neighbors and the neighbors described work as prosthomere to update
The label of point.
In embodiments of the present invention, geodesic distance is two sections in the social networks model before iteration renewal
Beeline between point.The geodesic distance of neighbors can be the ancestor node corresponding to the label of neighbors
Geodesic distance between neighbors, wherein, ancestor node can be the social networks mould before iteration renewal
In type, the node corresponding to the label of neighbors.
In embodiments of the present invention, server according to the weights on the side between the present node and the neighbors,
The geodesic distance of the label of the neighbors and the neighbors includes to update the label of the present node:
Server obtains the geodesic distance of neighbors, and the neighbors per class label is determined by the label of neighbors, wherein,
It is identical per the label entrained by the neighbors of class label, calculate the weights point on the side between present node and neighbors
Ratio not between the geodesic distance of neighbors, so as to obtain the ratio sum of the neighbors of every class label,
The ratio sum of maximum is obtained from the ratio sum of the neighbors of every class label, it is determined that maximum ratio sum
Corresponding label, and the label for updating present node is maximum weights and corresponding label.Wherein,
Specifically, it can obtain as follows to calculate per the ratio sum of the neighbors of class label:
In the formula, dijIt is the neighbors j of node i geodesic distance, wijRepresent present node i and neighbors
J weights, δ (Lj)=1, and δ (Lj) represent present node i neighbors j label.
In embodiments of the present invention, terminal, according to the social networks model of structure, can be built before iteration renewal
The corresponding relation of each node and label is found, and the shortest path obtained between each node enters as geodesic distance
Row storage, wherein, server can be deposited according to the geodesic distance generation list between each node of acquisition
Storage.Therefore, the geodesic distance of server acquisition neighbors can be specifically:The label of the neighbors is obtained,
According to each node and the corresponding relation of label, ancestor node corresponding to the label of the neighbors is obtained,
The mesh between the neighbors and the ancestor node is obtained in geodesic distance list between each node of storage
Geodesic distance is marked, the target geodesic distance got is set to the geodesic distance of the neighbors.
In a particular application, reference can be made to Fig. 6 provide topological diagram, as seen from Figure 6, it is assumed that present node B,
Then neighbors includes node A, node C, node D and node E.Node A label is F, node C
Label be F, node D label is D, and node E label is D, and therefore, server can be from storage
Geodesic distance list in get ancestor node corresponding to label F and node A geodesic distance be 3, mark
The geodesic distance for signing ancestor node corresponding to F and node C is 3, ancestor node and node corresponding to label D
D geodesic distance is 0, and ancestor node corresponding to label D and node E geodesic distance are 1, wherein, it is former
Beginning node is the node corresponding to label before renewal is iterated.Further, server can be saved by neighbour
The label of point is classified to node, wherein, the neighbors with identical label is divided into one kind, then class
Type includes node A and node C for the neighbors of label F, and type is that label D neighbors includes node D
And E.After server determines the neighbors per class label, server can calculate adding for the neighbors of every class label
Weigh weights and, as shown in Figure 5, the weights on the side between node A and current node B are 0.9, node C with
The weights on the side between current node B are 0.6, and the weights on the side between node D and current node B are 0.4, section
The weights on the side between point E and current node B are 0.4, then tag types are the ratio of the neighbors of label A
Sum and F=0.9* (1/4)+0.6* (1/4)=0.375, tag types be label D neighbors ratio it
With for D=0.4* (1/1)+0.4* (1/2)=0.6, therefore, the ratio sum that server can obtain maximum is 0.6,
It is label D so as to obtain the label corresponding to ratio sum 0.6, and is label by node B tag update
A。
In embodiments of the present invention, server can also obtain label to update present node according to below equation
Label, wherein, formula is as follows:
Due to dijIt is the geodesic distance in the neighbors j of node i label source, distance is more remote, and label is selected
Chance it is smaller.By taking Fig. 6 as an example, when node B neighbours have the label of label F and D, more new node B,
The result that label F calculates according to weights and geodesic distance is F=0.9* (1/4)+0.6* (1/4)=0.375, and
D result of calculation is D=0.4* (1/1)+0.4* (1/2)=0.6, so B label is updated to D, such as
If fruit only considers weights, B can be updated to F, be divided into one from its community farther out.Therefore,
Present invention introduces geodesic distance, it can make it that the node of same label will not be apart from far, so as to avoid result in
There is excessive community when dividing community in server.
In embodiments of the present invention, when server according to upper type to all nodes in social networks model
Label be iterated renewal, and server is updating node all in a wheel social networks model
During label, server will carry out counting current iteration update times, when iteration update times are more than predetermined threshold value
When, server can determine that iteration update times meet predetermined threshold value, wherein, predetermined threshold value can be by user voluntarily
It is configured, does not limit herein.Further, when server detects section all in social networks model
When the label of point all no longer changes, server can determine that iteration update times meet predetermined threshold value.
S303, when iteration update times meet predetermined threshold value, in the social networks model after renewal is handled
Carry user's mark corresponding to the node of same label and be divided into same community.
In embodiments of the present invention, step S300 and step S303 embodiment can be found in above-mentioned reality
The embodiment of the step S200 and step S203 in example are applied, will not be described here.
In embodiments of the present invention, social networks model is built based on the social networks in social networks, to institute
To state the side being connected with each other between the node in social networks model and assign weights, each node carries label, according to
The weights on the side between the label and node of each node, which are iterated in the renewal social networks model, to be owned
Node label, when iteration update times meet predetermined threshold value, the social networks mould after renewal is handled
User's mark corresponding to the node of same label is carried in type and is divided into same community.This effectively improves service
The stability of the label of device more new node, so as to effectively improve the stability of server division community.
Below in conjunction with accompanying drawing 7, server provided in an embodiment of the present invention is described in detail.Need to illustrate
, the server shown in accompanying drawing 7, the method for performing Fig. 1-embodiment illustrated in fig. 6 of the present invention, it is
It is easy to illustrate, illustrate only the part related to the embodiment of the present invention, particular technique details does not disclose,
It refer to the embodiment shown in Fig. 1-Fig. 6 of the present invention.
Fig. 7 is referred to, for the embodiments of the invention provide a kind of structural representation of server.As shown in fig. 7,
The server of the embodiment of the present invention can include:
Construction unit, for building social networks model based on the social networks in social networks, to the society
The side being connected with each other between the node in network model is handed over to assign weights, each node carries label.
Iteration updating block, the weights for the side between the label according to each node and the node
It is iterated the label for updating node all in the social networks model.
Divide neighbourhood unit, for when iteration update times meet predetermined threshold value, the society after renewal is handled
Hand over user corresponding to carrying the node of same label in network model to identify and be divided into same community.
In embodiments of the present invention, social networks can be that the various social tools for user's exchange are built
Social networks, social networks may include the social networks of at least one user's mark and each user mark.Its
In, user's mark can be user account, and social networks can be that each user identifies foundation in social networks
Relation chain composition, such as each user mark friend relation chain composition social networks social networks.
In embodiments of the present invention, construction unit can the social activity based on social networks in specified social tool
Relation builds social networks model, wherein, social networks model can a kind of be opened up by what node and side formed
Flutter figure or other complex networks.Specifically, construction unit can be by user's identity map in social networks
For the node in topological diagram, and node identification is assigned to the node in topological diagram according to user's mark, so as to use
Family identifies to be corresponded with node, and the social networks that construction unit identifies each user in social networks map
For the side in topological diagram, so as between node corresponding to user's mark with friend relation with connecting side.
In embodiments of the present invention, construction unit can also be the side that is connected with each other each node in topological diagram
Weights are assigned, wherein, the side that construction unit can be according to the relationship strength being connected with each other between each node to connection
Assign different weights.Further, construction unit and to the unique label of each node distribution.
The social networks model of construction unit structure can be in a particular application:Network G=(V, E, F),
Wherein V is the set of node, and E is the set on side, and F is the set of the weights on side.Wherein social networks model
In the topological diagram of part of nodes can be referring specifically to Fig. 2 topological diagram, the A in node A (A) can be
The identifier of node, (A) can be for the labels of node, the side connected between node A (A) and B (B)
Weights be 0.9.
In embodiments of the present invention, iteration updating block 200 is carrying out often taking turns iteration renewal social networks model
In all node label when, iteration updating block 200 can be by the side of the label of one node of following renewal
The label of all nodes in formula iteration renewal social networks model:Iteration updating block 200 can be arranged at random
Node in row social networks model, obtains the order of random alignment, and iteration updating block 200 can be by random
The order determination of arrangement need to currently be updated the present node of label, and neighbors is determined according to present node,
The weights for obtaining the side between the label and neighbors and present node of neighbors are carried out to obtain target labels
Update the label of present node.
In embodiments of the present invention, in the social networks model after division neighbourhood unit 300 can be handled renewal
Carry user's mark corresponding to the node of same label and be divided into same community, and can be to the use in the community
Family mark is uniformly processed, and such as launches identical advertisement.
Wherein, the construction unit 100 is specifically used for:
For the unique label of each node distribution.
The construction unit 100 is specifically used for:
The cohesion between the node for the node being connected with each other is calculated, according to the cohesion to the interconnection
Node between side carry out imparting weights.
The iteration updating block 200 includes:
Subelement is arranged, for the node in social networks model described in random alignment;
Iteration updates subelement, and the order according to random alignment is performed for iteration, starts successively according to current
The label of the weights on the side between node and neighbors and the neighbors updates the label of the present node,
And count iteration update times.
The iteration renewal subelement includes:
First determination subelement, for determining the neighbors per class label by the label of the neighbors, per class
The neighbors of label carries identical label;
First computation subunit, for the weights according to the side between the present node and the neighbors, meter
Calculate the neighbors per class label weights and;
First renewal subelement, for the weights from the neighbors per class label and the middle power for obtaining maximum
It is worth and determines the maximum weights and corresponding label, and the label for updating the present node is institute
State the weights of maximum and corresponding label.
Iteration renewal subelement also particularly useful for:
According to the weights on the side between the present node and the neighbors, the label of the neighbors and institute
The geodesic distance of neighbors is stated to update the label of the present node.
The iteration renewal subelement also includes:
First obtains subelement, for obtaining the geodesic distance of the neighbors;
Second determination subelement, for determining the neighbors per class label by the label of the neighbors, per class
The neighbors of label carries identical label;
Second computation subunit, distinguish for calculating the weights between the neighbors with the side of the present node
Ratio between the geodesic distance of the neighbors, obtain the ratio sum of the neighbors per class label;
Second renewal subelement, for obtaining maximum from the ratio sum of the neighbors per class label
Ratio sum, the maximum ratio sum and corresponding label are determined, and update the present node
Label is the maximum ratio sum and corresponding label.
The server also includes:
Memory cell, for according to the social networks model, establishing the corresponding relation of each node and label,
And the shortest path obtained between each node is stored as geodesic distance.
The first acquisition subelement includes:
Second obtains subelement, for obtaining the label of the neighbors, according to each node and label
Corresponding relation, obtain ancestor node corresponding to the label of the neighbors;
Subelement is set, for obtaining the neighbors and institute in the geodesic distance between each node of storage
The target geodesic distance between ancestor node is stated, the target geodesic distance got is set to the survey of the neighbors
Ground distance.
Wherein it is possible to understand, the function of each functional module of the unit in the server of the present embodiment can
Method specific implementation in above method embodiment, its specific implementation process are referred to above method reality
The associated description of example is applied, is no longer repeated herein.
In embodiments of the present invention, social networks model is built based on the social networks in social networks, to institute
To state the side being connected with each other between the node in social networks model and assign weights, each node carries label, according to
The weights on the side between the label and node of each node, which are iterated in the renewal social networks model, to be owned
Node label, when iteration update times meet predetermined threshold value, the social networks mould after renewal is handled
User's mark corresponding to the node of same label is carried in type and is divided into same community.This effectively improves service
The stability of the label of device more new node, so as to effectively improve the stability of server division community.
Fig. 8 is referred to, for the embodiments of the invention provide the structural representation of another server.Such as Fig. 8 institutes
Show, the server 1000 can include:At least one processor 1001, such as CPU, it is at least one
Network interface 1004, user interface 1003, memory 1005, at least one communication bus 1002.Wherein,
Communication bus 1002 is used to realize the connection communication between these components.Wherein, user interface 1003 can be with
Including display screen (Display), keyboard (Keyboard), optional user interface 1003 can also include standard
Wireline interface, wave point.Network interface 1004 can optionally include the wireline interface, wireless of standard
Interface (such as WI-FI interfaces).Memory 1005 can be high-speed RAM memory or non-shakiness
Fixed memory (non-volatile memory), for example, at least a magnetic disk storage.Memory 1005 can
Choosing can also be at least one storage device for being located remotely from aforementioned processor 1001.As shown in figure 8, make
For operating system, network communication module, use can be included in a kind of memory 1005 of computer-readable storage medium
Family interface module and communication connection application program.
In the terminal 1000 shown in Fig. 8, user interface 1003 is mainly used in providing the user connecing for input
Mouthful, obtain the data that user inputs;Network interface 1004 is used to be connected with server, obtains server hair
The data sent;And processor 1001 can be used for calling the communication connection application journey stored in memory 1005
Sequence, and specifically perform following operate:
Social networks model is built based on the social networks in social networks, in the social networks model
The side being connected with each other between node assigns weights, wherein, the social networks identifies including at least one user,
The social networks model includes at least one node, at least one user's mark with it is described at least one
Node corresponds, and each node carries label;
The renewal society is iterated according to the weights on the side between the label of each node and the node
Hand over the label of node all in network model;
When iteration update times meet predetermined threshold value, phase is carried in the social networks model after renewal is handled
Same community is divided into user's mark corresponding to the node of label.
Wherein, the processor 1001 is included based on the social networks structure social networks model in social networks:
For the unique label of each node distribution.
Wherein, the processor 1001 assigns the side being connected with each other the node in the social networks model
Weights include:
The cohesion between the node for the node being connected with each other is calculated, according to the cohesion to the interconnection
Node between side carry out imparting weights.
Wherein, the processor 1001 is according to the power on the side between the label of each node and the node
The label that value is iterated node all in the renewal social networks model includes:
Node in social networks model described in random alignment;
Iteration performs the order according to random alignment, starts successively according to the side between present node and neighbors
The label of weights and the neighbors updates the label of the present node, and counts iteration update times.
Wherein, weights and the adjacent section of the processor 1001 according to the side between present node and neighbors
The label of point includes to update the label of the present node:
The neighbors per class label is determined by the label of the neighbors, the neighbors per class label carries identical
Label;
According to the weights on the side between the present node and the neighbors, the neighbors of every class label is calculated
Weights and;
From the weights of the neighbors per class label and it is middle obtain maximum weights and, determine described maximum
Weights and corresponding label, and the label of the present node is updated for maximum weights and corresponding
Label.
Wherein, weights and the adjacent section of the processor 1001 according to the side between present node and neighbors
The label of point includes to update the label of the present node:
According to the weights on the side between the present node and the neighbors, the label of the neighbors and institute
The geodesic distance of neighbors is stated to update the label of the present node.
Wherein, weights of the processor 1001 according to the side between the present node and the neighbors, institute
The geodesic distance of the label and the neighbors of stating neighbors includes to update the label of the present node:
Obtain the geodesic distance of the neighbors;
The neighbors per class label is determined by the label of the neighbors, the neighbors per class label carries identical
Label;
Calculate between the neighbors with the weights on the side of the present node geodesic distance with the neighbors respectively
Ratio between, obtain the ratio sum of the neighbors per class label;
The ratio sum of maximum is obtained from the ratio sum of the neighbors per class label, it is determined that it is described most
Big ratio sum and corresponding label, and the label for updating the present node is the maximum ratio
Sum and corresponding label.
Wherein, in the processor 1001 according to the side between the label of each node and the node
Weights are iterated before the label for updating node all in the social networks model, the processor
1001 also perform:
According to the social networks model, the corresponding relation of each node and label is established, and obtains each section
Shortest path between point is stored as geodesic distance.
Wherein, the geodesic distance that the processor 1001 obtains the neighbors includes:
The label of the neighbors is obtained, according to each node and the corresponding relation of label, described in acquisition
Ancestor node corresponding to the label of neighbors;
The mesh between the neighbors and the ancestor node is obtained in geodesic distance between each node of storage
Geodesic distance is marked, the target geodesic distance got is set to the geodesic distance of the neighbors.
Wherein it is possible to understand, the function of each functional module of the unit in the server of the present embodiment can
Method specific implementation in above method embodiment, its specific implementation process are referred to above method reality
The associated description of example is applied, is no longer repeated herein.
In embodiments of the present invention, social networks model is built based on the social networks in social networks, to institute
To state the side being connected with each other between the node in social networks model and assign weights, each node carries label, according to
The weights on the side between the label and node of each node, which are iterated in the renewal social networks model, to be owned
Node label, when iteration update times meet predetermined threshold value, the social networks mould after renewal is handled
User's mark corresponding to the node of same label is carried in type and is divided into same community.This effectively improves service
The stability of the label of device more new node, so as to effectively improve the stability of server division community.
One of ordinary skill in the art will appreciate that all or part of flow in above-described embodiment method is realized,
It is that by computer program the hardware of correlation can be instructed to complete, described program can be stored in a calculating
In machine read/write memory medium, the program is upon execution, it may include such as the flow of the embodiment of above-mentioned each method.
Wherein, described storage medium can be magnetic disc, CD, read-only memory (Read-Only Memory,
) or random access memory (Random Access Memory, RAM) etc. ROM.
Above disclosure is only preferred embodiment of present invention, can not limit the present invention's with this certainly
Interest field, therefore the equivalent variations made according to the claims in the present invention, still belong to the scope that the present invention is covered.
Claims (18)
1. a kind of division methods of myspace, it is characterised in that methods described includes:
Social networks model is built based on the social networks in social networks, in the social networks model
The side being connected with each other between node assigns weights, wherein, the social networks identifies including at least one user,
The social networks model includes at least one node, at least one user's mark with it is described at least one
Node corresponds, and each node carries label;
The renewal society is iterated according to the weights on the side between the label of each node and the node
Hand over the label of node all in network model;
When iteration update times meet predetermined threshold value, phase is carried in the social networks model after renewal is handled
Same community is divided into user's mark corresponding to the node of label.
2. the method as described in claim 1, it is characterised in that the social activity based in social networks is closed
System's structure social networks model includes:
For the unique label of each node distribution.
3. the method as described in claim 1, mutually interconnected the node in the social networks model
The side connect, which assigns weights, to be included:
The cohesion between the node for the node being connected with each other is calculated, according to the cohesion to the interconnection
Node between side carry out imparting weights.
4. the method as described in claim 1, it is characterised in that the label according to each node
And the weights on the side between the node are iterated the mark for updating node all in the social networks model
Label include:
Node in social networks model described in random alignment;
Iteration performs the order according to random alignment, starts successively according to the side between present node and neighbors
The label of weights and the neighbors updates the label of the present node, and counts iteration update times.
5. method as claimed in claim 4, it is characterised in that described according between present node and neighbors
The weights on side and the label of the neighbors include to update the label of the present node:
The neighbors per class label is determined by the label of the neighbors, the neighbors per class label carries identical
Label;
According to the weights on the side between the present node and the neighbors, the neighbors of every class label is calculated
Weights and;
From the weights of the neighbors per class label and it is middle obtain maximum weights and, determine described maximum
Weights and corresponding label, and the label of the present node is updated for maximum weights and corresponding
Label.
6. method as claimed in claim 4, it is characterised in that described according between present node and neighbors
The weights on side and the label of the neighbors include to update the label of the present node:
According to the weights on the side between the present node and the neighbors, the label of the neighbors and institute
The geodesic distance of neighbors is stated to update the label of the present node.
7. method as claimed in claim 6, it is characterised in that it is described according to the present node with it is described
The geodesic distance of the weights on the side between neighbors, the label of the neighbors and the neighbors is to update
Stating the label of present node includes:
Obtain the geodesic distance of the neighbors;
The neighbors per class label is determined by the label of the neighbors, the neighbors per class label carries identical
Label;
Calculate between the neighbors with the weights on the side of the present node geodesic distance with the neighbors respectively
Ratio between, obtain the ratio sum of the neighbors per class label;
The ratio sum of maximum is obtained from the ratio sum of the neighbors per class label, it is determined that it is described most
Big ratio sum and corresponding label, and the label for updating the present node is the maximum ratio
Sum and corresponding label.
8. method as claimed in claim 7, it is characterised in that the label according to each node
And the weights on the side between the node are iterated the mark for updating node all in the social networks model
Include before label:
According to the social networks model, the corresponding relation of each node and label is established, and obtains each section
Shortest path between point is stored as geodesic distance.
9. method as claimed in claim 8, it is characterised in that the geodesic distance for obtaining the neighbors
From including:
The label of the neighbors is obtained, according to each node and the corresponding relation of label, described in acquisition
Ancestor node corresponding to the label of neighbors;
The mesh between the neighbors and the ancestor node is obtained in geodesic distance between each node of storage
Geodesic distance is marked, the target geodesic distance got is set to the geodesic distance of the neighbors.
10. a kind of server, it is characterised in that the server includes:
Construction unit, for building social networks model based on the social networks in social networks, to the society
The side being connected with each other between the node in network model is handed over to assign weights, wherein, the social networks is included at least
One user's mark, the social networks model include at least one node, at least one user's mark
Corresponded with least one node, each node carries label;
Iteration updating block, the weights for the side between the label according to each node and the node
It is iterated the label for updating node all in the social networks model;
Divide neighbourhood unit, for when iteration update times meet predetermined threshold value, the society after renewal is handled
Hand over user corresponding to carrying the node of same label in network model to identify and be divided into same community.
11. server as claimed in claim 10, it is characterised in that the construction unit is specifically used for:
For the unique label of each node distribution.
12. server as claimed in claim 10, the construction unit is specifically used for:
The cohesion between the node for the node being connected with each other is calculated, according to the cohesion to the interconnection
Node between side carry out imparting weights.
13. server as claimed in claim 10, it is characterised in that the iteration updating block includes:
Subelement is arranged, for the node in social networks model described in random alignment;
Iteration updates subelement, and the order according to random alignment is performed for iteration, starts successively according to current
The label of the weights on the side between node and neighbors and the neighbors updates the label of the present node,
And count iteration update times.
14. server as claimed in claim 13, it is characterised in that the iteration renewal subelement includes:
First determination subelement, for determining the neighbors per class label by the label of the neighbors, per class
The neighbors of label carries identical label;
First computation subunit, for the weights according to the side between the present node and the neighbors, meter
Calculate the neighbors per class label weights and;
First renewal subelement, for the weights from the neighbors per class label and the middle power for obtaining maximum
It is worth and determines the maximum weights and corresponding label, and the label for updating the present node is institute
State the weights of maximum and corresponding label.
15. server as claimed in claim 13, it is characterised in that the iteration renewal subelement also has
Body is used for:
According to the weights on the side between the present node and the neighbors, the label of the neighbors and institute
The geodesic distance of neighbors is stated to update the label of the present node.
16. server as claimed in claim 15, it is characterised in that the iteration renewal subelement also wraps
Include:
First obtains subelement, for obtaining the geodesic distance of the neighbors;
Second determination subelement, for determining the neighbors per class label by the label of the neighbors, per class
The neighbors of label carries identical label;
Second computation subunit, distinguish for calculating the weights between the neighbors with the side of the present node
Ratio between the geodesic distance of the neighbors, obtain the ratio sum of the neighbors per class label;
Second renewal subelement, for obtaining maximum from the ratio sum of the neighbors per class label
Ratio sum, the maximum ratio sum and corresponding label are determined, and update the present node
Label is the maximum ratio sum and corresponding label.
17. server as claimed in claim 16, the server also includes:
Memory cell, for according to the social networks model, establishing the corresponding relation of each node and label,
And the shortest path obtained between each node is stored as geodesic distance.
18. server as claimed in claim 17, it is characterised in that the first acquisition subelement includes:
Second obtains subelement, for obtaining the label of the neighbors, according to each node and label
Corresponding relation, obtain ancestor node corresponding to the label of the neighbors;
Subelement is set, for obtaining the neighbors and institute in the geodesic distance between each node of storage
The target geodesic distance between ancestor node is stated, the target geodesic distance got is set to the survey of the neighbors
Ground distance.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610458069.8A CN107529656B (en) | 2016-06-22 | 2016-06-22 | Method for dividing social network communities and server |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610458069.8A CN107529656B (en) | 2016-06-22 | 2016-06-22 | Method for dividing social network communities and server |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107529656A true CN107529656A (en) | 2018-01-02 |
CN107529656B CN107529656B (en) | 2021-01-15 |
Family
ID=60765435
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610458069.8A Active CN107529656B (en) | 2016-06-22 | 2016-06-22 | Method for dividing social network communities and server |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107529656B (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110110212A (en) * | 2018-01-22 | 2019-08-09 | 杭州橙鹰数据技术有限公司 | Combo discovering method, server, terminal installation and system |
CN110210867A (en) * | 2019-05-14 | 2019-09-06 | 无线生活(北京)信息技术有限公司 | The determination method and device of node label |
CN110738577A (en) * | 2019-09-06 | 2020-01-31 | 平安科技(深圳)有限公司 | Community discovery method, device, computer equipment and storage medium |
CN111182522A (en) * | 2019-11-28 | 2020-05-19 | 北京明略软件系统有限公司 | Group partner determining method, device, electronic equipment and computer storage medium |
CN111460233A (en) * | 2020-03-20 | 2020-07-28 | 北京五八信息技术有限公司 | Label propagation method and device, electronic equipment and storage medium |
CN111552846A (en) * | 2020-04-28 | 2020-08-18 | 支付宝(杭州)信息技术有限公司 | Method and device for identifying suspicious relationship |
CN111626887A (en) * | 2019-02-27 | 2020-09-04 | 北京奇虎科技有限公司 | Social relationship evaluation method and device |
CN111753154A (en) * | 2020-06-22 | 2020-10-09 | 北京三快在线科技有限公司 | User data processing method, device, server and computer readable storage medium |
CN111797663A (en) * | 2019-08-02 | 2020-10-20 | 北京京东尚科信息技术有限公司 | Collocation scene recognition method and device |
CN111899115A (en) * | 2020-05-30 | 2020-11-06 | 中国兵器科学研究院 | Method, device and storage medium for determining community structure in social network |
CN113177854A (en) * | 2021-04-23 | 2021-07-27 | 携程计算机技术(上海)有限公司 | Community division method and system, electronic device and storage medium |
CN117056239A (en) * | 2023-10-11 | 2023-11-14 | 腾讯科技(深圳)有限公司 | Method, device, equipment and storage medium for determining test function using characteristics |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102768670A (en) * | 2012-05-31 | 2012-11-07 | 哈尔滨工程大学 | Webpage clustering method based on node property label propagation |
CN105279187A (en) * | 2014-07-15 | 2016-01-27 | 天津科技大学 | Edge clustering coefficient-based social network group division method |
CN105335438A (en) * | 2014-08-11 | 2016-02-17 | 天津科技大学 | Local shortest loop based social network group division method |
-
2016
- 2016-06-22 CN CN201610458069.8A patent/CN107529656B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102768670A (en) * | 2012-05-31 | 2012-11-07 | 哈尔滨工程大学 | Webpage clustering method based on node property label propagation |
CN105279187A (en) * | 2014-07-15 | 2016-01-27 | 天津科技大学 | Edge clustering coefficient-based social network group division method |
CN105335438A (en) * | 2014-08-11 | 2016-02-17 | 天津科技大学 | Local shortest loop based social network group division method |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110110212A (en) * | 2018-01-22 | 2019-08-09 | 杭州橙鹰数据技术有限公司 | Combo discovering method, server, terminal installation and system |
CN111626887A (en) * | 2019-02-27 | 2020-09-04 | 北京奇虎科技有限公司 | Social relationship evaluation method and device |
CN110210867A (en) * | 2019-05-14 | 2019-09-06 | 无线生活(北京)信息技术有限公司 | The determination method and device of node label |
CN111797663A (en) * | 2019-08-02 | 2020-10-20 | 北京京东尚科信息技术有限公司 | Collocation scene recognition method and device |
WO2021043064A1 (en) * | 2019-09-06 | 2021-03-11 | 平安科技(深圳)有限公司 | Community detection method and apparatus, and computer device and storage medium |
CN110738577A (en) * | 2019-09-06 | 2020-01-31 | 平安科技(深圳)有限公司 | Community discovery method, device, computer equipment and storage medium |
CN110738577B (en) * | 2019-09-06 | 2022-02-22 | 平安科技(深圳)有限公司 | Community discovery method, device, computer equipment and storage medium |
CN111182522A (en) * | 2019-11-28 | 2020-05-19 | 北京明略软件系统有限公司 | Group partner determining method, device, electronic equipment and computer storage medium |
CN111182522B (en) * | 2019-11-28 | 2022-10-25 | 北京明智和术科技有限公司 | Group partner determining method, device, electronic equipment and computer storage medium |
CN111460233A (en) * | 2020-03-20 | 2020-07-28 | 北京五八信息技术有限公司 | Label propagation method and device, electronic equipment and storage medium |
CN111552846A (en) * | 2020-04-28 | 2020-08-18 | 支付宝(杭州)信息技术有限公司 | Method and device for identifying suspicious relationship |
CN111552846B (en) * | 2020-04-28 | 2023-09-08 | 支付宝(杭州)信息技术有限公司 | Method and device for identifying suspicious relationships |
CN111899115A (en) * | 2020-05-30 | 2020-11-06 | 中国兵器科学研究院 | Method, device and storage medium for determining community structure in social network |
CN111899115B (en) * | 2020-05-30 | 2021-03-23 | 中国兵器科学研究院 | Method, device and storage medium for determining community structure in social network |
CN111753154A (en) * | 2020-06-22 | 2020-10-09 | 北京三快在线科技有限公司 | User data processing method, device, server and computer readable storage medium |
CN111753154B (en) * | 2020-06-22 | 2024-03-19 | 北京三快在线科技有限公司 | User data processing method, device, server and computer readable storage medium |
CN113177854A (en) * | 2021-04-23 | 2021-07-27 | 携程计算机技术(上海)有限公司 | Community division method and system, electronic device and storage medium |
CN117056239A (en) * | 2023-10-11 | 2023-11-14 | 腾讯科技(深圳)有限公司 | Method, device, equipment and storage medium for determining test function using characteristics |
CN117056239B (en) * | 2023-10-11 | 2024-01-30 | 腾讯科技(深圳)有限公司 | Method, device, equipment and storage medium for determining test function using characteristics |
Also Published As
Publication number | Publication date |
---|---|
CN107529656B (en) | 2021-01-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107529656A (en) | The division methods and server of a kind of myspace | |
Mustafa et al. | Modelling built-up expansion and densification with multinomial logistic regression, cellular automata and genetic algorithm | |
CN106126521B (en) | The social account method for digging and server of target object | |
CN109598563A (en) | Brush single detection method, device, storage medium and electronic equipment | |
CN107123056B (en) | Social big data information maximization method based on position | |
Chen et al. | Mean-variance model for the build-operate-transfer scheme under demand uncertainty | |
CN111309824A (en) | Entity relationship map display method and system | |
CN105630800A (en) | Node importance ranking method and system | |
CN110111110A (en) | The method and apparatus of knowledge based map detection fraud, storage medium | |
CN110784883B (en) | Base station construction evaluation method, device, equipment and storage medium | |
Hu et al. | Trade structure and risk transmission in the international automotive Li-ion batteries trade | |
Yeshwanth et al. | Evolutionary churn prediction in mobile networks using hybrid learning | |
CN107742169A (en) | A kind of Urban Transit Network system constituting method and performance estimating method based on complex network | |
CN106780064A (en) | A kind of region partitioning method, device and the network equipment | |
CN112101577B (en) | XGboost-based cross-sample federal learning and testing method, system, device and medium | |
CN104820945A (en) | Online social network information transmision maximization method based on community structure mining algorithm | |
CN112020012B (en) | Moving track reconstruction and road matching method, storage medium and server | |
CN110321438A (en) | Real-time fraud detection method, device and electronic equipment based on complex network | |
CN106874931B (en) | User portrait clustering method and device | |
Chen et al. | Incorporating geographical location for team formation in social coding sites | |
CN107590189A (en) | Intelligent contract performs method, apparatus, equipment and storage medium | |
Cao et al. | Mapping strategy for virtual networks in one stage | |
CN105488247A (en) | K-mean community structure mining method and apparatus | |
Giovannetti et al. | An Internet periphery study: Network centrality and clustering for mobile access in Bhutan | |
CN107578136A (en) | The overlapping community discovery method extended based on random walk with seed |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |