CN109815414A - Social networks character relation analysis method based on multitiered network community division - Google Patents
Social networks character relation analysis method based on multitiered network community division Download PDFInfo
- Publication number
- CN109815414A CN109815414A CN201910061662.2A CN201910061662A CN109815414A CN 109815414 A CN109815414 A CN 109815414A CN 201910061662 A CN201910061662 A CN 201910061662A CN 109815414 A CN109815414 A CN 109815414A
- Authority
- CN
- China
- Prior art keywords
- network
- user
- community division
- social networks
- multitiered
- 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
Landscapes
- Information Transfer Between Computers (AREA)
Abstract
The present invention discloses a kind of social networks character relation analysis method based on multitiered network community division, comprising: S1, obtains people information data by Twitter open platform;S2, building multitiered network;Include: friend relation network, forwards relational network, thumb up relational network;S3, by friend relation network that step S2 is constructed, forward relational network, thumb up relational network and merge;S4, community division is carried out to the fused multitiered network of step S3;S5, according to being completely embedded between corporations' node or sparse feature, analysis obtains task nexus;Multilayer social networks community discovery algorithm of the present invention based on side cluster, can be effectively prevented from parameter uncertainty problem, community division result is more accurate.
Description
Technical field
The invention belongs to social media field, in particular to the user individual service technology of a kind of social media.
Background technique
With the development of development of Mobile Internet technology, social media is also risen therewith, various social media services
It emerges one after another, it is well-known such as external LinkedIn, Twitter, Facebook, Instagram, domestic such as Sina weibo, beans
Valve net and know.These famous social media sites are owned by a large amount of user.Such as the maximum professional social network sites in the whole world
LinkedIn has more than 500,000,000 users in the multiple countries and regions of global 200d.By in March, 2018, Twitter is shared
3.36 hundred million any active ues, total number of users also already exceed 500,000,000.Therefore for social media character relation analysis and research at
For a big hot spot of research.This is that user's customization individual service is of great significance for social network sites.
Since Watts in 1998 and Strogatz propose small-world network model, the research of complex network was in the past
It is rapidly developed within several years.It is directly right in view of the sociability of social media and the amount of user information of social media are huge
The information of user itself carries out analysis and research and has ignored complicated social networks network between user.Pass through various passes between user
It is one huge and complicated multitiered network of (such as friend relation, thumb up relationship, forwarding relationship etc.) composition, therefore can be to society
The structural property of multilayer complex network locating for media personality is handed over to be analyzed to study character relation.It is more both at home and abroad in recent years
Layer network research field flourishes, and Quality Research structural for multitiered network has had many achievements, analyzes complex web
The structural property of network plays an important roll analysis social media character relation.Excavation to character relation net is a community
Topology discovery process, this process can disclose the interaction of user in complicated network, have significant practical meaning
Justice.For example, carrying out personalized user in different communities recommends or formulates different degrees of public sentiment monitoring strategies.
Community structure is one of basic structure property of complex network, and community structure parser research is that research is practical
The community structure that network has, i.e. whole network are made of several corporations, and the connection between corporations is relatively sparse, inside corporations
Connection it is relatively dense.The community detecting algorithm of relatively mainstream has the community discovery algorithm based on modularity optimization at present: passing through
Optimization Q value improves the main thought that modularity is this kind of algorithm, typical algorithm have Newman fast algorithm (FN algorithm),
Louvain algorithm.The GN algorithm and spectral method of Newman.In recent years, the research of multitiered network gradually grows up, and then goes out
Showed many multilayer community discovery algorithms: multilayer α-core hash cluster abnormal data community discovery algorithm is based on multilayered particles
The community discovery algorithm of group, CLECC algorithm, multitiered network part community discovery algorithm and by comparing the pass between node degree
System is to find local community structure in multitiered network etc..
Summary of the invention
In order to solve the above technical problems, the present invention proposes that a kind of social networks personage based on multitiered network community division is closed
It is analysis method, suitable for having the live network community structure analysis of overlapping corporations, parameter uncertainty can be effectively prevented from
Problem, community division result are more accurate.
The technical solution adopted by the present invention are as follows: the social networks character relation analysis side based on multitiered network community division
Method characterized by comprising
S1, people information data are obtained by Twitter open platform;The mission bit stream data pass through tweepy packet
Interface obtains.
S2, building multitiered network;Include: friend relation network, forwards relational network, thumb up relational network;
S3, by friend relation network that step S2 is constructed, forward relational network, thumb up relational network and merge;
S4, community division is carried out to the fused multitiered network of step S3;
S5, according to being completely embedded between corporations' node or sparse feature, analysis obtains task nexus.
2, the social networks character relation analysis method according to claim 1 based on multitiered network community division,
It is characterized in that, step S1.
Further, friend relation network struction process are as follows:
A1, key figure is determined;
A2, call tweepy packet interface acquisition key figure essential information and first layer good friend's id list;
A3, tweepy packet interface is called to the user in one layer of good friend's id list, acquires second layer good friend's id list;
A4, the user in key figure, first layer good friend's id list, the user in second layer good friend's id list are added to
In total user list, and friend relation between the user in total user list is determined using each user's buddy list;
A5, using obtained friend relation as side information, user list is deposited into figure as nodal information, obtains good friend
Relational network figure.
Forward the building process of relational network are as follows: based on the nodal information in friend relation network, utilize tweepy packet
Interface obtains the event for whether having mutual forwarding to push away text between user;Using between user it is existing it is mutual forwarding push away text event as
Side information obtains forwarding relational network figure.
Thumb up the building process of relational network are as follows: based on the nodal information in friend relation network, utilize tweepy packet
Interface obtains whether having the event thumbed up mutually between user, using the event thumbed up mutually between user as side information, obtains
Thumb up relational graph.
Further, the neighbours that node is given in the fused multitiered network of step S3, are defined as follows:
MN (x, α)=y | card ({ l:<x, y, l>∈ E ∨<y, x, l>∈ E })>=α
Wherein, MN (x, α) indicates the neighbours for having α layers or α layers relation above with node x, and x ∈ V, V are a non-empty nodes
Collection, y indicate that node, x ≠ y, card refer to the size of set element, and α is the specified cyberrelationship number of plies.
Further, community division process described in step S4 are as follows:
B1, for every a pair of (x, y), as x ∈ MN (y), calculate CLEDCC (x, y), record initial modularization degree Q and
Community division situation;
B2, in all relationships, it is the smallest a pair of (x, y) to remove CLEDCC (x, y) value, when existing simultaneously multiple minimum values
When, randomly select one;
B3, CLEDCC (x, y) value is updated, recalculates the side that CLEDCC (x, y) value may be changed, logging modle degree
Q and community division situation;
B4, step S2 is repeated, until the side in multitiered network is all removed;
Modularization degree Q after B5, each community division of comparison therefrom finds maximum Q value and corresponding best corporations
Dividing condition.
Further, which is characterized in that CLEDCC (x, y) calculating formula are as follows:
Wherein, N is the maximum relationship number of multidimensional social networks;α is the relationship number of plies;ωα=2 α/(N (N+1)).
Beneficial effects of the present invention: the method for the present invention, it is contemplated that the sociability of social media personage passes through building
Social medias personage's multilayer relational network such as twitter, more traditional single network can more reflect locating for social media personage
Live network;The method for having supplied traditional single analysis user information text has ignored social networks network to character relation
The defect of dissection;Multilayer social networks community discovery algorithm (CLEDCC) of the present invention simultaneously based on side cluster
More traditional Louvain is more suitable for the live network community structure analysis with overlapping corporations, poly- with the cross-layer side of same type
Class coefficient (CLECC) algorithm is compared, and parameter uncertainty problem can be effectively prevented from, and community division result is more accurate.
Detailed description of the invention
Fig. 1 is the solution of the present invention flow chart.
Specific embodiment
The prior art of the present invention is briefly described first:
1, data acquisition technology
The data acquisition interface of Twitter providing method, the i.e. Python packet of Tweepy, can be extracted using this interface
The essential information of twitter user, friend relation list forward relationship, thumb up the information such as relationship to construct the multilayer between user
Complex network.
2, community discovery technology
Community, intuitively from the point of view of refer to some pods in network, the connection squad between node inside each community
Closely, each intercommunal connection is relatively sparse.Community discovery is a basic task in social network analysis,
That is: a network is given, the process of community structure is found out, " community " is unified for " corporations " below, in order in the present invention
The description of appearance.
Louvain algorithm: Louvain algorithm is a kind of algorithm for being based on multi-layer optimized modularity (Modularity),
Its advantages are quick, accurate, it is considered to be best one of the community discovery algorithm of performance.Modularity function initially by with
In the quality for measuring community discovery arithmetic result, it can portray the tightness degree of the community of discovery.Since can so portray society
The tightness degree of group, can also be used to as a majorized function, i.e., by the society where its some neighbour of node join
In group, if it is possible to promote the modularity (modularity) of existing community structure.
Modularity is defined as follows:
Wherein m is the total quantity on side in figure, kiIndicate all the sum of company's side right weights for being directed toward node i, kjSimilarly.Ai,jIt indicates
Node i, company's side right weight between j.When network is not weighted graph, the weight on all sides can regard 1 as.
Two step Iterative Designs of Louvain algorithm:
Most start, each ancestor node regards an independent corporations as, and company's side right weight in corporations is 0.
1, all nodes in algorithm scan data are measured for all neighbor nodes of each node traverses node
The node is added the income of modularity brought by the corporations where its neighbor node.And select the neighbours of corresponding maximum return
The corporations where it are added in node.This proceduring corporations' ownership for repeating to instruct each node is not becoming
Change.
2, the corporations formed in step 1 are folded, each corporations is folded into a single-point, it is new to calculate separately these
The sum of company's side right weight between " corporations' point " and the weight of company's side right between all the points in corporations for generating.For next round
Step 1.Until the modularity of entire figure no longer changes.
The sharpest edges of the algorithm are exactly that speed is fast, effect is good.The time complexity of each iteration of step 1 is O (n), n
For the quantity on the side in input data.The time complexity of step 2 is O (m+n), and m is the number at epicycle iteration midpoint.
CLECC algorithm: the algorithm divided suitable for network community proposes cross-layer side cluster coefficients (cross-layer
Edge clustering coefficient) it is defined as follows:
Wherein: molecule is the triangle number of rings composed by y>and the multilayer neighbor node that they are common by side<x;Denominator be by
The triangle number of rings that the multilayer neighbor node of side<x, y>and all of which may form.CLEC is by node x, and y's is common more
The ratio of layer neighbours' number and their all possible neighbours' numbers describes the tightness degrees of two nodes, can be used as multilayer community network
In cluster intensity index between two nodes.Wherein parameter alpha is adjustable, and when network is sparse, α takes smaller value that can obtain preferably
Community division result;When network is dense, α takes the larger value that can obtain better community division result.
Algorithm flow:
Input: multilayer social networks MSN.
Output: the community division situation of MSN.
1, it for every a pair of (x, y), as x ∈ MN (y), calculates CLECC (x, y), records initial modularization degree Q and society
Group's dividing condition;
2, in all relationships, the smallest a pair (x, y) of CLECC value is removed, when existing simultaneously multiple minimum values, at random
Choose one;
3, CLECC (x, y) value is updated, the side that CLECC value may be changed, that is, last removed side are recalculated
All neighbor nodes, logging modle degree Q and community division situation;
4, step 2 is repeated, until the side in MSN is all removed;
Modularization degree Q after comparing each community division therefrom finds maximum Q value and corresponding best community division
Situation.
For convenient for those skilled in the art understand that technology contents of the invention, with reference to the accompanying drawing to the content of present invention into one
Step is illustrated.
Existing social networks character relation analyzes the information and text information based on personage itself being generally used, example
Such as the brief introduction of Twitter user, user user of interest, the bean vermicelli of user, the topic of user's concern, user, which sends out, pushes away text,
The relationship of social media personage is analyzed with this.This is easy the network social intercourse for ignoring social media.The present invention makes full use of society
The character relation information of network is handed over, the character relation multitiered network of social personage, such as friend relation is constructed, thumbs up relationship, turn
Hair relationship etc. analyzes the community structure property of the network using community detecting algorithm to realize that character relation is analyzed, to traditional text
This information analysis method is supplied.
As shown in Figure 1, realization process of the invention are as follows:
1, social media character data is acquired
In general, to obtain webpage information needs to design targeted crawler strategy and tool, traditional reptile instrument is all
It is that information scratching is carried out by webpage.Twitter provides open Python packet tweepy for developer, and developer only needs
Registration, which can be authorized to, on Twitter open platform calls API, obtains relevant information.Therefore, by tweepy packet to data
It is acquired and has not only saved the time, but also reduce the link of data cleansing.The people information data that will acquire are stored in
In MongoDB.
2, multitiered network constructs
The character relation network that the present invention is built includes friend relation network, forwards relational network, thumbs up relational network.
Friend relation network establishment step are as follows:
A1, key figure, the i.e. central person of relational network are determined.
A2, call tweepy packet interface acquisition key figure essential information and first layer good friend's id list.
A3, tweepy packet interface is called to obtain the second layer good friend id list user in one layer of buddy list.
A4, by key figure, first layer good friend's id list, the user in second layer good friend's id list is added to total user's column
In table, and friend relation between the user in total user list is determined using the buddy list of each user.
A5, using obtained friend relation as side information, user list is deposited into figure as nodal information, obtains good friend
Relational network figure.
Forward relational network build process are as follows: based on the nodal information in friend relation network, connect using tweepy packet
Mouth obtains the event for whether having mutual forwarding to push away text between user.Such as A, B are forwarded over mutually pushes away text, then exist between A, B and turn
Hair relationship is forwarded relationship as side information, obtains forwarding relational network figure.
It thumbs up relational network build process similarly: based on the nodal information in friend relation network, utilizing tweepy packet
Interface obtains whether having the event thumbed up mutually between user, using the event thumbed up mutually as side information, obtains thumbing up relationship
Figure.
3, multitiered network merges
Friend relation network is obtained in multitiered network building, forward relational network and thumbs up relational network as relatively independent
Network can analyze network community structure one by one, but it is not comprehensive enough to obtain information.The present invention melts the network of these three levels
It is combined to get to a network, there are many types on its side, not only represent a kind of character relation, but respectively represent
Friend relation, forwarding relationship and thumbs up relationship.Such network is the multilayer complex network that the present invention needs to construct.
4, community division
Group dividing method in embodiment of the present invention is using a kind of multilayer social networks corporations hair based on side cluster
Existing algorithm (CLEDCC).The algorithm synthesis considers that the relationship of any two nodes neighbors and node itself in every layer of network of personal connections is strong
It is weak, it can be effectively prevented from parameter uncertainty problem, and more than the community division result of cross-layer side cluster coefficients (CLECC) algorithm
Precisely.
In multilayer social networks (multi-layered social network), a ternary<V, E, L>group are defined.
Wherein, V is a non-empty node collection;E is a triple<x, y, l>, x, y ∈ V, l ∈ L, x ≠ y, for any two tuple
<x, y, l>∈ E,<x ', y ', l '>∈ E, if x=x ', y=y ', then l ≠ l ';L is the set of a number of plies.In multilayer society
Hand in network,<x, y, l>refer in multitiered network have in l layers of social networks one article from node x to the line of node y.This
Meaning that in network might have between any node | L | plant different relationships, thus triple<x, y, l>can be with systematic mathematical
Change all relationships of all nodes in ground description multitiered network.
Social networks is the figure being made of node and line, and for single layer social networks, the neighbours for giving node x are fixed
Justice is as follows:
N (x)=y |<x, y>∈ E ∨<y, x>∈ E }
Wherein: line of the E between nodes.
In multilayer social networks, MN (x, α) refers to the neighbours for having α layers or α layers relation above with node x ∈ V, and definition is such as
Under:
MN (x, α)=y | card ({ l:<x, y, l>∈ E ∨<y, x, l>∈ E })>=α
Wherein: card refers to the size of set element, and α is the specified cyberrelationship number of plies.In single layer network, two nodes
Between there is the line to be then neighbours.And in multitiered network, if a node is the neighbours of another specified node, two nodes are wanted
Satisfaction at least has line in α network layer.
The algorithm proposes cross-layer side difference cluster coefficients (cross- on the idea basis of the CLECC of Brodka et al.
layer edge differential clustering coefficient).A difference multilayer corporations neighbours are constructed first
DMN(x,α)(differential muti-layered neighbors):
By two formula above, comprehensively consider α and take relationship all over all probable values, and carry out weight processing, thus construct across
Layer side difference cluster coefficients are as follows:
Wherein: N is the maximum relationship number of multidimensional social networks;α is the relationship number of plies;ωα=2 α/(N (N+1)).
Algorithm flow:
Input: multilayer social networks MSN.
Output: the community division situation of MSN.
B1, for every a pair of (x, y), as x ∈ MN (y), calculate CLEDCC (x, y), record initial modularization degree Q and
Community division situation;X ∈ MN (y) indicates that node x belongs to the multilayer social activity neighbours of node y.
B2, in all relationships, remove CLEDCC value it is the smallest a pair (x, y), when existing simultaneously multiple minimum values, with
Machine chooses one;
B3, CLEDCC (x, y) value is updated, recalculates the side that CLEDCC value may be changed, that is, the last time is moved
All neighbor nodes of flash trimming, logging modle degree Q and community division situation;
B4, step B2 is repeated, until the side in MSN is all removed;
Modularization degree Q after B5, each community division of comparison therefrom finds maximum Q value and corresponding best corporations
Dividing condition.
The twitter multitiered network that the algorithm is applied to construct can be obtained to the community division situation of the network.
5, community division interpretation of result character relation is utilized
By one of the basic structure property that community structure is in complex network, there are same corporations' interior nodes to be completely embedded,
Different corporations' intermediate nodes connect sparse feature.According to this feature it is found that in twitter multilayer relational network, same corporations
Interior user member necessarily have certain denominator, than if any same interest, identical concern topic or identical society
Circle etc. is handed over, such policymaker can carry out personalized recommendation, customization according to the character relation of the twitter user of each corporations
Property service etc..For example, the personage of same community may be interested in the topic of artificial intelligence field, and some are paid close attention to
The scholar and worker of work smart field.The user of this corporation can be so classified as " intelligently feeling manual
The people of interest ", twitter can recommend page to recommend some accounts in relation to artificial intelligence or news to push away text to it.
The present embodiment is only illustrated for a kind of this social networks by twitter, in practical applications side of the invention
Method, which also may extend to other, can construct the network of social networks, such as facebook.
Those of ordinary skill in the art will understand that the embodiments described herein, which is to help reader, understands this hair
Bright principle, it should be understood that protection scope of the present invention is not limited to such specific embodiments and embodiments.For ability
For the technical staff in domain, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made
Any modification, equivalent substitution, improvement and etc. should be included within scope of the presently claimed invention.
Claims (8)
1. the social networks character relation analysis method based on multitiered network community division characterized by comprising
S1, people information data are obtained by Twitter open platform;
S2, building multitiered network;Include: friend relation network, forwards relational network, thumb up relational network;
S3, by friend relation network that step S2 is constructed, forward relational network, thumb up relational network and merge;
S4, community division is carried out to the fused multitiered network of step S3;
S5, according to being completely embedded between corporations' node or sparse feature, analysis obtains task nexus.
2. the social networks character relation analysis method according to claim 1 based on multitiered network community division, special
Sign is that the mission bit stream data of step S1 are obtained by the interface of tweepy packet.
3. the social networks character relation analysis method according to claim 2 based on multitiered network community division, special
Sign is, friend relation network struction process are as follows:
A1, key figure is determined;
A2, call tweepy packet interface acquisition key figure essential information and first layer good friend's id list;
A3, tweepy packet interface is called to the user in one layer of good friend's id list, acquires second layer good friend's id list;
A4, the user in key figure, first layer good friend's id list, the user in second layer good friend's id list are added to total use
In the list of family, and friend relation between the user in total user list is determined using each user's buddy list;
A5, using obtained friend relation as side information, user list is deposited into figure as nodal information, obtains friend relation
Network.
4. the social networks character relation analysis method according to claim 3 based on multitiered network community division, special
Sign is, forwards the building process of relational network are as follows: based on the nodal information in friend relation network, is connect using tweepy packet
Mouth obtains the event for whether having mutual forwarding to push away text between user;Mutual forwarding existing between user is pushed away into the event of text as side
Information obtains forwarding relational network figure.
5. the social networks character relation analysis method according to claim 4 based on multitiered network community division, special
Sign is, thumbs up the building process of relational network are as follows: based on the nodal information in friend relation network, is connect using tweepy packet
Mouth obtains whether having the event thumbed up mutually between user, using the event thumbed up mutually between user as side information, obtains a little
Praise relational graph.
6. the social networks character relation analysis method according to claim 4 based on multitiered network community division, special
Sign is, the neighbours of node are given in the fused multitiered network of step S3, are defined as follows:
MN (x, α)=y | card ({ l:<x, y, l>∈ E ∨<y, x, l>∈ E })>=α
Wherein, MN (x, α) indicates the neighbours for having α layers or α layers relation above with node x, and x ∈ V, V are a non-empty node collection, y
Indicate that node, x ≠ y, card refer to the size of set element, α is the specified cyberrelationship number of plies.
7. the social networks character relation analysis method according to claim 6 based on multitiered network community division, special
Sign is, community division process described in step S4 are as follows:
B1, for every a pair of (x, y), as x ∈ MN (y), calculate CLEDCC (x, y), record initial modularization degree Q and corporations
Dividing condition;
B2, in all relationships, it is the smallest a pair of (x, y) to remove CLEDCC (x, y) value, when existing simultaneously multiple minimum values,
Randomly select one;
B3, update CLEDCC (x, y) value, recalculate the side that CLEDCC (x, y) value may be changed, logging modle degree Q and
Community division situation;
B4, step S2 is repeated, until the side in multitiered network is all removed;
Modularization degree Q after B5, each community division of comparison therefrom finds maximum Q value and corresponding best community division
Situation.
8. the social networks character relation analysis method according to claim 7 based on multitiered network community division, special
Sign is, CLEDCC (x, y) calculating formula are as follows:
Wherein, N is the maximum relationship number of multidimensional social networks;α is the relationship number of plies;ωα=2 α/(N (N+1)).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910061662.2A CN109815414A (en) | 2019-01-23 | 2019-01-23 | Social networks character relation analysis method based on multitiered network community division |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910061662.2A CN109815414A (en) | 2019-01-23 | 2019-01-23 | Social networks character relation analysis method based on multitiered network community division |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109815414A true CN109815414A (en) | 2019-05-28 |
Family
ID=66604863
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910061662.2A Pending CN109815414A (en) | 2019-01-23 | 2019-01-23 | Social networks character relation analysis method based on multitiered network community division |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109815414A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110275941A (en) * | 2019-06-21 | 2019-09-24 | 福州大学 | A kind of mail method for digging based on roughening with local overlay module degree |
CN110807129A (en) * | 2019-09-29 | 2020-02-18 | 北京淇瑀信息科技有限公司 | Method and device for generating multilayer user relationship graph set and electronic equipment |
CN111581530A (en) * | 2020-05-07 | 2020-08-25 | 北京互金新融科技有限公司 | Community division method and device |
CN112052404A (en) * | 2020-09-23 | 2020-12-08 | 西安交通大学 | Group discovery method, system, device and medium for multi-source heterogeneous relation network |
CN113111114A (en) * | 2021-04-21 | 2021-07-13 | 北京易数科技有限公司 | Data processing method, device, medium and electronic equipment based on social network |
CN113269653A (en) * | 2021-06-18 | 2021-08-17 | 北京市科学技术情报研究所 | Social network management method and system based on circled thought |
CN113486258A (en) * | 2021-07-05 | 2021-10-08 | 北京易数科技有限公司 | Data analysis method, device, medium and electronic equipment based on social network |
CN114826870A (en) * | 2022-04-22 | 2022-07-29 | 西北工业大学 | Rumor propagation source tracing method based on multilayer social platform |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105141499A (en) * | 2015-07-03 | 2015-12-09 | 电子科技大学 | Social network relationship recommendation method based on privacy degree and publicity degree |
CN107122447A (en) * | 2017-04-25 | 2017-09-01 | 南京邮电大学 | The network searching system and control method of a kind of multi-data source fusion based on preference |
CN107767279A (en) * | 2017-10-12 | 2018-03-06 | 燕山大学 | A kind of average weighted personalized friend recommendation method based on LDA |
CN109062914A (en) * | 2017-06-12 | 2018-12-21 | 东软集团股份有限公司 | User's recommended method and device, storage medium and server |
-
2019
- 2019-01-23 CN CN201910061662.2A patent/CN109815414A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105141499A (en) * | 2015-07-03 | 2015-12-09 | 电子科技大学 | Social network relationship recommendation method based on privacy degree and publicity degree |
CN107122447A (en) * | 2017-04-25 | 2017-09-01 | 南京邮电大学 | The network searching system and control method of a kind of multi-data source fusion based on preference |
CN109062914A (en) * | 2017-06-12 | 2018-12-21 | 东软集团股份有限公司 | User's recommended method and device, storage medium and server |
CN107767279A (en) * | 2017-10-12 | 2018-03-06 | 燕山大学 | A kind of average weighted personalized friend recommendation method based on LDA |
Non-Patent Citations (2)
Title |
---|
张宗辉: ""基于社团的社交网络影响力传播分析"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
袁辉辉: ""基于多层社会网络的信息扩散研究"", 《中国优秀硕士论文全文数据库 信息科技辑》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110275941A (en) * | 2019-06-21 | 2019-09-24 | 福州大学 | A kind of mail method for digging based on roughening with local overlay module degree |
CN110807129A (en) * | 2019-09-29 | 2020-02-18 | 北京淇瑀信息科技有限公司 | Method and device for generating multilayer user relationship graph set and electronic equipment |
CN110807129B (en) * | 2019-09-29 | 2024-02-20 | 北京淇瑀信息科技有限公司 | Method and device for generating multi-layer user relation graph set and electronic equipment |
CN111581530A (en) * | 2020-05-07 | 2020-08-25 | 北京互金新融科技有限公司 | Community division method and device |
CN112052404A (en) * | 2020-09-23 | 2020-12-08 | 西安交通大学 | Group discovery method, system, device and medium for multi-source heterogeneous relation network |
CN112052404B (en) * | 2020-09-23 | 2023-08-15 | 西安交通大学 | Group discovery method, system, equipment and medium of multi-source heterogeneous relation network |
CN113111114A (en) * | 2021-04-21 | 2021-07-13 | 北京易数科技有限公司 | Data processing method, device, medium and electronic equipment based on social network |
CN113269653A (en) * | 2021-06-18 | 2021-08-17 | 北京市科学技术情报研究所 | Social network management method and system based on circled thought |
CN113269653B (en) * | 2021-06-18 | 2024-03-29 | 北京市科学技术情报研究所 | Social network management method and system based on layering thought |
CN113486258A (en) * | 2021-07-05 | 2021-10-08 | 北京易数科技有限公司 | Data analysis method, device, medium and electronic equipment based on social network |
CN114826870A (en) * | 2022-04-22 | 2022-07-29 | 西北工业大学 | Rumor propagation source tracing method based on multilayer social platform |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109815414A (en) | Social networks character relation analysis method based on multitiered network community division | |
Newman | The structure and function of complex networks | |
CN103379158B (en) | The method and system of commending friends information in a kind of social networks | |
Chakrabarti et al. | Graph mining: Laws, generators, and algorithms | |
Budak et al. | Structural trend analysis for online social networks | |
CN106503230A (en) | A kind of microblogging based on complex network propagates source tracing method | |
CN105849764A (en) | Systems and methods for identifying influencers and their communities in a social data network | |
Nguyen et al. | Least cost influence in multiplex social networks: Model representation and analysis | |
Huang et al. | Top-k structural diversity search in large networks | |
CN107679239B (en) | Personalized community recommendation method based on user behaviors | |
Ahmed et al. | Space-efficient sampling from social activity streams | |
Wang et al. | Relational graph neural network with neighbor interactions for bundle recommendation service | |
CN104317881B (en) | One kind is based on the authoritative microblogging method for reordering of user's topic | |
Hashmi et al. | Are all social networks structurally similar? | |
CN107506455A (en) | A kind of friend recommendation method for merging user social contact circle similarity | |
Liu et al. | A novel hybrid-jump-based sampling method for complex social networks | |
CN116383519A (en) | Group recommendation method based on double weighted self-attention | |
CN107346333A (en) | A kind of online social networks friend recommendation method and system based on link prediction | |
CN103051476B (en) | Topology analysis-based network community discovery method | |
Kou et al. | Bignn: A bilateral-branch graph neural network to solve popularity bias in recommendation | |
Wu et al. | A directed link prediction method using graph convolutional network based on social ranking theory | |
Cai et al. | Knowledge-aware graph collaborative filtering for recommender systems | |
Li et al. | Heterogeneous information diffusion model for social recommendation | |
CN106844790A (en) | A kind of quick Temporal query method for user's group analysis in social networks | |
Pan et al. | Detecting link communities based on local approach |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190528 |
|
RJ01 | Rejection of invention patent application after publication |