CN110197397A - The partitioning method and device of node level - Google Patents
The partitioning method and device of node level Download PDFInfo
- Publication number
- CN110197397A CN110197397A CN201910408213.0A CN201910408213A CN110197397A CN 110197397 A CN110197397 A CN 110197397A CN 201910408213 A CN201910408213 A CN 201910408213A CN 110197397 A CN110197397 A CN 110197397A
- Authority
- CN
- China
- Prior art keywords
- destination node
- node
- reference mode
- association
- degree
- 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
Classifications
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
Landscapes
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The disclosure is directed to a kind of partitioning method and devices of node level.It include the destination node and reference mode associated with the destination node in the relationship subgraph of the destination node, and the reference mode is updated with data this method comprises: obtaining the relationship subgraph of the destination node currently updated；According to the degree of association between the destination node, the reference mode and the reference mode and the destination node associated with the destination node and with data update, the current scoring of the destination node is determined；The present level of the destination node is divided according to the current scoring of the destination node, the present level corresponds to the relationship subgraph of the destination node currently updated.The disclosure can based on destination node all related informations come the real-time grade for determining destination node, therefore grade classification can be realized easily and fast, to ensure that the real-time of grade classification and comprehensive.
Description
Technical field
This disclosure relates to data processing technique more particularly to a kind of division side of node level in Internet technical field
Method and device.
Background technique
The high development of internet it is deep change people's lives.In e-commerce field, as transaction
The attributive character of the Primary Actor of system, user have important influence for the processing result of electronic commerce data.
Currently, the method in e-commerce field for determining user gradation mainly includes following two:
The first, the classification to user is realized according to the preset rules pre-established.Specifically, based on default rule
When being then classified to user, regular quality has a very big impact the result of classification, since preset rules often can only
Reflect the information of user's some aspects, it is difficult to comprehensively delineate user image, therefore rating result there are certain one-sidedness and
Limitation, while there is also the problems of cumbersome trouble for a large amount of rule maintenance.
Second, the classification to user is realized by means of training pattern.Specifically, using machine learning model pair
When user is classified, it is assumed that model is suitable and trains in place, can naturally also obtain good effect, but this is needed largely
Hierarchic user is as training set, and the accuracy rate for improving result is also required to the continuous iteration of model, and therefore, it is difficult to provide in real time
The service of property.
It follows that about the determination method of user gradation, there is also many problems in the prior art, therefore need to seek
Better stage division meets the needs of e-commerce transaction.
Summary of the invention
To overcome the problems in correlation technique, the embodiment of the present disclosure provides the division methods and dress of a kind of node level
It sets.The technical solution is as follows:
According to the first aspect of the embodiments of the present disclosure, a kind of division methods of node level are provided, comprising:
The relationship subgraph of the destination node currently updated is obtained, includes the target in the relationship subgraph of the destination node
Node and reference mode associated with the destination node, and the reference mode is updated with data；
According to the destination node, it is associated with the destination node and with data update the reference mode, with
And the degree of association between the reference mode and the destination node, determine the current scoring of the destination node；
The present level of the destination node is divided according to the current scoring of the destination node, the present level is corresponding
In the relationship subgraph of the destination node currently updated.
In one embodiment, the relationship subgraph for obtaining the destination node currently updated includes:
Obtain the relationship subgraph of the destination node before updating, and the relator of the destination node before Monitoring and Update
The data variation of each reference mode in figure, the data variation include the change of increase and decrease and the node scoring of the reference mode
Change；
When the state of any one reference mode changes, the relator of the destination node after obtaining variation
Scheme, in the relationship subgraph of the destination node after the variation, the destination node is node to be updated.
In one embodiment, it is described according to the destination node, it is associated with the destination node and have data more
The degree of association between the new reference mode and the reference mode and the destination node, determines the destination node
It is current scoring include:
History scoring, the scoring of the reference mode associated with the destination node for obtaining the destination node become
The weighted value of change and the degree of association between the reference mode and the destination node；
On the basis of the scoring of the history of the destination node, according to the scoring of reference mode variation and the reference
The weighted value of the degree of association between node and the destination node determines the current scoring of the destination node.
In one embodiment, the degree of association between the reference mode and the destination node include first degree of association,
Second degree of association and the third degree of association；
The weighted value of the degree of association includes corresponding first weighted value of first degree of association, second degree of association pair
The second weighted value and the corresponding third weighted value of the third degree of association answered；
Wherein, first degree of association, second degree of association and the third degree of association are by by force to weak, first power
Weight values, second weighted value and the third weighted value are descending.
In one embodiment, the division methods further include:
Node relationships network is constructed, and each node in the node relationships network is initialized, the node
It include the relationship subgraph of the destination node in relational network.
In one embodiment, the building node relationships network includes:
Using default node as core, the associated nodes of the default node, the default node are obtained based on preset relation
Including the destination node；
The node relationships network, the node are constructed according to the associated nodes of the default node and the default node
It include the relationship subgraph of the destination node in relational network.
In one embodiment, the division methods further include:
Monitor the quantity of the reference mode increased newly in the relationship subgraph of the destination node in preset period of time and described new
The attribute of the reference mode of increasing；
Attribute according to the preset period of time, the quantity of the newly-increased reference mode and the newly-increased reference mode is true
The attribute of the fixed destination node, the attribute of the destination node indicate the risk class of the destination node.
According to the second aspect of an embodiment of the present disclosure, a kind of dividing device of node level is provided, comprising:
Module is obtained, for obtaining the relationship subgraph of the destination node currently updated, the relationship subgraph of the destination node
In include the destination node and reference mode associated with the destination node, and the reference mode have data more
Newly；
First determining module, for according to the destination node, it is associated with the destination node and with data update
The reference mode and the reference mode and the destination node between the degree of association, determine the destination node
Current scoring；
Division module divides the present level of the destination node, institute for the current scoring according to the destination node
State the relationship subgraph that present level corresponds to the destination node currently updated.
In one embodiment, the acquisition module includes:
First acquisition unit, for obtaining the relationship subgraph of the destination node before updating, and the institute before Monitoring and Update
The data variation of each reference mode in the relationship subgraph of destination node is stated, the data variation includes the increasing of the reference mode
Subtract and node scoring variation；
Second acquisition unit, for the institute when the state of any one reference mode changes, after obtaining variation
The relationship subgraph for stating destination node, in the relationship subgraph of the destination node after the variation, the destination node is to more
New node.
In one embodiment, first determining module includes:
Third acquiring unit, for obtain the destination node history score, institute associated with the destination node
State scoring variation and the weighted value of the degree of association between the reference mode and the destination node of reference mode；
Determination unit, for being become on the basis of the scoring of the history of the destination node according to the scoring of the reference mode
The weighted value of change and the degree of association between the reference mode and the destination node determines currently commenting for the destination node
Point.
In one embodiment, the degree of association between the reference mode and the destination node include first degree of association,
Second degree of association and the third degree of association；
The weighted value of the degree of association includes corresponding first weighted value of first degree of association, second degree of association pair
The second weighted value and the corresponding third weighted value of the third degree of association answered；
Wherein, first degree of association, second degree of association and the third degree of association are by by force to weak, first power
Weight values, second weighted value and the third weighted value are descending.
In one embodiment, the dividing device further include:
Module is constructed, is carried out just for constructing node relationships network, and to each node in the node relationships network
Beginningization includes the relationship subgraph of the destination node in the node relationships network.
In one embodiment, the building module includes:
4th acquiring unit, for obtaining the association of the default node based on preset relation using default node as core
Node, the default node include the destination node；
Construction unit, for constructing the node relationships according to the associated nodes of the default node and the default node
Network includes the relationship subgraph of the destination node in the node relationships network.
In one embodiment, the dividing device further include:
Monitoring modular, the number of the reference mode increased newly in the relationship subgraph for monitoring the destination node in preset period of time
The attribute of amount and the newly-increased reference mode；
Second determining module, for according to the quantity of the preset period of time, the newly-increased reference mode and described newly-increased
The attribute of reference mode determine that the attribute of the destination node, the attribute of the destination node indicate the wind of the destination node
Dangerous grade.
According to the third aspect of an embodiment of the present disclosure, a kind of dividing device of node level is provided, comprising:
Processor；
Memory for storage processor executable instruction；
Wherein, the processor is configured to the step of executing first aspect any embodiment the method.
According to a fourth aspect of embodiments of the present disclosure, a kind of computer readable storage medium is provided, calculating is stored thereon with
Machine instruction, when which is executed by processor the step of realization first aspect any embodiment the method.
The technical scheme provided by this disclosed embodiment can include the following benefits:
Disclosed technique scheme can be true in real time according to the data variation of each reference mode in the relationship subgraph of destination node
Set the goal the current scoring of node, and the newest classification of destination node is determined according to the current scoring of destination node.Due to mesh
The grade of mark node is to be determined in real time based on its all related information, therefore can realize that grade is drawn easily and fast
Point, to ensure that the real-time of grade classification and comprehensive.Compared to the stage division generallyd use at present, disclosed technique
On the one hand scheme can be avoided the one-sidedness and limitation of preset rules, be on the other hand not necessarily to the training set of preparatory mark, and
Data iteration can be realized easily and fast, therefore there is better grading effect.
It should be understood that above general description and following detailed description be only it is exemplary and explanatory, not
The disclosure can be limited.
Detailed description of the invention
The drawings herein are incorporated into the specification and forms part of this specification, and shows the implementation for meeting the disclosure
Example, and together with specification for explaining the principles of this disclosure.
Fig. 1 is the flow chart one of the division methods of the node level shown accoding to exemplary embodiment；
Fig. 2 is the schematic diagram of the relationship subgraph of the destination node shown accoding to exemplary embodiment；
Fig. 3 is the flowchart 2 of the division methods of the node level shown accoding to exemplary embodiment；
Fig. 4 is the flow chart 3 of the division methods of the node level shown accoding to exemplary embodiment；
Fig. 5 a is the module map one of the dividing device of the node level shown accoding to exemplary embodiment；
Fig. 5 b is the module map two of the dividing device of the node level shown accoding to exemplary embodiment；
Fig. 5 c is the module map three of the dividing device of the node level shown accoding to exemplary embodiment；
Fig. 5 d is the module map four of the dividing device of the node level shown accoding to exemplary embodiment；
Fig. 5 e is the module map five of the dividing device of the node level shown accoding to exemplary embodiment；
Fig. 5 f is the module map six of the dividing device of the node level shown accoding to exemplary embodiment；
Fig. 6 is the block diagram of the dividing device for node level shown accoding to exemplary embodiment.
Specific embodiment
Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to
When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment
Described in embodiment do not represent all implementations consistent with this disclosure.On the contrary, they be only with it is such as appended
The example of the consistent device and method of some aspects be described in detail in claims, the disclosure.
Technical solution provided by the embodiment of the present disclosure is related to the division methods of node level, for realizing to user node
Classification, the node include user in e-commerce system and phone associated with the user, bank card, equipment and its
Its user etc..In the related technology, commonplace user's stage division includes according to the preset rules pre-established come to user
It carries out classification and classification two ways is carried out to user by means of training pattern, but the former lacking there are one-sidedness and limitation
It falls into, and there is also be difficult to the problem of providing real-time sex service by the latter.Based on this, the embodiment of the present disclosure can be according to the pass of destination node
It is that the data variation of each reference mode in subgraph determines the current scoring of destination node in real time, and according to the current of destination node
It scores to determine the newest classification of destination node.Since the grade of destination node is true in real time based on its all related information
It is fixed, therefore can realize grade classification easily and fast, to ensure that the real-time of grade classification and comprehensive.Compared to
On the one hand the stage division generallyd use at present, disclosed technique scheme can be avoided the one-sidedness and limitation of preset rules,
On the other hand it is not necessarily to the training set of preparatory mark, and can realize data iteration easily and fast, therefore has and preferably divides
Grade effect.
Fig. 1 illustrates the flow chart of the division methods of node level provided by the embodiment of the present disclosure.According to Fig. 1
Shown, the division methods of the node level specifically comprise the following steps S101 to S103:
In step s101, the relationship subgraph of the destination node currently updated, the relationship subgraph of the destination node are obtained
In include destination node and reference mode associated with the destination node, and the reference mode have data update.
Wherein, destination node refers to the first user node such as purchaser node in e-commerce system, and reference mode
Refer to associated with first user node second user node such as seller's node or good friend's node in e-commerce system,
Bank card node, calling node and device node etc..
In step s 102, according to destination node, it is associated with destination node and with data update reference mode, with
And the degree of association between reference mode and destination node determines the current scoring of destination node.
Wherein, the degree of association between reference mode and destination node refers to strong and weak correlation degree between the two, such as general
Clearance connection, strong association and weak rigidity, or according to incidence relation made of different degrees of divide.It should be understood that here
The degree of association can be handled in practical applications by calculation system, that is, by the different degrees of association as different numerical value, so as to full
The requirement of sufficient data processing.
In step 103, the present level of destination node, the present level pair are divided according to the current scoring of destination node
It should be in the relationship subgraph of the destination node currently updated.
Wherein, the current scoring of destination node and present level both correspond to the relator of the destination node currently updated
Figure is the newest grade of the data variation based on reference mode and destination node determining in real time.
Based on above-mentioned steps S101-S103 it is found that the embodiment of the present disclosure can be according to each in the relationship subgraph of destination node
The data variation of reference mode determines the current scoring of destination node in real time, and determines mesh according to the current scoring of destination node
Mark the newest classification of node.Since the grade of destination node is determined in real time based on its all related information, energy
It is enough to realize grade classification easily and fast, to ensure that the real-time of grade classification and comprehensive.Compared to generally adopting at present
On the one hand stage division, disclosed technique scheme can be avoided the one-sidedness and limitation of preset rules, on the other hand without
The training set of preparatory mark is needed, and can realize data iteration easily and fast, therefore there is better grading effect.
Each step of disclosed technique scheme is described in detail combined with specific embodiments below.
In step s101, the relationship subgraph of the destination node currently updated, the relationship subgraph of the destination node are obtained
In include destination node and reference mode associated with the destination node, and the reference mode have data update.
Here with destination node for buyer user, reference mode includes seller user associated with the buyer user, good
For friendly user a and good friend user b, bank card and phone, carried out specifically in conjunction with specific embodiment of the Fig. 2 to this step
It is bright.Fig. 2 illustrates the schematic diagram of the relationship subgraph of destination node in the embodiment of the present disclosure.Certainly, the destination node
Relationship subgraph actual capabilities are increasingly complex, here only using simple example as explanation.As can be seen from FIG. 2, the present embodiment can be certainly
It is extracted in the node relationships network constructed in advance using buyer user as the relationship subgraph of the i.e. buyer user of relationship subgraph of core,
It becomes reconciled in the relationship subgraph of the buyer user including buyer user, seller user associated with the buyer user, good friend user a
Friendly user b and several nodes such as bank card and phone, and each node carries the scoring of oneself, in this approach just
The relationship subgraph of buyer user can be got.It will be appreciated that Fig. 2 just show in node relationships network with destination node
That is the associated part of buyer user, but it further includes more complicated relational network that it is practical, does not do specific introduction here.It is based on
This, this step S101 is specifically included:
S1011, the relationship subgraph for obtaining the destination node before updating, and the relationship subgraph of the destination node before Monitoring and Update
In each reference mode data variation, the data variation include reference mode increase and decrease and node scoring variation；
S1012, when the state of any one reference mode changes, obtain variation after destination node relator
Figure, in the relationship subgraph of the destination node after the variation, the destination node is node to be updated.
For example, the present embodiment can extract the relationship subgraph of buyer user in the node relationships network constructed in advance
It include history scoring, the seller user of buyer user as the relationship subgraph before update, in the relationship subgraph of the buyer user
The history of history scoring, the history scoring of good friend user a, the history scoring of good friend user b, the history scoring and phone of bank card
Scoring and the corresponding weighted value of the degree of association between each reference mode and buyer user, wherein weighted value is to preset
Good value.Relationship subgraph based on the buyer user obtained, real-time monitoring relationship subgraph, to judge it with the presence or absence of reference
The data situations such as the increase and decrease of node or the scoring of reference mode variation.Once monitoring that data occurs in any one node
Variation, the relationship subgraph of the buyer user after can obtaining variation is as updated relationship subgraph, the relationship of the buyer user
The current scoring of history scoring, seller user, the current scoring of good friend user a and good friend user in subgraph including buyer user
Between the current scoring of b, the current scoring of bank card and the current scoring and each reference mode and buyer user of phone
The corresponding weighted value of the degree of association, wherein weighted value generally remains constant, and but not limited to this.Based on this it is found that in the updated
In the relationship subgraph of buyer user, the scoring of buyer user is still history scoring, and associated with the buyer user sells household
The scoring of family, good friend user a and good friend user b and each reference mode such as bank card and phone is current scoring, at this time root
It can determine the current scoring of buyer user according to the current scoring of each reference mode, it is specific to determine that method is detailed in subsequent step.
In step s 102, according to destination node, it is associated with destination node and with data update reference mode, with
And the degree of association between reference mode and destination node determines the current scoring of destination node.
It here include by being associated with to the first weak degree of association, second by force with the degree of association between reference mode and destination node
For degree and the third degree of association, i.e. association, the third degree of association are corresponding during the corresponding strong association of first degree of association, second degree of association are corresponding
Weak rigidity, then the third weight of the first weighted value of first degree of association, the second weighted value of second degree of association and the third degree of association
Be worth it is descending successively successively decrease, such as the first weighted value be the 3, second weighted value be 2, third weighted value is 1.Based on this, this step
Rapid S102 is specifically included:
S1021, the history for obtaining destination node score, the scoring of reference mode associated with the destination node changes,
And the weighted value of the degree of association between reference mode and destination node；
S1022, on the basis of the scoring of the history of destination node, according to the scoring variation of reference mode and reference mode with
The weighted value of the degree of association between destination node determines the current scoring of destination node.
For example, the history that the present embodiment can obtain buyer user based on the relationship subgraph of the buyer user before update is commented
Divide, the history of seller user scores, the history of good friend user a scores and the history of the scoring of the history of good friend user b, bank card is commented
Divide and the history of phone scores, and the relationship subgraph based on the buyer user after variation obtains the current of seller user and scores, is good
The current scoring of current scoring, the current scoring and phone of bank card of the current scoring and good friend user b of friendly user a, thus
The scoring variation of each reference mode is obtained, while it is right also to obtain the institute of the degree of association between each reference mode and buyer user
The weighted value answered.In view of weighted value corresponding to the degree of association between each reference mode and buyer user may become
Change, therefore is preferably based on the relationship subgraph of the buyer user after variation to obtain, so as to avoid when obtaining the weighted value
The problem for causing the scoring of buyer user inaccurate when weighted value changes.Assuming that by tying up between buyer user and bank card
The degree of association for determining relationship generation is Qiang Guanlian, between buyer user and seller user by the degree of association that order relationship generates is
Association, between buyer user and good friend user a and good friend user b by session relationship generate the degree of association be middle association, buyer
Between user and phone by binding relationship generate the degree of association be weak rigidity, at this time buyer user it is current score can be according to such as
Lower principle determines: on the basis of the scoring of the history of buyer user, the scoring of seller user can be changed to corresponding with middle association the
The product of two weighted values 2, the scoring of good friend user a change the product of the second weighted value 2 corresponding with middle association, good friend user b
Scoring variation change corresponding with being associated with by force first with the scoring of the product of corresponding second weighted value 2 of middle association, bank card
The product of weighted value 3 and the scoring of phone change the product of third weighted value 1 corresponding with weak rigidity and going through for buyer user
Commentary on historical events or historical records point sum up operation, to obtain the current scoring of updated buyer user.Wherein, in good friend user a or good friend
User b is judged as junk user or when bank card catchs a packet, corresponding to scoring variation be negative value, be based on it at this time
The score contribution to buyer user of scoring variation and the degree of association between buyer user and determination is also negative value, i.e., can make
The score of buyer user reduces.
In step 103, the present level of destination node, the present level pair are divided according to the current scoring of destination node
It should be in the relationship subgraph of the destination node currently updated.
Here by taking node level includes diamond user, star user, public users and the user that breaks one's promise as an example, different brackets pair
There should be different scoring requirements, the present embodiment, can be true according to the current scoring of buyer user and the scoring requirement of each grade
Make the present level belonging to it.Based on this it is found that the present level of buyer user is according to each reference associated there
Node and determination, therefore the grading to user can be realized according to all related informations of user, to ensure that grade is drawn
That divides is comprehensive.At the same time, due in the relationship subgraph of buyer user, once it is associated with the buyer user any one
Reference mode changes, and just needs the current scoring for updating the buyer user, therefore can realize user rating easily and fast
Update, ensure that the real-time of grade classification.
Based on the above process it is found that disclosed technique scheme is to extract the relator of destination node in node relationships network
Figure, therefore the division methods of the node level also need to include the steps that constructing node relationships network.Fig. 3 is illustrated
The flow chart of the division methods of node level provided by the embodiment of the present disclosure.According to Fig.3, the division side of the node level
Method further includes following steps S100:
In the step s 100, node relationships network is constructed, and each node in node relationships network is initialized,
It include the relationship subgraph of destination node in the node relationships network.
It should be understood that the node relationships network constructed here, which refers to, contains the pass of all users and its related information
It is network, and the relationship subgraph of destination node is the small part in the relational network.
Based on this, the detailed process of the building node relationships network includes:
S1001, using default node as core, the associated nodes of default node, the default section are obtained based on preset relation
Point includes destination node；
S1002, node relationships network, the node relationships net are constructed according to the associated nodes for presetting node and default node
It include the relationship subgraph of destination node in network.
Wherein, default node refers to the node that its relationship subgraph is constructed as core, and associated nodes refer to and default section
The associated node of point, and preset relation then refers to the condition for establishing relationship between default node and associated nodes, packet
Include but be not limited to binding relationship, order relationship and session relationship.
For example, several pre-set users can be considered as core node by the present embodiment, which includes but is not limited to
Buyer user and seller user set up each pre-set user by various preset relations using each pre-set user as core
Relationship subgraph.In the case where pre-set user is enough, by preset relation can realize each pre-set user relationship subgraph it
Between association, to be built into required node relationships network.Based on this, the present embodiment can also be in the node relationships network
Each node is initialized, and as different nodes assigns different initialization values, such as the initialization benchmark of ordinary user
Score is 100, and the initialization benchmark score of bank card is 0, if encountering punishment subtracts 1 point, the initialization benchmark score of phonecard is
0, if encountering virtual-number subtracts 1 point, it can score in this way for subsequent node and basis be provided.
In view of the building of node relationships network is to observe user behavior for convenience, so as to user's implement general plan control
System, therefore disclosed technique scheme can also further monitor the process of node relationships network.Fig. 4 illustrates disclosure reality
Apply the flow chart of the division methods of node level provided by example.According to Fig.4, the division methods of the node level are further
Further include following steps S104 to S105:
In step S104, the quantity of the reference mode increased newly in the relationship subgraph of destination node in preset period of time is monitored
And the attribute of newly-increased reference mode；
In step s105, according to the preset period of time, the quantity of newly-increased reference mode and newly-increased reference mode
Attribute determines the attribute of destination node, the risk class of the attribute instruction destination node of the destination node.
For example, in e-commerce system, buyer user can be by between order relational implementation and seller user
Association, by chat relationship realize and being associated between good friend user, and by binding relationship realize and bank card and electricity
Association between words, therefore the present embodiment can monitor the growth of the relationship subgraph based on the relationship subgraph of destination node
Variation.It, will in the relationship subgraph of destination node if establishing order relationship between buyer user and new seller user
Increase new seller user using the reference mode as the buyer user；If being established between buyer user and new good friend user
Chat relationship then also will increase new good friend user using the reference node as the buyer user in the relationship subgraph of destination node
Point；If buyer user is bundled with new bank card, new bank card will increase also in the relationship subgraph of destination node to make
For the reference mode of the buyer user.Wherein, newly-increased reference mode carries respective attribute, for example, user belong to it is high-quality
User or junk user, bank card belong to platinum card or generic card, telephone number belong to advanced level user or defaulting subscriber
Deng.Based on this, the present embodiment can be according to the duration of preset period of time, the quantity of newly-increased reference mode and newly-increased reference mode
Attribute further determines that the attribute of destination node, the attribute of the destination node may be used to indicate the risk class of destination node.
Such as a large amount of good friend user is associated in the chat relationship of a buyer user, and when session content belongs to rubbish message,
The attribute of the buyer user will be judged as junk user, can further control at this time buyer user's implement general plan.
Based on above-mentioned every embodiment it is found that disclosed technique scheme can be realized based on the correlation behavior of user to
The grade classification at family, due to the relationship subgraph that classification basis is destination node, iteration speed is fast, is based on existing data
User rating can be updated easily and fast, so that it is guaranteed that the real-time of grade classification and comprehensive.
Following is embodiment of the present disclosure, can be used for executing embodiments of the present disclosure.
Fig. 5 a is the structural schematic diagram of the dividing device of the node level shown accoding to exemplary embodiment, which can lead to
Cross being implemented in combination with as some or all of of electronic equipment of software, hardware or both.According to Fig. 5 a, the node etc.
The dividing device of grade includes obtaining module 501, the first determining module 502 and division module 503.Wherein, module 501 is obtained to be used for
Obtain the relationship subgraph of destination node currently updated, include in the relationship subgraph of the destination node destination node and
Reference mode associated with the destination node, and the reference mode is updated with data；First determining module 502 is used for
According to the destination node, the reference mode and the ginseng associated with the destination node and with data update
The degree of association between node and the destination node is examined, determines the current scoring of the destination node；Division module 503 is used for root
The present level of the destination node is divided according to the current scoring of the destination node, the present level corresponds to current update
The destination node relationship subgraph.
In one embodiment, with reference to shown in Fig. 5 b, the acquisition module 501 includes first acquisition unit 5011 and second
Acquiring unit 5012.Wherein, first acquisition unit 5011 is used to obtain the relationship subgraph of the destination node before updating, and supervises
The data variation of each reference mode in the relationship subgraph of the destination node before updating is surveyed, the data variation includes described
The variation of the increase and decrease of reference mode and node scoring；Second acquisition unit 5012 is used for the shape in any one reference mode
When state changes, the relationship subgraph of the destination node after obtaining variation, the pass of the destination node after the variation
It is in subgraph, the destination node is node to be updated.
In one embodiment, with reference to shown in Fig. 5 c, first determining module 502 includes 5021 He of third acquiring unit
Determination unit 5022.Wherein, the history that third acquiring unit 5021 is used to obtain the destination node scores and the target section
The scoring variation of the associated reference mode of point and the degree of association between the reference mode and the destination node
Weighted value；Determination unit 5022 is used on the basis of the scoring of the history of the destination node, according to the scoring of the reference mode
The weighted value of variation and the degree of association between the reference mode and the destination node, determines the current of the destination node
Scoring.
In one embodiment, the degree of association between the reference mode and the destination node include first degree of association,
Second degree of association and the third degree of association；
The weighted value of the degree of association includes corresponding first weighted value of first degree of association, second degree of association pair
The second weighted value and the corresponding third weighted value of the third degree of association answered；
Wherein, first degree of association, second degree of association and the third degree of association are by by force to weak, first power
Weight values, second weighted value and the third weighted value are descending.
In one embodiment, with reference to shown in Fig. 5 d, the dividing device of the node level further includes building module 500, should
Building module 500 initializes each node in the node relationships network, institute for constructing node relationships network
State the relationship subgraph in node relationships network including the destination node.
In one embodiment, with reference to shown in Fig. 5 e, the building module 500 includes the 4th acquiring unit 5001 and building
Unit 5002.Wherein, the 4th acquiring unit 5001 is used to obtain the default section based on preset relation using default node as core
The associated nodes of point, the default node includes the destination node；Construction unit 5002 be used for according to the default node and
The associated nodes of the default node construct the node relationships network, include the destination node in the node relationships network
Relationship subgraph.
In one embodiment, with reference to shown in Fig. 5 f, the dividing device of the node level further includes monitoring modular 504 and
Two determining modules 505.Wherein, monitoring modular 504 is used to monitor in the relationship subgraph of the destination node in preset period of time and increase newly
Reference mode quantity and the newly-increased reference mode attribute；Second determining module 505 is used for according to described default
The attribute of period, the quantity of the newly-increased reference mode and the newly-increased reference mode determine the category of the destination node
Property, the attribute of the destination node indicates the risk class of the destination node.
The dividing device of node level provided by the embodiment of the present disclosure, according to each ginseng in the relationship subgraph of destination node
The data variation for examining node determines the current scoring of destination node in real time, and determines target according to the current scoring of destination node
The newest classification of node.It, can since the grade of destination node is determined in real time based on its all related information
Grade classification is realized easily and fast, to ensure that the real-time of grade classification and comprehensive.Compared to generalling use at present
Stage division, on the one hand disclosed technique scheme can be avoided the one-sidedness and limitation of preset rules, is on the other hand not necessarily to
The training set of preparatory mark, and data iteration can be realized easily and fast, therefore there is better grading effect.
About the device in above-described embodiment, modules execute the concrete mode of operation in related this method
It is described in detail in embodiment, no detailed explanation will be given here.
The embodiment of the present disclosure also provides a kind of dividing device of node level, which includes:
Processor；
Memory for storage processor executable instruction；
Wherein, processor is configured as executing:
The relationship subgraph of the destination node currently updated is obtained, includes the target in the relationship subgraph of the destination node
Node and reference mode associated with the destination node, and the reference mode is updated with data；
According to the destination node, it is associated with the destination node and with data update the reference mode, with
And the degree of association between the reference mode and the destination node, determine the current scoring of the destination node；
The present level of the destination node is divided according to the current scoring of the destination node, the present level is corresponding
In the relationship subgraph of the destination node currently updated.
In one embodiment, above-mentioned processor is also configured to: obtaining the relationship of the destination node before updating
Subgraph, and in the relationship subgraph of the destination node before Monitoring and Update each reference mode data variation, the data become
Change the variation of increase and decrease and node scoring including the reference mode；
When the state of any one reference mode changes, the relator of the destination node after obtaining variation
Scheme, in the relationship subgraph of the destination node after the variation, the destination node is node to be updated.
In one embodiment, above-mentioned processor is also configured to: obtaining history scoring and the institute of the destination node
It states between the scoring variation and the reference mode and the destination node of the associated reference mode of destination node
The weighted value of the degree of association；
On the basis of the scoring of the history of the destination node, according to the scoring of reference mode variation and the reference
The weighted value of the degree of association between node and the destination node determines the current scoring of the destination node.
In one embodiment, above-mentioned processor is also configured to: building node relationships network, and is closed to the node
It is that each node in network is initialized, includes the relationship subgraph of the destination node in the node relationships network.
In one embodiment, above-mentioned processor is also configured to: using default node as core, being obtained based on preset relation
The associated nodes of the default node are taken, the default node includes the destination node；
The node relationships network, the node are constructed according to the associated nodes of the default node and the default node
It include the relationship subgraph of the destination node in relational network.
In one embodiment, above-mentioned processor is also configured to: the pass of the destination node in monitoring preset period of time
It is the quantity of the reference mode increased newly in subgraph and the attribute of the newly-increased reference mode；
Attribute according to the preset period of time, the quantity of the newly-increased reference mode and the newly-increased reference mode is true
The attribute of the fixed destination node, the attribute of the destination node indicate the risk class of the destination node.
Fig. 6 is the structural block diagram of the dividing device for node level shown accoding to exemplary embodiment.For example, device
60 may be provided as a server.Device 60 includes processing component 602, further comprises one or more processors, and
The memory resource as representated by memory 604, can be by the instruction of the execution of processing component 602, such as using journey for storing
Sequence.The application program stored in memory 604 may include it is one or more each correspond to one group of instruction mould
Block.In addition, processing component 602 is configured as executing instruction, to execute the above method.
Device 60 can also include the power management that a power supply module 606 is configured as executive device 60, and one wired
Or radio network interface 608 is configured as device 60 being connected to network and input/output (I/O) interface 610.Device
60 can operate based on the operating system for being stored in memory 604, such as Windows ServerTM, Mac OS XTM,
UnixTM, LinuxTM, FreeBSDTM or similar.
The embodiment of the present disclosure also provides a kind of non-transitorycomputer readable storage medium, when the instruction in storage medium by
When the processor of device 60 executes, so that device 60 is able to carry out the division methods of above-mentioned node level, which comprises
The relationship subgraph of the destination node currently updated is obtained, includes the target in the relationship subgraph of the destination node
Node and reference mode associated with the destination node, and the reference mode is updated with data；
According to the destination node, it is associated with the destination node and with data update the reference mode, with
And the degree of association between the reference mode and the destination node, determine the current scoring of the destination node；
The present level of the destination node is divided according to the current scoring of the destination node, the present level is corresponding
In the relationship subgraph of the destination node currently updated.
In one embodiment, the relationship subgraph for obtaining the destination node currently updated includes:
Obtain the relationship subgraph of the destination node before updating, and the relator of the destination node before Monitoring and Update
The data variation of each reference mode in figure, the data variation include the change of increase and decrease and the node scoring of the reference mode
Change；
When the state of any one reference mode changes, the relator of the destination node after obtaining variation
Scheme, in the relationship subgraph of the destination node after the variation, the destination node is node to be updated.
In one embodiment, it is described according to the destination node, it is associated with the destination node and have data more
The degree of association between the new reference mode and the reference mode and the destination node, determines the destination node
It is current scoring include:
History scoring, the scoring of the reference mode associated with the destination node for obtaining the destination node become
The weighted value of change and the degree of association between the reference mode and the destination node；
On the basis of the scoring of the history of the destination node, according to the scoring of reference mode variation and the reference
The weighted value of the degree of association between node and the destination node determines the current scoring of the destination node.
In one embodiment, the degree of association between the reference mode and the destination node include first degree of association,
Second degree of association and the third degree of association；
The weighted value of the degree of association includes corresponding first weighted value of first degree of association, second degree of association pair
The second weighted value and the corresponding third weighted value of the third degree of association answered；
Wherein, first degree of association, second degree of association and the third degree of association are by by force to weak, first power
Weight values, second weighted value and the third weighted value are descending.
In one embodiment, the method also includes:
Node relationships network is constructed, and each node in the node relationships network is initialized, the node
It include the relationship subgraph of the destination node in relational network.
In one embodiment, the building node relationships network includes:
Using default node as core, the associated nodes of the default node, the default node are obtained based on preset relation
Including the destination node；
The node relationships network, the node are constructed according to the associated nodes of the default node and the default node
It include the relationship subgraph of the destination node in relational network.
In one embodiment, the method also includes:
Monitor the quantity of the reference mode increased newly in the relationship subgraph of the destination node in preset period of time and described new
The attribute of the reference mode of increasing；
Attribute according to the preset period of time, the quantity of the newly-increased reference mode and the newly-increased reference mode is true
The attribute of the fixed destination node, the attribute of the destination node indicate the risk class of the destination node.
Those skilled in the art will readily occur to its of the disclosure after considering specification and practicing disclosure disclosed herein
Its embodiment.This application is intended to cover any variations, uses, or adaptations of the disclosure, these modifications, purposes or
Person's adaptive change follows the general principles of this disclosure and including the undocumented common knowledge in the art of the disclosure
Or conventional techniques.The description and examples are only to be considered as illustrative, and the true scope and spirit of the disclosure are by appended
Claim is pointed out.
It should be understood that the present disclosure is not limited to the precise structures that have been described above and shown in the drawings, and
And various modifications and changes may be made without departing from the scope thereof.The scope of the present disclosure should be limited by the accompanying claims.
Claims (16)
1. a kind of division methods of node level characterized by comprising
The relationship subgraph of the destination node currently updated is obtained, includes the destination node in the relationship subgraph of the destination node
And reference mode associated with the destination node, and the reference mode is updated with data；
According to the destination node, the reference mode, Yi Jisuo associated with the destination node and with data update
The degree of association between reference mode and the destination node is stated, determines the current scoring of the destination node；
The present level of the destination node is divided according to the current scoring of the destination node, the present level corresponds to and works as
The relationship subgraph of the destination node of preceding update.
2. division methods according to claim 1, which is characterized in that the relationship for obtaining the destination node currently updated
Subgraph includes:
The relationship subgraph of the destination node before updating is obtained, and in the relationship subgraph of the destination node before Monitoring and Update
The data variation of each reference mode, the data variation include the variation of increase and decrease and the node scoring of the reference mode；
When the state of any one reference mode changes, the relationship subgraph of the destination node after obtaining variation,
In the relationship subgraph of the destination node after the variation, the destination node is node to be updated.
3. division methods according to claim 2, which is characterized in that described according to the destination node and the target
Pass between the node reference mode and the reference mode and the destination node associated and with data update
Connection degree determines that the current scoring of the destination node includes:
Obtain the destination node history scoring, the reference mode associated with the destination node scoring change,
And the weighted value of the degree of association between the reference mode and the destination node；
On the basis of the scoring of the history of the destination node, according to the scoring of reference mode variation and the reference mode
The weighted value of the degree of association between the destination node determines the current scoring of the destination node.
4. division methods according to claim 3, which is characterized in that between the reference mode and the destination node
The degree of association includes first degree of association, second degree of association and the third degree of association；
The weighted value of the degree of association includes that corresponding first weighted value of first degree of association, second degree of association are corresponding
Second weighted value and the corresponding third weighted value of the third degree of association；
Wherein, first degree of association, second degree of association and the third degree of association are by by force to weak, first weight
Value, second weighted value and the third weighted value are descending.
5. division methods according to claim 1, which is characterized in that further include:
Node relationships network is constructed, and each node in the node relationships network is initialized, the node relationships
It include the relationship subgraph of the destination node in network.
6. division methods according to claim 5, which is characterized in that the building node relationships network includes:
Using default node as core, the associated nodes of the default node are obtained based on preset relation, the default node includes
The destination node；
The node relationships network, the node relationships are constructed according to the associated nodes of the default node and the default node
It include the relationship subgraph of the destination node in network.
7. division methods according to claim 1-6, which is characterized in that further include:
Monitor the quantity of reference mode increased newly in the relationship subgraph of the destination node in preset period of time and described newly-increased
The attribute of reference mode；
Institute is determined according to the attribute of the preset period of time, the quantity of the newly-increased reference mode and the newly-increased reference mode
The attribute of destination node is stated, the attribute of the destination node indicates the risk class of the destination node.
8. a kind of dividing device of node level characterized by comprising
Module is obtained to wrap in the relationship subgraph of the destination node for obtaining the relationship subgraph of the destination node currently updated
The destination node and reference mode associated with the destination node are included, and the reference mode is updated with data；
First determining module, for according to the destination node, institute associated with the destination node and with data update
The degree of association between reference mode and the reference mode and the destination node is stated, determines the current of the destination node
Scoring；
Division module divides the present level of the destination node for the current scoring according to the destination node, described to work as
Preceding grade corresponds to the relationship subgraph of the destination node currently updated.
9. dividing device according to claim 8, which is characterized in that the acquisition module includes:
First acquisition unit, for obtaining the relationship subgraph of the destination node before updating, and the mesh before Monitoring and Update
Mark the data variation of each reference mode in the relationship subgraph of node, the data variation include the increase and decrease of the reference mode with
And the variation of node scoring；
Second acquisition unit, for the mesh when the state of any one reference mode changes, after obtaining variation
The relationship subgraph for marking node, in the relationship subgraph of the destination node after the variation, the destination node is section to be updated
Point.
10. dividing device according to claim 9, which is characterized in that first determining module includes:
Third acquiring unit, for obtain the destination node history score, the ginseng associated with the destination node
Examine scoring variation and the weighted value of the degree of association between the reference mode and the destination node of node；
Determination unit, for by the history of the destination node scoring on the basis of, according to the scoring of the reference mode change with
And the weighted value of the degree of association between the reference mode and the destination node, determine the current scoring of the destination node.
11. dividing device according to claim 10, which is characterized in that between the reference mode and the destination node
The degree of association include first degree of association, second degree of association and the third degree of association；
The weighted value of the degree of association includes that corresponding first weighted value of first degree of association, second degree of association are corresponding
Second weighted value and the corresponding third weighted value of the third degree of association；
Wherein, first degree of association, second degree of association and the third degree of association are by by force to weak, first weight
Value, second weighted value and the third weighted value are descending.
12. dividing device according to claim 8, which is characterized in that further include:
Module is constructed, is initialized for constructing node relationships network, and to each node in the node relationships network,
It include the relationship subgraph of the destination node in the node relationships network.
13. dividing device according to claim 12, which is characterized in that the building module includes:
4th acquiring unit, for obtaining the associated nodes of the default node based on preset relation using default node as core,
The default node includes the destination node；
Construction unit, for constructing the node relationships net according to the associated nodes of the default node and the default node
Network includes the relationship subgraph of the destination node in the node relationships network.
14. according to the described in any item dividing devices of claim 8-13, which is characterized in that further include:
Monitoring modular, the quantity of the reference mode increased newly in the relationship subgraph for monitoring the destination node in preset period of time with
And the attribute of the newly-increased reference mode；
Second determining module, for the quantity and the newly-increased ginseng according to the preset period of time, the newly-increased reference mode
The attribute for examining node determines that the attribute of the destination node, the attribute of the destination node indicate the risk etc. of the destination node
Grade.
15. a kind of dividing device of node level characterized by comprising
Processor；
Memory for storage processor executable instruction；
Wherein, the processor is configured to the step of perform claim requires any one of 1-7 the method.
16. a kind of computer readable storage medium, is stored thereon with computer instruction, which is characterized in that the instruction is by processor
The step of any one of claim 1-7 the method is realized when execution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910408213.0A CN110197397A (en) | 2019-05-15 | 2019-05-15 | The partitioning method and device of node level |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910408213.0A CN110197397A (en) | 2019-05-15 | 2019-05-15 | The partitioning method and device of node level |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110197397A true CN110197397A (en) | 2019-09-03 |
Family
ID=67751542
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910408213.0A Pending CN110197397A (en) | 2019-05-15 | 2019-05-15 | The partitioning method and device of node level |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110197397A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111914180A (en) * | 2020-08-19 | 2020-11-10 | 腾讯科技（深圳）有限公司 | User characteristic determination method, device, equipment and medium based on graph structure |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107257312A (en) * | 2017-05-10 | 2017-10-17 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device |
| CN107808283A (en) * | 2016-09-09 | 2018-03-16 | 腾讯科技（深圳）有限公司 | Order processing method, apparatus and system |
| CN107993144A (en) * | 2017-11-30 | 2018-05-04 | 平安科技（深圳）有限公司 | Customer risk grade determines method, apparatus, equipment and readable storage medium storing program for executing |
| CN108564467A (en) * | 2018-05-09 | 2018-09-21 | 平安普惠企业管理有限公司 | A kind of determination method and apparatus of consumer's risk grade |
| CN109472615A (en) * | 2017-09-06 | 2019-03-15 | 菜鸟智能物流控股有限公司 | Service processing method, device and equipment and readable medium |
-
2019
- 2019-05-15 CN CN201910408213.0A patent/CN110197397A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107808283A (en) * | 2016-09-09 | 2018-03-16 | 腾讯科技（深圳）有限公司 | Order processing method, apparatus and system |
| CN107257312A (en) * | 2017-05-10 | 2017-10-17 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device |
| CN109472615A (en) * | 2017-09-06 | 2019-03-15 | 菜鸟智能物流控股有限公司 | Service processing method, device and equipment and readable medium |
| CN107993144A (en) * | 2017-11-30 | 2018-05-04 | 平安科技（深圳）有限公司 | Customer risk grade determines method, apparatus, equipment and readable storage medium storing program for executing |
| CN108564467A (en) * | 2018-05-09 | 2018-09-21 | 平安普惠企业管理有限公司 | A kind of determination method and apparatus of consumer's risk grade |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111914180A (en) * | 2020-08-19 | 2020-11-10 | 腾讯科技（深圳）有限公司 | User characteristic determination method, device, equipment and medium based on graph structure |
| CN111914180B (en) * | 2020-08-19 | 2024-04-16 | 腾讯科技（深圳）有限公司 | User characteristic determining method, device, equipment and medium based on graph structure |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107330706A (en) | A kind of electricity battalion's customer service system and commercial operation pattern based on artificial intelligence | |
| CN107194623A (en) | A kind of discovery method and device of clique's fraud | |
| CN107330715A (en) | The method and apparatus for selecting display advertising material | |
| EP3852014A1 (en) | Method and apparatus for training learning model, and computing device | |
| Yeshwanth et al. | Evolutionary churn prediction in mobile networks using hybrid learning | |
| CN109743286A (en) | A kind of IP type mark method and apparatus based on figure convolutional neural networks | |
| CN109767052A (en) | Autotask distribution method and system | |
| CN110445939A (en) | The prediction technique and device of capacity resource | |
| CN117350640A (en) | Project progress management method and system | |
| JP2018005666A (en) | Human resources development assistance system | |
| CN110197397A (en) | The partitioning method and device of node level | |
| CN110413722A (en) | Address choice method, apparatus and non-transient storage medium | |
| CN107909496B (en) | User influence analysis method and device in social network and electronic equipment | |
| Kabicher-Fuchs et al. | Experience breeding in process-aware information systems | |
| CN107493315A (en) | A kind of collection method of behavioral data, Resource Server and storage medium | |
| Talpová et al. | Scrum anti-patterns, team performance and responsibility | |
| CN106874741B (en) | A kind of business same device identification method, calculates equipment and storage medium at device | |
| CN115293275A (en) | Data identification method and device, electronic equipment and storage medium | |
| CN112712002B (en) | CGAN-based environment monitoring method, CGAN-based environment monitoring device, CGAN-based environment monitoring system and storage medium | |
| CN115545088A (en) | Model construction method, classification method and device and electronic equipment | |
| do Amaral et al. | Attribute selection with filter and wrapper: an application on incident management process | |
| Christoforou et al. | A multilayer fuzzy cognitive maps approach to the cloud adoption decision support problem | |
| Yannou et al. | The UIPC-Monitor Tool for Augmenting Idea Maturity with Proofs of Utility Innovation Profitablilty and Concept | |
| JP6199480B2 (en) | Simulation analysis method and information processing system | |
| CN111178489B (en) | Conversation robot engine flow distribution method and device |
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: 20190903 |
|
| RJ01 | Rejection of invention patent application after publication |