CN104504264B - Visual human's method for building up and device - Google Patents
Visual human's method for building up and device Download PDFInfo
- Publication number
- CN104504264B CN104504264B CN201410814330.4A CN201410814330A CN104504264B CN 104504264 B CN104504264 B CN 104504264B CN 201410814330 A CN201410814330 A CN 201410814330A CN 104504264 B CN104504264 B CN 104504264B
- Authority
- CN
- China
- Prior art keywords
- node
- account
- visual human
- similarity
- rho
- 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.)
- Active
Links
- 230000000007 visual effect Effects 0.000 title claims abstract description 76
- 238000000034 method Methods 0.000 title claims abstract description 45
- 241000282412 Homo Species 0.000 claims description 6
- 239000006185 dispersion Substances 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims 1
- 230000006399 behavior Effects 0.000 description 14
- 230000000694 effects Effects 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
- G16Z99/00—Subject matter not provided for in other main groups of this subclass
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a kind of visual human's method for building up of Behavior-based control daily record and device.Visual human's method for building up includes：Account and landing time corresponding with account are extracted in subordinate act daily record, end message is logged in；According to the similarity between the situation calculating account that appearance is cooperateed between account, construct the connected graph that account is characterized with node, and the side characterized with the length on the side between node between the similarity between account, node is shorter, similarity is higher between the account that node is characterized；Node in the connected graph is clustered, visual human is set up according to cluster result.Device is set up the invention further relates to a kind of visual human.The visual human's method for building up and device Behavior-based control daily record of the present invention sets up visual human, and complexity is low, and accuracy rate is high, is suitable for handling big data.
Description
Technical field
The present invention relates to technical field of data processing, more particularly to a kind of Behavior-based control daily record visual human's method for building up and
Device.
Background technology
Currently, instant messaging, Email, online game, P2P software downloads, network forum, E-Recruit, electronics business
Business is merchandised, and the various network services such as predetermined air ticket hotel of network bring great convenience to the life of the network user.Various networks
Service can typically distribute an account number to each user, and account is associated with the log-on message of user and to enter to each user
Row record and identification, such as the instant communication number (such as QQ accounts) or e-mail address of the network user, online game account number,
Forum logs in account number, and P2P software account numbers etc..
Each network user is owned by the various account of type, and the account number for the flood tide that the substantial amounts of network user then brings
According to for relevant departments, effectively management network subscriber information has become difficult task.For effectively management network user letter
Breath, realizes analysis to network account attaching relation, i.e. which account number belongs to same person (visual human), as needing solution badly
Certainly the problem of.
When prior art is the problem of in face of building visual human, attributes match mode is attributed to mostly.The scheme of attributes match
Approximately as：
A the rule of network account attributes match) is specified, is matched in the case of which kind of with which attribute, and accordingly
The match is successful decision method.Such as, when matching a QQ account number and during Taobao's account number, if " name " of two account numbers and
The editing distance (edit distance) of " contact method " two fields is respectively less than 3, then it is assumed that the match is successful for the two account numbers.
B) according to the situation of attributes match, the degree (similarity) of same person is belonged between structure account number.And final root
Tell which account number belongs to same person according to similarity.Such as, in upper example, then think to belong to same as long as the match is successful
People.
But, there is following situation in real life：
1. often occurring the situation of attribute missing in account data, such as it only fill in part property value when account is registered.
2. different types of account data, shared attribute is few.And in shared attribute, not necessarily can be used for attribute
Matching.
3. different types of account data, to same semantic attribute difference, it is necessary to align, which in turns increases difficulty
Degree.Such as in A class account numbers, the corresponding field of name is exactly " name " this field, but in B class account numbers, name is actual
On be to be represented with " surname " and " name " two fields.
4. in actual account number data, the confidence level of property value is not very high.For example, in default of real-name authentication, may
There is the false situation of identification card number.
5. needing to carry out the comparison of properties level, complexity is higher.
These situations make it that the process of attributes match is complicated, computationally intensive and actual result is undesirable, in particular for big
When measuring data processing, the degree of accuracy is relatively low.
The content of the invention
Therefore, it is an object of the invention to provide a kind of visual human's method for building up of Behavior-based control daily record, solve because of account number
The type visual human brought such as various builds the problem of complicated, the degree of accuracy is low.
Another object of the present invention is to provide a kind of visual human of Behavior-based control daily record to set up device, solve because of account number class
The type visual human brought such as various builds the problem of complicated, the degree of accuracy is low.
To achieve the above object, the invention provides a kind of visual human's method for building up, comprise the following steps：
Account and landing time corresponding with account are extracted in subordinate act daily record, end message is logged in；
According to the similarity between the situation calculating account that appearance is cooperateed between account, the company that account is characterized with node is constructed
Logical figure, and the side characterized with the length on the side between node between the similarity between account, node is shorter, what node was characterized
Similarity is higher between account；
Node in the connected graph is clustered, visual human is set up according to cluster result.
Wherein, the factor beyond the situation of appearance is also cooperateed with to calculate the similarity between the account between introducing account.
Wherein, the process that the node in the connected graph is clustered is comprised the following steps：
The local density Rho of each node is obtained respectively, and the length that Rho is defined as connecting this node is predefined less than some
The number of value Dc adjacent side；
The dispersion Delta, Delta for obtaining each node respectively are defined as all higher Rho values neighbours of connection of this node
The length of side of most short side in the adjacent side of node；If in the absence of such neighbor node, taking the length of side of the most long adjacent side of this node.
By the Centroid that the node identification that Rho values and Delta values are respectively higher than predetermined threshold value R_T and D_T is class；
Non-central node is classified as that the non-central nodal distance is most short and Rho values are higher than the center of the non-central node
Class belonging to node；
Each mutually similar node together constitutes with a visual human, that is, belongs to same visual human.
Wherein, the node in the connected graph is clustered using K-Means methods or hierarchy clustering method.
Wherein, in addition to merge all visual humans and account corresponding with visual human and turn into Virtual Human Data storehouse.
Device is set up present invention also offers a kind of visual human, including：
Information extraction unit, for extracting account and landing time corresponding with account in subordinate act daily record, logging in terminal
Information；
Connected graph structural unit, for according to the similarity between the situation calculating account that appearance is cooperateed between account, structure
The connected graph that account is characterized with node is made, and is characterized with the length on the side between node between the similarity between account, node
Side it is shorter, similarity is higher between the account that node is characterized；
Visual human sets up unit, for being clustered to the node in the connected graph, sets up virtual according to cluster result
People.
Wherein, in addition to external model introduce unit, for introducing the factor beyond the situation that appearance is cooperateed between account
Calculate the similarity between the account.
Wherein, the process that the node in the connected graph is clustered is comprised the following steps：
The local density Rho of each node is obtained respectively, and the length that Rho is defined as connecting this node is predefined less than some
The number of value Dc adjacent side；
The dispersion Delta, Delta for obtaining each node respectively are defined as all higher Rho values neighbours of connection of this node
The length of side of most short side in the adjacent side of node；If in the absence of such neighbor node, taking the length of side of the most long adjacent side of this node.
By the Centroid that the node identification that Rho values and Delta values are respectively higher than predetermined threshold value R_T and D_T is class；
Non-central node is classified as that the non-central nodal distance is most short and Rho values are higher than the center of the non-central node
Class belonging to node；
Each mutually similar node together constitutes with a visual human.
Wherein, the node in the connected graph is clustered using K-Means methods or hierarchy clustering method.
Wherein, in addition to visual human's combining unit, turn into for merging all visual humans and account corresponding with visual human
Virtual Human Data storehouse.
In summary, visual human is set up in visual human's method for building up of the invention and device Behavior-based control daily record, and complexity is low,
Accuracy rate is high, is suitable for handling big data.
Brief description of the drawings
In accompanying drawing,
Fig. 1 is the flow chart of the preferred embodiment of visual human's method for building up one of the present invention；
Fig. 2 is the logical schematic of the preferred embodiment of visual human's method for building up one of the present invention；
Fig. 3 is the Rho value-Delta Distribution value schematic diagrames in the preferred embodiment of visual human's method for building up one of the present invention；
Fig. 4 is the structural representation that visual human of the present invention sets up the preferred embodiment of device one.
Embodiment
Below in conjunction with the accompanying drawings, it is described in detail by the embodiment to the present invention, technical scheme will be made
And its advantage is apparent.
Referring to Fig. 1, it is the flow chart of the preferred embodiment of visual human's method for building up one of the present invention.The key step of the present invention
Including：
Account and landing time corresponding with account are extracted in subordinate act daily record, end message is logged in；
According to the similarity between the situation calculating account that appearance is cooperateed between account, the company that account is characterized with node is constructed
Logical figure, and the side characterized with the length on the side between node between the similarity between account, node is shorter, what node was characterized
Similarity is higher between account；
Node in the connected graph is clustered, visual human is set up according to cluster result.
The present invention, which can also include all visual humans of merging and account corresponding with visual human, turns into Virtual Human Data storehouse
Step.
Complicated, the low practical problem of the degree of accuracy, the present invention are built to tackle the visual human brought because account number type is various etc.
Propose a kind of analysis method of Behavior-based control daily record.User behaviors log have recorded the situation of network user's application network service, can
Gather from server end, user terminal etc..This method is based on observation as follows to reality：
1. in a period of time, the account number for having activity on same station terminal may belong to same person.We claim a certain
Multiple account numbers had activity in same terminal in the section time, were that the collaboration of these account numbers occurs.
2. situation that the collaboration of many account numbers occurs is more approximate-such as number of times is more, these account numbers belong to same person can
Energy property (claiming, similarity) is bigger.
3. in multiple account numbers that unique user possesses, always there is part account number to use more frequent.
4. between the part account number of different user, even if having collaboration to occur once in a while, it cooperates with situation about occurring not compare
The situation of appearance is cooperateed between each account number of user oneself closer to seemingly.
Referring to Fig. 2, it is the logical schematic of the preferred embodiment of visual human's method for building up one of the present invention.
Final steps in the preferred embodiment include：
Step 1. is by recording in user behaviors log is abstract【Time, terminal, account number】, so as to obtain including timestamp, account
Number ID and Termination ID data, so as to know when which account number has activity in which terminal, by each
Account count the account it is interior for a period of time had activity with other account numbers in same terminal cooperate with occurrence number, can obtain
Go out between account to cooperate with the number of times of appearance.
" number of times " is a kind of mode for weighing " situation ", using the saying of " number of times " merely to simplification is said in this embodiment
It is bright.In fact, the information such as period can also be added weighed together as weights " situation "-such as, the collaboration of quitting time goes out
Existing weight can slightly overweight work hours-work hours more likely can shared computer terminal.
Step 2. cooperates with the observation for situation occur based on above-mentioned account, calculates the similarity between account number.If abstract
Into connected graph, then the node on behalf account number in connected graph, the length on side characterizes the similarity between account number.Under normal circumstances, phase
Higher like spending, side is shorter.
Step 3. can equally regard the matching result of correspondence model as influence side if any other models, such as attributes match
One factor of length.
Step 4. is obtained after above-mentioned figure, can be calculated as below, and show which account number belongs to same person：
Step 4.1 obtains its local density Rho to each node.Rho definition is that this node's length is predetermined less than some
The number on adopted value Dc side.
The each node of step 4.2 pair, obtains its dispersion Delta.Delta is defined as all higher Rho of connection of this node
It is worth the length of side of most short side in the adjacent side of neighbor node；If in the absence of such neighbor node, taking the side of the most long adjacent side of this node
It is long.
Rho values and Delta values are respectively higher than specific threshold R_T and D_T node by step 4.3, are designated the center of class
Node.Each such one class of node on behalf, that is, a visual human.
Other non-central nodes are classified as its most short and Rho value of distance higher than the Centroid of oneself by step 4.4
That class.
Each mutually similar node of step 4.5 is that expression belongs to same visual human.Each class of correspondence is set up accordingly respectively
Visual human
To clustering method shown in final steps 4, such as K-Means, hierarchical clustering (Hierarchical can be also used
) etc clustering other conventional clustering methods, they can also reach similar result, simply in complexity or effect
It is different.User behaviors log is analyzed with reference to the clustering algorithm in the preferred embodiment, with other K-Means, hierarchical clustering
Compared etc. cluster mode, reduce the complicated degree of analysis of whole system.Meanwhile, nationality is by the two sources of Delta and Rho values
From the distribution characteristics amount of data in itself, there is provided a kind of objective reference mode selected to clusters number.
In committed step 4.3, shown class central point identification method is the Rho values and Delta values of node simultaneously above some
Respective threshold.Other methods based on Rho values or Delta values can be taken in practice.Such as Rho values are higher than 3, then delta values are in 4-5
Between, Rho values are higher than 5, then Delta values are between 5-6.
The implication combination simple examples to various values in visual human's method for building up of the present invention are described as follows below.
The length of side is characterized：Belong to the measurement of the possibility (similarity) of same person between node.
Rho is characterized:Importance of the present node to its abutment points.
Delta is characterized：If using present node as class center, the distinguishability at its other relative class center.
For example：
The length of side may be defined as：Two account numbers cooperate with the number of times (c occurred in user behaviors loga,b) (c of inverse 1/a,b).I.e.
The inverse of two account numbers number of times that priority activity is crossed in certain time in same terminal.
Rho may be defined as：In the adjacent side of present node, length is less than the quantity on parameter value Dc side.
Delta is defined as the length of side of most short side in the adjacent side of all higher Rho values neighbor nodes of connection of this node；If not depositing
In such neighbor node, then the length of side of the most long adjacent side of this node is taken.
Corresponding formula is expressed as under above-mentioned definition example：
Make c (a, b) for the account number a and b that are counted in subordinate act daily record collaboration occurrence number, then have:
The length of side between 1.a, b:
D (a, b)=1/c (a, b) [equation 1].
2. item to a all N number of neighbor node bn, n=1 ... N (N is natural number), a's
Rho values:
Wherein, X (x) definition is：If 1. x<0, then X (x)=1, otherwise X (x)=0.
3.a Delta values：
The bN that makes node a neighbor node be followed successively by b1 ..., then Delta (a) may be defined as:
1) if there is meeting Rho (bx)>Rho (a) adjacent side, then have：
Delta (a)=min d (a, bn)) | n=1..N and Rho (bn)>Rho(a)}.
2) otherwise:
Delta (a)=max { d (a, bn), n=1..N }
Particularly, for the node of not any adjacent side, when marking its class mark, their own can be directly designated, i.e.,
It is individually formed a visual human.
Trying to achieve for Delta values is relevant with Rho values, and the definition of Rho values can also use other definition sides such as common centrad
Formula.
Dc value is relevant with specific data in practice, and generally we can be after connected graph be obtained, then determines Dc's
Value.That is, as in other common cluster modes, it is an input parameter.But with the K values in K-Means
Selection unlike, the number of the selection directly determination class of K values, but Dc here can pass through Rho values and Delta values and R_
T and D_T value and the influence for weakening subjective factor, because the selection of these parameters can introduce the visitor to data self character
See and consider.
A kind of method of R_T and D_T selection is as follows.As shown in figure 3, it is that visual human's method for building up one of the present invention is preferable
The point of each in Rho value-Delta Distribution value schematic diagrames in embodiment, figure represents a node.Each point is drawn first
Rho value-Delta Distribution value figures, observe the distribution situation of Delta values (Rho values) afterwards, see that, in which value, distribution situation is sent out
Mutation is given birth to, then it is D_T (R_T) to take the value.As in Fig. 3, at d ' (r ') place, the distribution situations of Delta values there occurs interruption/dash forward
Become, then D_T (R_T) value is d ' (r ').It if data point is more, can be sampled, then be taken with the distribution map of sample point
The reference of value.
By introducing other models, such as attributes match, the matching result of correspondence model can be equally regard as the influence length of side
One factor of degree.That is, cooperateing with the factor beyond the number of times of appearance to calculate between the account between introducing account
Similarity.
With attributes match for example, it is to regard the result of attributes match as calculating side with mathematical symbolism
A long parameter.That is, a that Match (a, b) is arrived for attributes match and b account number similarity are made, then the length of side can be defined as below：
D (a, b)=f (c (a, b), match (a, b)).
By taking [equation 1] as an example, it may be selected to be specifically defined as：
The length of side introduced after attributes match model
Referring to Fig. 4, it is the structural representation that visual human of the present invention sets up the preferred embodiment of device one.The preferred embodiment
Visual human set up device including information extraction unit 1, connected graph structural unit 2, external model introduces unit 3, and visual human builds
Vertical unit 4 and visual human's combining unit 5.
Information extraction unit 1, for extracting account and landing time corresponding with account in subordinate act daily record, logging in terminal
Information；
Connected graph structural unit 2, for according to the similarity between the situation calculating account that appearance is cooperateed between account, structure
The connected graph that account is characterized with node is made, and is characterized with the length on the side between node between the similarity between account, node
Side it is shorter, similarity is higher between the account that node is characterized；
External model introduces unit 3, and the account is calculated for introducing the factor beyond the situation that appearance is cooperateed between account
Similarity between number；
Visual human sets up unit 4, for being clustered to the node in the connected graph, sets up virtual according to cluster result
People；
Visual human's combining unit 5, Virtual Human Data is turned into for merging all visual humans and account corresponding with visual human
Storehouse.
Visual human sets up the mode that unit 4 clustered to the node in connected graph and referred in preceding description to the present invention
The description of visual human's method for building up.
In the visual human's method for building up and device of the present invention, by way of analytical behavior daily record, obtained by actual analysis
The result gone out is " which account number is to belong to same person operation ".In reality system demand, user is than account number owner
It is often more meaningful, while this can be also reduced because key values such as " ID card No. " is untrue, and cause attribution of account numbers relation knot
Deviation on fruit.Analyzed with user behaviors log, add the usability of whole system-only need to account number mark, and simultaneously
It is not necessarily required to specific account attributes.The reduction of feature and above-mentioned complexity from user behaviors log, the present invention can be fitted more preferably
With under wider, in the range of the longer time, the environment of more data amount.In fact, the scope of data acquisition certainly is wider, the time
The more conference of longer, data volume make it that the actual accuracy rate of system is higher.After the present invention is according to the above-mentioned analysis to user behaviors log, cluster
The attribution of account numbers relation drawn, can combine the excessive datas such as account attributes, be further depicted as out name, address of the visual human etc.
Attribute information.
In summary, visual human is set up in visual human's method for building up of the invention and device Behavior-based control daily record, and complexity is low,
Accuracy rate is high, is suitable for handling big data.
It is described above, for the person of ordinary skill of the art, can be with technique according to the invention scheme and technology
Other various corresponding changes and deformation are made in design, and all these changes and deformation should all belong to appended right of the invention
It is required that protection domain.
Claims (8)
1. a kind of visual human's method for building up, it is characterised in that comprise the following steps：
Account and landing time corresponding with account are extracted in subordinate act daily record, end message is logged in；
According to the similarity between the situation calculating account that appearance is cooperateed between account, the connection that account is characterized with node is constructed
Figure, and the side characterized with the length on the side between node between the similarity between account, node is shorter, the account that node is characterized
Similarity is higher between number；
Node in the connected graph is clustered, visual human is set up according to cluster result；Wherein, in the connected graph
The process that node is clustered comprises the following steps：
The local density Rho of each node is obtained respectively, and Rho is defined as connecting the neighbour of the length less than predefined value Dc of this node
The number on side；
The dispersion Delta, Delta for obtaining each node respectively are defined as all higher Rho values neighbor nodes of connection of this node
Adjacent side in most short side the length of side；If in the absence of such neighbor node, taking the length of side of the most long adjacent side of this node；
By the Centroid that the node identification that Rho values and Delta values are respectively higher than predetermined threshold value R_T and D_T is class；
Non-central node is classified as that the non-central nodal distance is most short and Rho values are higher than the Centroid of the non-central node
Affiliated class；
Each mutually similar node together constitutes with a visual human.
2. visual human's method for building up as claimed in claim 1, it is characterised in that can also introduce between account and cooperate with the feelings of appearance
Factor beyond condition calculates the similarity between the account.
3. visual human's method for building up as claimed in claim 1, it is characterised in that use K-Means methods or hierarchical clustering side
Method is clustered to the node in the connected graph.
4. visual human's method for building up as claimed in claim 1, it is characterised in that also including merge all visual humans and with it is virtual
The corresponding account of people turns into Virtual Human Data storehouse.
5. a kind of visual human sets up device, it is characterised in that including：
Information extraction unit, for extracting account and landing time corresponding with account in subordinate act daily record, logging in end message；
Connected graph structural unit, for according to cooperateed between account appearance situation calculate account between similarity, construction with
Node characterizes the connected graph of account, and characterizes the side between the similarity between account, node with the length on the side between node
Shorter, similarity is higher between the account that node is characterized；
Visual human sets up unit, and for being clustered to the node in the connected graph, visual human is set up according to cluster result；Its
In, the process that the node in the connected graph is clustered is comprised the following steps：
The local density Rho of each node is obtained respectively, and Rho is defined as connecting the neighbour of the length less than predefined value Dc of this node
The number on side；
The dispersion Delta, Delta for obtaining each node respectively are defined as all higher Rho values neighbor nodes of connection of this node
Adjacent side in most short side the length of side；If in the absence of such neighbor node, taking the length of side of the most long adjacent side of this node；
By the Centroid that the node identification that Rho values and Delta values are respectively higher than predetermined threshold value R_T and D_T is class；
Non-central node is classified as that the non-central nodal distance is most short and Rho values are higher than the Centroid of the non-central node
Affiliated class；
Each mutually similar node together constitutes with a visual human.
6. visual human as claimed in claim 5 sets up device, it is characterised in that also introduces unit including external model, is used for
The factor beyond the situation of appearance is cooperateed with to calculate the similarity between the account between introducing account.
7. visual human as claimed in claim 5 sets up device, it is characterised in that use K-Means methods or hierarchical clustering side
Method is clustered to the node in the connected graph.
8. visual human as claimed in claim 5 sets up device, it is characterised in that also including visual human's combining unit, for closing
And all visual humans and account corresponding with visual human turn into Virtual Human Data storehouse.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410814330.4A CN104504264B (en) | 2014-12-08 | 2014-12-23 | Visual human's method for building up and device |
| PCT/CN2015/072487 WO2016090748A1 (en) | 2014-12-08 | 2015-02-09 | Virtual human creating method and apparatus |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410741334 | 2014-12-08 | ||
| CN2014107413344 | 2014-12-08 | ||
| CN201410814330.4A CN104504264B (en) | 2014-12-08 | 2014-12-23 | Visual human's method for building up and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104504264A CN104504264A (en) | 2015-04-08 |
| CN104504264B true CN104504264B (en) | 2017-09-01 |
Family
ID=52945661
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410814330.4A Active CN104504264B (en) | 2014-12-08 | 2014-12-23 | Visual human's method for building up and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN104504264B (en) |
| WO (1) | WO2016090748A1 (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104965846B (en) * | 2014-12-31 | 2018-10-02 | 深圳市华傲数据技术有限公司 | Visual human's method for building up in MapReduce platform |
| RU2617918C2 (en) * | 2015-06-19 | 2017-04-28 | Иосиф Исаакович Лившиц | Method to form person's image considering psychological portrait characteristics obtained under polygraph control |
| CN106372977B (en) * | 2015-07-23 | 2019-06-07 | 阿里巴巴集团控股有限公司 | A kind of processing method and equipment of virtual account |
| CN106469413B (en) * | 2015-08-20 | 2021-08-03 | 深圳市腾讯计算机系统有限公司 | Data processing method and device for virtual resources |
| CN105224606B (en) * | 2015-09-02 | 2019-04-02 | 新浪网技术(中国)有限公司 | A kind of processing method and processing device of user identifier |
| CN105897667A (en) * | 2015-10-22 | 2016-08-24 | 乐视致新电子科技（天津）有限公司 | Device access history tracking method, apparatus, server and system |
| CN107291760A (en) * | 2016-04-05 | 2017-10-24 | 阿里巴巴集团控股有限公司 | Unsupervised feature selection approach, device |
| CN106604264A (en) * | 2017-01-04 | 2017-04-26 | 北京奇虎科技有限公司 | Application installation method and system, server, and mobile terminal |
| CN107248929B (en) * | 2017-05-27 | 2020-08-11 | 北京知道未来信息技术有限公司 | Strong correlation data generation method of multi-dimensional correlation data |
| CN110032603A (en) * | 2019-01-22 | 2019-07-19 | 阿里巴巴集团控股有限公司 | The method and device that node in a kind of pair of relational network figure is clustered |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103544289A (en) * | 2013-10-28 | 2014-01-29 | 公安部第三研究所 | Feature extraction achieving method based on deploy and control data mining |
| CN103970752A (en) * | 2013-01-25 | 2014-08-06 | 北京思博途信息技术有限公司 | Estimating method and system for amount of unique visitors |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090293121A1 (en) * | 2008-05-21 | 2009-11-26 | Bigus Joseph P | Deviation detection of usage patterns of computer resources |
| CN103368917B (en) * | 2012-04-01 | 2017-11-14 | 阿里巴巴集团控股有限公司 | Risk control method and system for network virtual users |
| CN103927307B (en) * | 2013-01-11 | 2017-03-01 | 阿里巴巴集团控股有限公司 | A kind of method and apparatus of identification website user |
-
2014
- 2014-12-23 CN CN201410814330.4A patent/CN104504264B/en active Active
-
2015
- 2015-02-09 WO PCT/CN2015/072487 patent/WO2016090748A1/en active Application Filing
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103970752A (en) * | 2013-01-25 | 2014-08-06 | 北京思博途信息技术有限公司 | Estimating method and system for amount of unique visitors |
| CN103544289A (en) * | 2013-10-28 | 2014-01-29 | 公安部第三研究所 | Feature extraction achieving method based on deploy and control data mining |
Non-Patent Citations (1)
| Title |
|---|
| 《Graph based k-means clustering》;Laurent Galluccio 等;《Signal Processing》;20120121;第92卷(第9期);摘要，第2节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2016090748A1 (en) | 2016-06-16 |
| CN104504264A (en) | 2015-04-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104504264B (en) | Visual human's method for building up and device | |
| US9059897B2 (en) | Method and apparatus to identify outliers in social networks | |
| EP4198775B1 (en) | Method and apparatus for determining abnormal users, electronic device, and storage medium | |
| CN106375339B (en) | Attack mode detection method based on event sliding window | |
| Jeong et al. | User migration across multiple social media platforms | |
| WO2016119275A1 (en) | Network account identifying and matching method | |
| CN104965846B (en) | Visual human's method for building up in MapReduce platform | |
| CN109241062A (en) | A kind of generation method and device of government data catalogue | |
| CN109951499A (en) | A kind of method for detecting abnormality based on network structure feature | |
| CN110019996A (en) | A kind of family relationship recognition methods and system | |
| Ackland et al. | Political homophily on the web | |
| Han et al. | Edge sample and discard: A new algorithm for counting triangles in large dynamic graphs | |
| US8700756B2 (en) | Systems, methods and devices for extracting and visualizing user-centric communities from emails | |
| CN112052399B (en) | Data processing method, device and computer readable storage medium | |
| Srivastava et al. | A privacy settings recommender system for online social networks | |
| KR101928822B1 (en) | System and method for computing a user's trust value of unknown device in IoT | |
| CN116049438B (en) | Knowledge graph-based group membership analysis method | |
| CN105574139A (en) | Social network friend making recommendation method and system based on double attractiveness calculation | |
| CN113810329B (en) | Method and system for detecting abnormality of mailbox account | |
| JP2009187395A (en) | Topic analyzing device, method and program | |
| CN106301880B (en) | A method and device for determining the stability of a network relationship and recommending an Internet service | |
| CN116257803A (en) | User state detection method and device, electronic equipment and storage medium | |
| CN107274058A (en) | A kind of determination methods of mechanics | |
| Yu et al. | Identifying interaction groups in social network using a game-theoretic approach | |
| CN106656758A (en) | Method and device for constructing social network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP02 | Change in the address of a patent holder |
Address after: 518057 2203/2204, Building 1, Huide Building, North Station Community, Minzhi Street, Longhua District, Shenzhen, Guangdong Province Patentee after: SHENZHEN AUDAQUE DATA TECHNOLOGY Ltd. Address before: 518057 Rooms 713, 715 and 716, 7/F, Software Building, No. 9, High-tech Middle Road, High-tech Zone, Nanshan District, Shenzhen, Guangdong Province Patentee before: SHENZHEN AUDAQUE DATA TECHNOLOGY Ltd. |
|
| CP02 | Change in the address of a patent holder |