CN110309419A - A kind of overlapping anatomic framework method for digging and device propagated based on balance multi-tag - Google Patents
A kind of overlapping anatomic framework method for digging and device propagated based on balance multi-tag Download PDFInfo
- Publication number
- CN110309419A CN110309419A CN201810457626.3A CN201810457626A CN110309419A CN 110309419 A CN110309419 A CN 110309419A CN 201810457626 A CN201810457626 A CN 201810457626A CN 110309419 A CN110309419 A CN 110309419A
- Authority
- CN
- China
- Prior art keywords
- node
- neighbor node
- tally set
- ownership
- neighbor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000000644 propagated effect Effects 0.000 title claims abstract description 25
- 238000010586 diagram Methods 0.000 claims abstract description 31
- 238000001914 filtration Methods 0.000 claims abstract description 20
- 230000000717 retained effect Effects 0.000 claims description 16
- 238000009412 basement excavation Methods 0.000 claims description 5
- 210000002230 centromere Anatomy 0.000 claims description 3
- 230000001186 cumulative effect Effects 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 238000013178 mathematical model Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000006854 communication Effects 0.000 abstract description 7
- 238000004422 calculation algorithm Methods 0.000 description 15
- 238000004590 computer program Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 5
- 235000013162 Cocos nucifera Nutrition 0.000 description 3
- 244000060011 Cocos nucifera Species 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000005325 percolation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The invention discloses a kind of overlapping anatomic framework method for digging and device propagated based on balance multi-tag, this method comprises: 1) reading and constructing social network diagram;2) label assigning process;3) neighbor node process is filtered;4) label communication process is executed;5) overlapping anatomic framework Result process is obtained.Invention defines organizational structures to belong to tendency degree S, before label communication process, the neighbor node of each node is filtered, filters out and label, which helps lesser part neighbor node, to be determined to central node, is i.e. organizational structure belongs to the lesser neighbor node of tendency degree S.When overlapping anatomic architectural configurations are more complex, connect intensive between node when, by filtration fraction neighbor node, can effectively promote the speed of label propagation, improve the efficiency that organizational structure is excavated.
Description
Technical field
Excavating the present invention relates to complex networks framework is complex network community discovery technique field, and in particular to a kind of
The overlapping anatomic framework method for digging and device propagated based on balance multi-tag.
Background technique
There are information and mode abundant in complex network, and is side building taking human as node, human relationships
It equally include the information such as relationship between organizational structure structure, member in social networks.Therefore by excavating in social networks
Organizational structure structure, i.e. community structure can obtain a large amount of valuable information, this theoretically be all ten in practical application
Divide beneficial.
In the past 10 years, many complex network community method for digging are suggested, including classical GN algorithm, modularity
Optimization algorithm, based on dynamic (dynamical) method and the method etc. propagated based on label.Based on label propagate community discovery algorithm by
It is widely used in its time complexity with near-linear.
Label propagation algorithm was suggested in 2002 earliest, it is a kind of semi-supervised learning method based on figure,
Algorithm basic thought is the label information for removing to predict other also unmarked nodes with the label information of marked node.Earlier
RAK algorithm, LPAm algorithm etc. belong to non-overlap Combo discovering method, but the community structure that these algorithms excavate all is
Non-overlap, that is to say, that each node can only be subordinated to some corporation, and there is no overlay structures between corporations.Obviously this
Be not consistent the case where network in society, therefore it is subsequent in succession develop clique percolation method (CPM), EAGLE calculate
Method, GCE algorithm, LFK algorithm, COPRA algorithm etc. are overlapped Combo discovering method.
In COPRA algorithm, node x can update the label of oneself according to the label of its neighbours' point set.In tag update
During if it exists if multiple optional labels, algorithm will randomly choose v label therein as updating as a result, v
It is used to limit the parameter for the number of tags that each node can possess, setting parameter v avoids all labels and is updated to
Identical result.But existing overlapping anatomic framework label propagation algorithm is when each nodes neighbors node is excessive, the propagation of label
Process can be complex.Therefore, how to solve the problems, such as this, become an important research side for those skilled in the art
To.
Summary of the invention
The present invention provides a kind of overlapping anatomic framework excavation propagated based on balance multi-tag according to above-mentioned technical background
Method and device can filter out unnecessary neighbor node, to effectively improve the efficiency of label propagation.
In a first aspect, the present invention provides a kind of overlapping anatomic framework method for digging propagated based on balance multi-tag, packet
Include following steps:
Step 1: reading social network data, construct the social networks that the relationship using user as node, between user is side
Figure;
Step 2: distributing multiple labels, multiple labels composition of each node for each node in the social network diagram
The tally set of the node;
Step 3: each node in the social network diagram is traversed according to random sequence, to each label node to be determined,
That is central node traverses its neighbor node, the tally set of each neighbor node is obtained, according to the information mistake in the tally set
Filter the part of nodes in the neighbor node;
Step 4: according to the neighbor node retained after neighbor node filtering, the balance for calculating the central node is returned
Belong to coefficient, the part labels of the central node are retained according to the value of the balance ownership coefficient;
Step 5: continuing to execute step 3 and step 4, until the tally set of each node no longer changes, finally obtain overlapping group
The result that stretching frame structure excavates.
Preferably, in above-mentioned steps 1, the mathematical model of the social network diagram is G=(V, E);Wherein, V represents section
The set of point, E represent the set on connection side.
It preferably, in step 2 above, include k label in the tally set, the k value of the tally set of each node differs,
The number of the organizational structure of i.e. each node-home is unrestricted.
It is highly preferred that in step 2 above, multiple labels in the tally set are the binary number that form is (c, b)
Right, wherein c is organizational structure identifier, and b is ownership coefficient, indicates the relationship strength of node and organizational structure.
Preferably, above-mentioned steps 3 filter the specific steps of neighbor node are as follows:
Step 3-1: each node in the social network diagram is traversed according to random sequence, each node is label to be determined
Node, also referred to as central node;Find the neighbor node set N (x) of each central node;
Step 3-2: the tally set ((c of each neighbor node in the neighbor node set N (x) is obtained1,b1),(c2,
b2),...,(ck, bk), wherein b1+b2+…+bk=1;
Step 3-3: according to the tally set of each neighbor node of the central node of acquisition, each neighbor node is calculated
Organizational structure belong to tendency degree S, is defined as:
Wherein Is defined as: all ownership coefficients is averaged in neighbor node tally set
Value;
Organizational structure in the neighbor node of the central node is selected to belong to the lesser part of nodes of tendency degree S, by its mistake
Filter, even if it does not play dissemination in label communication process;If the S value of each neighbor node is equal, by the k value of tally set compared with
Big neighbor node filters out.
It is highly preferred that the organizational structure ownership tendency degree S characterization is certain in the neighbor node in step 3-3
Ownership tendentiousness of one node to each organizational structure, ownership gender gap of the lower neighbor node of S value to each organizational structure
Less, which or which organizational structure cannot be belonged to by clearly judgement, can not be provided for the determination of the label of center node more
It helps, therefore can be filtered.
Preferably, above-mentioned steps 4 execute the specific steps that label is propagated are as follows:
Step 4-1: cumulative ownership coefficient: by organizational structure identifier c identical in the tally set of all neighbor nodes
Corresponding ownership coefficient b is added, and obtains the tally set { (c of central node01, b01),(c02,b02) ..., (com, bom)};
Step 4-2: filtering ownership coefficient: the maximum ownership coefficient b in the tally set of the central node is foundmaxAnd its
Respective labels cmax, given threshold parameter p, if some ownership coefficient b meets following formula in the tally set of the central node:
Then this ownership coefficient and its organizational structure identifier c are retained, and otherwise will be filtered;After filtering it is described in
The tally set of heart node becomes { (c01, b01),(c02, b02) ..., (con, bon), wherein n≤m, p are threshold parameters, p ∈ (0,
1];
Step 4-3: normalization ownership coefficient: the ownership coefficient retained after the step 4-2 be not able to satisfy addition and
It is 1, it is normalized;Wherein, formula is normalized are as follows:
It is highly preferred that in step 4-2, the p is threshold parameter, and p ∈ (0,1], indicate the label of the central node
The equilibrium degree in some ownership coefficient and tally set between maximum ownership coefficient is concentrated, the value of p is set depending on the height of the equilibrium degree
Fixed, then p value is larger for equilibrium degree height, and the low then p value of equilibrium degree is smaller.
Second aspect, the present invention provides a kind of overlapping anatomic framework excavating gear propagated based on balance multi-tag, packets
It includes:
Input module: for reading social network data;
Constructing module: it for the social network data according to reading, constructs using user as node, the relationship between user is
The social network diagram on side;
Label distribution module: for distributing multiple labels for each node in the social network diagram, each node
Multiple labels form the tally set of the node;
Filter neighbor node module: for traversing each node in the social network diagram according to random sequence, to each
Central node traverses its neighbor node, the tally set of each neighbor node is obtained, according to the information filtering in the tally set
Fall the part of nodes in the neighbor node;
Execute label propagation module: for according to the neighbor node that retains after neighbor node filtering, described in calculating
The balance of central node belongs to coefficient, and the part labels of the central node are retained according to the value of the balance ownership coefficient;
Determination module: for determining whether the tally set of each node changes again, to select next step;
Output module: for obtaining the result of overlapping anatomic framework excavation.
The beneficial effect of above-mentioned technical proposal provided by the invention includes at least:
Compared with prior art, the present invention is based on the balance community discovery algorithms (BMLPA) that multi-tag is propagated to calculate in COPRA
The more new strategy that label has been redesigned on the basis of method controls the label that each node can possess by given threshold p
Number, therefore do not need setting parameter v, that is to say, that the number of the overlapping corporations of excavation is no longer limited by parameter v.
Meanwhile invention defines organizational structures to belong to tendency degree S, before label communication process, to each node
Neighbor node is filtered, and is filtered out and is determined that label helps lesser part neighbor node to central node, is i.e. organizational structure is returned
Belong to the lesser neighbor node of tendency degree S.When overlapping anatomic architectural configurations are more complex, connect intensive between node when, each node
Neighbor node is more, this will lead to the inefficiency of label communication process, takes a long time.The present invention is saved by filtration fraction neighbours
Point can effectively promote the speed of label propagation, improve the efficiency that organizational structure is excavated.
Other features and advantages of the present invention will be illustrated in the following description, also, partly becomes from specification
It obtains it is clear that understand through the implementation of the invention.The objectives and other advantages of the invention can be by written explanation
Specifically noted structure is achieved and obtained in book, claims and attached drawing.
Below by the drawings and specific embodiments, technical scheme of the present invention will be described in further detail.
Detailed description of the invention
Attached drawing is used to provide further understanding of the present invention, and constitutes part of specification, with reality of the invention
It applies example to be used to explain the present invention together, not be construed as limiting the invention.In the accompanying drawings:
Fig. 1 is that the overlapping anatomic framework method for digging provided in an embodiment of the present invention propagated based on balance multi-tag realizes stream
Cheng Tu;
Fig. 2 is that neighbor node provided in an embodiment of the present invention filters schematic diagram;
Fig. 3 is the frame of the overlapping anatomic framework excavating gear provided in an embodiment of the present invention propagated based on balance multi-tag
Figure.
Specific embodiment
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
It is fully disclosed to those skilled in the art.
Embodiment:
Referring to Fig.1 shown in -2, the overlapping anatomic framework provided in an embodiment of the present invention propagated based on balance multi-tag is excavated
Method, comprising the following steps:
Step 1: reading social network data, construct the social networks that the relationship using user as node, between user is side
Figure, the mathematical model of social network diagram are G=(V, E);Wherein, V represents the set of node, and E represents the set on connection side.
Step 2: label assigning process: distributing multiple labels for each node in social network diagram, each node it is more
A label forms the tally set of the node, includes k label in tally set, the k value of the tally set of each node differs, i.e., each node
The number of the organizational structure of ownership is unrestricted.Multiple labels in tally set are the binary number pair that form is (c, b), wherein
C is organizational structure identifier, and b is ownership coefficient, indicates the relationship strength of node and organizational structure.
Step 3: filtering neighbor node: according to each node in random sequence traversal social network diagram, to each centromere
Point traverses its neighbor node, obtains the tally set of each neighbor node, is fallen in neighbor node according to the information filtering in tally set
Part of nodes.Detailed process is as follows:
Step 3-1: according to each node in random sequence traversal social network diagram, each node is label section to be determined
Point, also referred to as central node;Find the neighbor node set N (x) of each central node;
Step 3-2: the tally set { (c of each neighbor node in neighbor node set N (x) is obtained1, b1), (c2, b2) ...,
(ck, bk), wherein b1+b2+…+bk=1;
Step 3-3: according to the tally set of each neighbor node of the central node of acquisition, the group stretching frame of each neighbor node is calculated
Structure belongs to tendency degree S, is defined as:
Wherein Is defined as: all ownership coefficients is averaged in neighbor node tally set
Value;
It selects organizational structure in the neighbor node of central node to belong to the lesser part of nodes of tendency degree S, is filtered, i.e.,
It is set not play dissemination in label communication process;If the S value of each neighbor node is equal, and the k value of tally set is biggish
Neighbor node filters out.
What above-mentioned organizational structure ownership tendency degree S was characterized be in neighbor node a certain node to the ownership of each organizational structure
Tendentiousness, it is little in view of ownership gender gap of the lower neighbor node of S value to each organizational structure, it cannot be belonged to by clearly judgement
Which or which organizational structure, can not provide more help, therefore can be filtered for the determination of the label of center node.
Step 4: execute label and propagate: the neighbor node retained after being filtered according to neighbor node calculates central node
Balance ownership coefficient retains the part labels of central node according to the value of balance ownership coefficient.Detailed process is as follows:
Step 4-1: cumulative ownership coefficient: organizational structure identifier c identical in the tally set of all neighbor nodes is corresponding
Ownership coefficient b be added, obtain the tally set { (c of central node01, b01),(c02, b02) ..., (com, bom)};
Step 4-2: filtering ownership coefficient: the maximum ownership coefficient b in the tally set of central node is foundmaxAnd its it is corresponding
Label cmax, given threshold parameter p, if some ownership coefficient b meets following formula in the tally set of central node:
Then this ownership coefficient and its organizational structure identifier c are retained, and otherwise will be filtered;Filter rear center's section
The tally set of point becomes { (c01, b01), (c02, b02) ..., (con, bon), wherein n≤m, p are threshold parameters, and p ∈ (0,1], table
Show that some in the tally set of central node belongs to the equilibrium degree in coefficient and tally set between maximum ownership coefficient, the value view of p is equal
The height of weighing apparatus degree is set, and then p value is larger for equilibrium degree height, and the low then p value of equilibrium degree is smaller, and the value of p is too high or too low can all influence group
The result that stretching frame structure excavates;
Step 4-3: normalization ownership coefficient: the ownership coefficient retained after step 4-2 is not able to satisfy addition and is 1,
Therefore it is normalized;Wherein, formula is normalized are as follows:
Step 5: continuing to execute step 3 and step 4, until the tally set of each node no longer changes, finally obtain overlapping group
The result that stretching frame structure excavates.
The above-mentioned overlapping anatomic framework method for digging propagated based on balance multi-tag provided by the invention defines a group stretching frame
Structure ownership tendency degree S is filtered the neighbor node of each node, filters out to centromere before label communication process
Point determines that label helps lesser part neighbor node, i.e. organizational structure belongs to the lesser neighbor node of tendency degree S.When overlapping group
It knits that architectural configurations are more complex, when connecting intensive between node, by filtration fraction neighbor node, can effectively promote label propagation
Speed improves the efficiency that organizational structure is excavated.Based on the same inventive concept, the embodiment of the invention also provides based on the more marks of balance
The overlapping anatomic framework excavating gear propagated is signed, since the principle of the solved problem of the device and the aforementioned balance multi-tag that is based on pass
The overlapping anatomic framework method for digging broadcast is similar, therefore the implementation of the device may refer to the implementation of preceding method, repeats place
It repeats no more.
The embodiment of the invention also provides a kind of overlapping anatomic framework excavating gear propagated based on balance multi-tag, references
Shown in Fig. 3, comprising:
Input module 31: for reading social network data;
Constructing module 32: for the social network data according to reading, the relationship using user as node, between user is constructed
For the social network diagram on side;
Label distribution module 33: for distributing multiple labels, each node for each node in the social network diagram
Multiple labels form the tally set of the node;
Filter neighbor node module 34: for traversing each node in the social network diagram according to random sequence, to every
A central node traverses its neighbor node, the tally set of each neighbor node is obtained, according to the information mistake in the tally set
Filter the part of nodes in the neighbor node;
Execute label propagation module 35: for calculating institute according to the neighbor node retained after neighbor node filtering
The balance ownership coefficient for stating central node, the part labels of the central node are retained according to the value of the balance ownership coefficient;
Determination module 36: for determining whether the tally set of each node changes again, to select next step;
Output module 37: for obtaining the result of overlapping anatomic framework excavation.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The shape for the computer program product implemented in usable storage medium (including but not limited to magnetic disk storage and optical memory etc.)
Formula.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range.In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to include these modifications and variations.
Claims (9)
1. a kind of overlapping anatomic framework method for digging propagated based on balance multi-tag, it is characterised in that: the following steps are included:
Step 1: reading social network data, construct the social network diagram that the relationship using user as node, between user is side;
Step 2: distributing multiple labels for each node in the social network diagram, multiple labels of each node form the section
The tally set of point;
Step 3: traversing each node in the social network diagram according to random sequence, to each central node, traverse its neighbour
Node obtains the tally set of each neighbor node, is fallen in the neighbor node according to the information filtering in the tally set
Part of nodes;
Step 4: according to the neighbor node retained after neighbor node filtering, calculating the balance ownership system of the central node
Number retains the part labels of the central node according to the value of the balance ownership coefficient;
Step 5: continuing to execute step 3 and step 4, until the tally set of each node no longer changes, finally obtain overlapping anatomic frame
The result that structure excavates.
2. the overlapping anatomic framework method for digging according to claim 1 propagated based on balance multi-tag, it is characterised in that:
In step 1, the mathematical model of the social network diagram is G=(V, E);Wherein, V represents the set of node, and E represents connection side
Set.
3. the overlapping anatomic framework method for digging according to claim 1 propagated based on balance multi-tag, it is characterised in that:
It in step 2, include k label in the tally set, the k value of the tally set of each node differs.
4. the overlapping anatomic framework method for digging according to claim 1 or 3 propagated based on balance multi-tag, feature are existed
In: in step 2, multiple labels in the tally set are the binary number pair that form is (c, b), and wherein c is organizational structure
Identifier, b are ownership coefficient, indicate the relationship strength of node and organizational structure.
5. the overlapping anatomic framework method for digging according to claim 1 propagated based on balance multi-tag, it is characterised in that:
In step 3, each node in the social network diagram is traversed according to random sequence, to each central node, traverses its neighbour
Node obtains the tally set of each neighbor node, is fallen in the neighbor node according to the information filtering in the tally set
The specific steps of part of nodes are as follows:
Step 3-1: each node in the social network diagram is traversed according to random sequence, each node is label section to be determined
Point, also is indicated as central node;Find the neighbor node set N (x) of each central node;
Step 3-2: the tally set { (c of each neighbor node in the neighbor node set N (x) is obtained1, b1), (c2, b2) ...,
(ck, bk), wherein b1+b2+…+bk=1;
Step 3-3: according to the tally set of each neighbor node of the central node of acquisition, the group of each neighbor node is calculated
Stretching frame structure belongs to tendency degree S, is defined as:
Wherein Is defined as: the average value of all ownership coefficients in neighbor node tally set;
It selects organizational structure in the neighbor node of the central node to belong to the lesser part of nodes of tendency degree S, is filtered;If
The S value of each neighbor node is equal, then filters out the biggish neighbor node of k value of tally set.
6. the overlapping anatomic framework method for digging according to claim 5 propagated based on balance multi-tag, it is characterised in that:
In step 3-3, what organizational structure ownership tendency degree S was characterized be in the neighbor node a certain node to each group of stretching frame
The ownership tendentiousness of structure.
7. the overlapping anatomic framework method for digging according to claim 1 propagated based on balance multi-tag, it is characterised in that:
In step 4, according to the neighbor node retained after neighbor node filtering, the balance ownership system of the central node is calculated
Number retains the specific steps of the part labels of the central node according to the value of the balance ownership coefficient are as follows:
Step 4-1: cumulative ownership coefficient: organizational structure identifier c identical in the tally set of all neighbor nodes is corresponding
Ownership coefficient b be added, obtain the tally set { (c of central node01, b01), (c02, b02) ..., (com, bom)};
Step 4-2: filtering ownership coefficient: the maximum ownership coefficient b in the tally set of the central node is foundmaxAnd its it is corresponding
Label cmax, given threshold parameter p, if some ownership coefficient b meets following formula in the tally set of the central node:
Then this ownership coefficient and its organizational structure identifier c are retained, and otherwise will be filtered;The centromere after filtering
The tally set of point becomes { (c01, b01), (c02, b02) ..., (con, bon), wherein n≤m, p are threshold parameters, p ∈ (0,1];
Step 4-3: normalization ownership coefficient: the ownership coefficient retained after the step 4-2 is not able to satisfy addition and is 1,
It is normalized;Wherein, formula is normalized are as follows:
8. the overlapping anatomic framework method for digging according to claim 7 propagated based on balance multi-tag, it is characterised in that:
In step 4-2, p is threshold parameter, and p ∈ (0,1], indicate some ownership coefficient and label in the tally set of the central node
The equilibrium degree between maximum ownership coefficient is concentrated, the value of p regards the height setting of the equilibrium degree.
9. a kind of overlapping anatomic framework excavating gear propagated based on balance multi-tag characterized by comprising
Input module: for reading social network data;
Constructing module: for the social network data according to reading, constructing the relationship using user as node, between user is side
Social network diagram;
Label distribution module: for distributing multiple labels for each node in the social network diagram, each node it is multiple
Label forms the tally set of the node;
Filter neighbor node module: for traversing each node in the social network diagram according to random sequence, to each center
Node traverses its neighbor node, obtains the tally set of each neighbor node, falls institute according to the information filtering in the tally set
State the part of nodes in neighbor node;
Execute label propagation module: for calculating the center according to the neighbor node retained after neighbor node filtering
The balance of node belongs to coefficient, and the part labels of the central node are retained according to the value of the balance ownership coefficient;
Determination module: for determining whether the tally set of each node changes again, to select next step;
Output module: for obtaining the result of overlapping anatomic framework excavation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810457626.3A CN110309419A (en) | 2018-05-14 | 2018-05-14 | A kind of overlapping anatomic framework method for digging and device propagated based on balance multi-tag |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810457626.3A CN110309419A (en) | 2018-05-14 | 2018-05-14 | A kind of overlapping anatomic framework method for digging and device propagated based on balance multi-tag |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110309419A true CN110309419A (en) | 2019-10-08 |
Family
ID=68073919
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810457626.3A Pending CN110309419A (en) | 2018-05-14 | 2018-05-14 | A kind of overlapping anatomic framework method for digging and device propagated based on balance multi-tag |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110309419A (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103020267A (en) * | 2012-12-26 | 2013-04-03 | 上海交通大学 | Complex network community structure mining method based on triangular cluster multi-label transmission |
CN103729475A (en) * | 2014-01-24 | 2014-04-16 | 福州大学 | Multi-label propagation discovery method of overlapping communities in social network |
CN104636978A (en) * | 2015-02-12 | 2015-05-20 | 西安电子科技大学 | Detection method for overlapping community based on multi-label propagation |
CN105138588A (en) * | 2015-07-31 | 2015-12-09 | 南开大学 | Database overlap mode abstract generating method based on multi-label propagation |
CN106251230A (en) * | 2016-07-22 | 2016-12-21 | 福建师范大学 | A kind of community discovery method propagated based on election label |
US20170169215A1 (en) * | 2015-12-10 | 2017-06-15 | Ústav informatiky AV CR, v.v.i. | Identification of mislabeled samples via phantom nodes in label propagation |
CN106951524A (en) * | 2017-03-21 | 2017-07-14 | 哈尔滨工程大学 | Overlapping community discovery method based on node influence power |
-
2018
- 2018-05-14 CN CN201810457626.3A patent/CN110309419A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103020267A (en) * | 2012-12-26 | 2013-04-03 | 上海交通大学 | Complex network community structure mining method based on triangular cluster multi-label transmission |
CN103729475A (en) * | 2014-01-24 | 2014-04-16 | 福州大学 | Multi-label propagation discovery method of overlapping communities in social network |
CN104636978A (en) * | 2015-02-12 | 2015-05-20 | 西安电子科技大学 | Detection method for overlapping community based on multi-label propagation |
CN105138588A (en) * | 2015-07-31 | 2015-12-09 | 南开大学 | Database overlap mode abstract generating method based on multi-label propagation |
US20170169215A1 (en) * | 2015-12-10 | 2017-06-15 | Ústav informatiky AV CR, v.v.i. | Identification of mislabeled samples via phantom nodes in label propagation |
CN106251230A (en) * | 2016-07-22 | 2016-12-21 | 福建师范大学 | A kind of community discovery method propagated based on election label |
CN106951524A (en) * | 2017-03-21 | 2017-07-14 | 哈尔滨工程大学 | Overlapping community discovery method based on node influence power |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
DE112019004076T5 (en) | DECENTRALIZED DISTRIBUTED DEEP LEARNING | |
CN107193896A (en) | A kind of diagram data division methods based on cluster | |
CN112100450A (en) | Graph calculation data segmentation method, terminal device and storage medium | |
CN108009437A (en) | Data publication method and apparatus and terminal | |
CN104881415A (en) | Tree graph generation method and tree graph generation equipment | |
CN106446223A (en) | Map data processing method and device | |
DE102013209934B4 (en) | Start or stop virtual servers in proper order | |
CN112965710A (en) | Processing method, device and system of computation graph | |
CN104268243A (en) | Position data processing method and device | |
CN106095843B (en) | Social account method for digging and device | |
CN104662842A (en) | Lattice based traffic measurement | |
CA2743466A1 (en) | Path calculation order deciding method, program and calculating apparatus | |
CN110309419A (en) | A kind of overlapping anatomic framework method for digging and device propagated based on balance multi-tag | |
CN110876072A (en) | Batch registered user identification method, storage medium, electronic device and system | |
CN111737647A (en) | Method and device for detecting abnormal connected subgraph | |
CN110414689A (en) | Update method and device on a kind of machine learning model line | |
CN107155215A (en) | The distribution method and device of a kind of application home service cluster | |
CN104504719A (en) | Image edge detection method and equipment | |
CN108734545A (en) | Collaborative filtering recommending method and system | |
CN114490835A (en) | High-utility item set mining method and device, electronic equipment and medium | |
Tripathi et al. | Improving Applicability of DE Algorithm by Designing Multi-Operator Variant of DE. | |
CN110688623B (en) | Training optimization method, device, equipment and storage medium for high-order LR model | |
CN114021031A (en) | Financial product information pushing method and device | |
CN114037008A (en) | Node classification method and system for multi-granularity attribute network embedding based on attribute continuous edges | |
Wang et al. | Resilience of hybrid herbivore–plant–pollinator networks |
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: 20191008 |