CN108629624A - A kind of implementation method and its marketing message method for pushing of the neural network label of community portrait - Google Patents
A kind of implementation method and its marketing message method for pushing of the neural network label of community portrait Download PDFInfo
- Publication number
- CN108629624A CN108629624A CN201810424781.5A CN201810424781A CN108629624A CN 108629624 A CN108629624 A CN 108629624A CN 201810424781 A CN201810424781 A CN 201810424781A CN 108629624 A CN108629624 A CN 108629624A
- Authority
- CN
- China
- Prior art keywords
- label
- marketing
- community
- portrait
- marketing message
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 20
- 230000001960 triggered effect Effects 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 abstract description 5
- 230000000694 effects Effects 0.000 description 8
- 230000006399 behavior Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 241001122767 Theaceae Species 0.000 description 2
- 235000011850 desserts Nutrition 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 239000008267 milk Substances 0.000 description 2
- 210000004080 milk Anatomy 0.000 description 2
- 235000013336 milk Nutrition 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000012163 sequencing technique Methods 0.000 description 2
- 241000238876 Acari Species 0.000 description 1
- 238000012098 association analyses Methods 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 235000021152 breakfast Nutrition 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Marketing (AREA)
- Finance (AREA)
- General Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Entrepreneurship & Innovation (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Human Resources & Organizations (AREA)
- Primary Health Care (AREA)
- Tourism & Hospitality (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses the implementation methods and its marketing message method for pushing of a kind of neural network label of community portrait.The implementation method includes:Structure portrait label corresponding with marketing message;Marketing message is pushed to the community;Capture the exchange element of community user;Analyze the degree of association of exchange element portrait label corresponding with marketing message;After the degree of association is reached the label of effective threshold value as the last significant notation node in same significant notation nodes records to label threaded file;By the degree of association to do not reach effective threshold value label be used as contrast, by by according to label remove and as the last invalid flag node in same invalid flag nodes records to label chained file after;It repeats the above steps, forms the portrait with the neural network label.Method provided by the invention forms a kind of continuous renewal and perfect portrait and Marketing Mechanism so that the portrait for being directed to community is more accurate, and marketing efficiency increases.
Description
Technical field
A kind of implementation method for the neural network label drawn a portrait the present invention relates to community marketing domain more particularly to community and
Its marketing message method for pushing.
Background technology
Community refers to user's shape on the internet with essentially identical value orientation or hobby
At a kind of network colony, these users link up, are got in touch with, discussed by all kinds of social platforms.By to similar
The user of feature clusters, and forms community so that become possible for the precision marketing of specific realm user, it means that society
The commercial value of group's operation is huge.However even there is above-mentioned original intention, during community is established, due to user's differentiation feelings
The presence of condition so that the operation and management of community becomes difficult, such as how to maintain the active amount for the user for meeting community feature,
How to the suitable marketing activity of validated user push, reality is become difficult to not knowing about for user characteristics due to manager
It applies.
User draws a portrait, and is the technical foundation of the differentiated services such as personalized recommendation, diversification marketing, the core of user's portrait
Work is exactly to label for user.By carrying out labeling work to user, user behavior can be divided from macroscopic perspective
Analysis and prediction, contribute to enterprise for the precision of the marketing behavior of specific user.
However, existing user's Portrait brand technology needs to analyze based on a large amount of user behavior, and single community due to
Family amount is relatively little, and user behavior tends not to directly occur in the community, if thus by direct to society user inside the group
Portrait marketing is carried out, the numerous labels for forming its user portrait are difficult to follow the depth restriction of community and present and more match with it
State, redundancy label is more and miscellaneous so that the inaccurate, lag according to the marketing of user portrait, effect is poor, is not suitable for society
Group's operation and marketing.
Invention content
It is an object of the invention to solve the problems, such as that above-mentioned user's portrait is not suitable for community marketing, and provide a kind of community
The implementation method and its marketing message method for pushing of the neural network label of portrait.
To achieve the above object, the first aspect of the present invention provides a kind of realization of the neural network label of community portrait
Method includes the following steps:
Step 1:Structure portrait label corresponding with marketing message;
Step 2:Marketing message is pushed to the community;
Step 3:Capture exchange element of the community user after step 2;
Step 4:Analyze the degree of association of exchange element portrait label corresponding with the marketing message;
Step 5:The degree of association in step 4 is reached into the label of effective threshold value as same significant notation nodes records to mark
After signing the last significant notation node in threaded file, and mark timestamp;
Step 6:By the degree of association in step 4 to do not reach effective threshold value label be used as contrast, by label chained file by than
The label shone removes and as the last invalid flag section in same invalid flag nodes records to label chained file
After point, and mark timestamp;If a significant notation node only according to removal, is deleted there are one label and also this effectively
Markup node and its timestamp;
Step 7:It repeats step 1 and arrives step 6;Form the portrait with the neural network label.
In a certain embodiment:The step 5 and the step 6 can exchange.
In a certain embodiment:The significant notation node further includes marketing message other than label information and timestamp
Or the abstract of marketing message.
In a certain embodiment:The invalid flag node further includes marketing message other than label information and timestamp
Or the abstract of marketing message.
In a certain embodiment:If the corresponding portrait label of marketing message has been built up, step 1 can be skipped.
In a certain embodiment:The exchange element of the community is monitored before step 1, is established exchange element and is believed with marketing
Cease associated trigger;When the trigger is triggered, step 2 is executed.
To achieve the above object, the second aspect of the present invention provides a kind of marketing message method for pushing of community, described
Method for pushing is according to above-mentioned implementation method, by several marketing messages according to matching strategy by its corresponding portrait label and above-mentioned reality
The label in significant notation node in existing method is matched;Pair that several marketing messages are obtained according to its matching after matching
There should be criterion memory node sequence to sort to form marketing information link, and identification code is generated for the marketing information link, be believed according to marketing
Breath chain pushes the marketing message on the chain to the community successively.
In a certain embodiment, the method for pushing further includes:Using the single marketing message of marketing information link as step 2
The step 2 substituted into above-mentioned implementation method executes;And record produce the degree of association reach effective threshold value label situation whether
Occur, marked on significant notation node corresponding to the marketing message if occurring the marketing information link identification code and
Timestamp;It repeats, the identification code of the marketing information link is made to be distributed on several significant notation nodes.
In a certain embodiment, the method for pushing further includes:Above-mentioned method for pushing is repeated, significant notation is made
The identification code of marketing information link described in more than one is recorded on node.
To achieve the above object, the third aspect of the present invention provides a kind of marketing message method for pushing of community, described
Method directly extracts the identification code of above-mentioned marketing information link, and the significant notation node with same identification code is pressed the identification
Timestamp arrangement when code is labeled is pushed according to the order of the arrangement and is reached with the label degree of association on each significant notation node
To the marketing message of effective threshold value.
The present invention is in the building process of portrait, by being carried out to the exchange element generated in community after push marketing message
Crawl, by the carry out the matching analysis of the exchange element and marketing message corresponding label that grab, to which verification constitutes portrait repeatedly
Each label for particular communities validity, it is with strong points, be suitble to community operation, and solve the prior art since community is used
Family is relatively fewer, and does not often have real consumption behavior, so that label does not have representativeness, inaccurate ask of drawing a portrait
Topic.In addition, the embodiment of the present invention establishes a kind of Marketing Mechanism based on the portrait for being building up to the portrait of certain phase, it should
Marketing Mechanism does not carry out corresponding marketing only in accordance with the portrait in existing stage, has preferable specific aim and certain foresight, battalion
Efficiency is sold to improve;Can also the nested building process for substituting into above-mentioned portrait, to the respective labels in existing stage carry out verification and
Update forms a kind of benign cycle and combination of marketing and portrait, and allows marketing management person from the label chain
Valuable marketing information link identification code and portrait label are extracted in file, further community marketing are carried out going deep into excavation, be made
Surely it is suitble to the marketing program in community Historical Evolution direction.
Specific implementation mode
Below will by specific embodiment, the invention will be further described, pass through the following contents, work side of the invention
Formula and its advantage more clearly from will be understood and know.
The embodiment of the present invention one provides a kind of implementation method of the neural network label of community portrait comprising following
Step:
Step 1:Structure portrait label corresponding with marketing message;
Step 2:Marketing message is pushed to the community;
Step 3:Capture exchange element of the community user after step 2;
Step 4:Analyze the degree of association of exchange element portrait label corresponding with the marketing message;
Step 5:The degree of association in step 4 is reached into the label of effective threshold value as same significant notation nodes records to mark
After signing the last significant notation node in threaded file, and mark timestamp;
Step 6:By the degree of association in step 4 to do not reach effective threshold value label be used as contrast, by label chained file by than
The label shone removes and as the last invalid flag section in same invalid flag nodes records to label chained file
After point, and mark timestamp;If a significant notation node only according to removal, is deleted there are one label and also this effectively
Markup node and its timestamp;
Step 7:It repeats step 1 and arrives step 6;Form the portrait with the neural network label.
In the present embodiment, the single marketing message described in step 1 can correspond to several possible relevant portrait labels, example
Such as the marketing message first about the price reduction of the C food in the areas A B desserts shop, corresponding label may include the areas A, B desserts
Shop, C food, and thereon, subordinate's label and peer content similar in label.
In step 2-4, by pushing above-described this marketing message to the community, monitors user in community and receiving
The exchange element generated after to the push, and analyze these exchange elements and several portrait labels of the corresponding marketing message
Correlation degree.The exchange element include mutually exchange between community user and word content, audio content, picture material, regard
Frequency content, linked contents etc., will be in audio content, image by the methods of speech-to-text, image recognition, web page code extraction
After appearance, video content, linked contents first switch to word content, then by word semantic analysis in word derived above
Hold and carry out the matching analysis with each label, to obtain the degree of association of the exchange element and the portrait label.So, no
Marketing result success or not of the secondary marketing activity in the community can only be fed back, moreover it is possible to several pictures of the corresponding marketing message
As label carries out the judge with the actual association degree of the marketing message, belong to having for the marketing message and the community to filter out
The label of effect.
The step 5-6 then carried out, as executes above-mentioned screening.Specifically, obtain community exchange element with it is described
After the degree of association of several portrait labels, using the wherein degree of association be more than the label of a preset effective threshold value as can more represent so that
The label for the power source property that the marketing message can constantly be transmitted in the community is established and by above-mentioned label record to label chain
In a certain significant notation node in file;And the label for being less than effective threshold value is then used as contrasting, that is, is used as one
According to masterplate, in label chained file into line search, if in label chained file there are a certain label and this as contrasting masterplate
Label is identical, then removes it from corresponding node, and an invalid flag node is established in label chained file, will be upper
The invalid flag node is recorded in the label for stating not up to effective threshold value together, with indicate the label in the community for
The marketing message, or even the promotion degree of marketing result be not high.Particularly, if having criterion where the removed label
Memory node only includes the removed label, then by the significant notation knot removal, to maintain significant notation node
Accurate and effective.
The significant notation node is also marked every except several labels that the above-mentioned degree of association reaches effective threshold value are had recorded
Timestamp when minor node is established so that several significant notation nodes having in the label chained file rely on each node not
Same timestamp ordering forms an effective label chain with time sequencing.Likewise, invalid flag node also has recorded not
Reach the label of effective threshold value, and the timestamp of node foundation is equally marked so that invalid flag node is also according to the time
It sorts on the label chained file.
It is worth noting that the execution sequencing of step 5 and step 6 described in the present embodiment, has institute for structure
The portrait for stating neural network label has no actual influence, thus, in a possible embodiment, the step 5 and the step 6 can
To exchange.
After executing the step 5-6 for the first time, a significant notation node and a nothing are obtained in label chained file
Effect property flag node has recorded the corresponding label filtration result of the secondary marketing message.And works as and executing a certain marketing letter next time
When the push of breath, above-mentioned steps 1-6 is performed again so that the label in the significant notation node being previously obtained can be again
Obtain the verification of community user.It is illustrated again with above-mentioned example, first such as second of trial has been pushed about the areas A D milk tea
The marketing message of the E money milk tea price reduction in shop, but the corresponding response of community user cannot be obtained, then not only the corresponding label of D, E is answered
When being considered as invalid, and the corresponding labels of A are not to be regarded as pushing really effective label for the information of the community, it should also
It is removed.Thus, with the deep evolution of community operation, step 1-6 is repeatedly carried out so that really meet community Inherent spirit
Label, and can the label of authentic representative community user be saved.
By the way that above-mentioned steps 1-6 is repeatedly carried out, the portrait with the neural network label is formd.Build the portrait
Basis, i.e. neural network label, be made of each label in effective label node in the label chained file, and these
Effective label be continuous feedback based on actual marketing message in community, continuous renewal and be saved, thus it is right
Good matching degree is all had in each independent community, and is directed toward accurately, update is timely.Make the embodiment of the present invention in this way
Portrait method has abandoned that correlation degree between traditional label is poor, each label disperses independent portrait mode, forms a kind of mark
There is the portrait mode being closely connected between label, the effect of the corresponding marketing activity formed based on this kind portrait is also carried
It is high.
In concrete structure, the carrier of the community is community social client used by a user, such as wechat, QQ are immediately
Chat tool, the supplier of above-mentioned client can provide the first api interface of the exchange element and to community pushed informations
The second api interface, enabling it is first that the exchange from social client crawl is obtained by first api interface
Element.Further, the marketing message, portrait label, label chained file are stored in server, and the server passes through
Second api interface is connect with social client, to push marketing message to it according to portrait.In addition, the degree of association
Analysis and calculating are then executed by processor.
Further, the significant notation node further includes marketing message or marketing other than label information and timestamp
The abstract of information;Likewise, the invalid flag node further includes marketing message or battalion other than label information and timestamp
Sell the abstract of information.In this way so that significant notation node, invalid flag node are connected with specific marketing message, in the later stage
Marketing push when, the label in effective significant notation node can be directly selected according to corresponding marketing message, it is invalid to avoid
Label in markup node.
In addition, if the corresponding portrait label of marketing message has been built up, step 1 can be skipped.Thus the present embodiment can
The hot information that is there is provided using internet big data supplier and label system replace the execution of step 1, convenient and have
Good suitability.
Further, the present embodiment also monitors the exchange element of the community before step 1, establishes exchange element and battalion
Sell the trigger of information association;When community user exchanges to a certain topic information, related marketing letter existing for trigger match
Breath, when successful match, trigger is triggered, and executes step 2, and the marketing message is pushed to community, in user for this
When the liveness highest of topic information, the corresponding marketing of the corresponding topic information, the screening of respective labels and user's portrait are completed
It is perfect, efficiency greatly improves.
Next, the embodiment of the present invention two also provides a kind of marketing message method for pushing of community, this method is intended to
The foundation of above-mentioned portrait and improve during provide how to community push marketing message guidance.
The marketing message method for pushing includes:Several marketing messages are pressed into its corresponding portrait label according to matching strategy
It is matched with the label in the significant notation node in above-mentioned implementation method;Several marketing messages are matched according to it after matching
The obtained criterion memory node sequence that is corresponding with sorts to form a marketing information link, and generates identification code for the marketing information link,
Push to the community marketing message on the chain successively according to marketing information link.
In the method for pushing of above-mentioned marketing message, by by several marketing messages and having in significant notation node
Criterion label are matched so that several marketing messages have first passed through primary screener, if a marketing message can not be any one
It is matched with wherein several effective labels in a significant notation node, then represents the portrait of the marketing message and the community
It can not obviously match, thus will be excluded and push outside the plan, just obtain the one of the corresponding significant notation node in this way
Marketing information link will include N on the marketing information linkiIt is a may be in the effective marketing message of the community.Also, as described above
, timestamp when significant notation vertex ticks has the node to establish, during the foundation of marketing information link, the timestamp
Put in order and be carved on marketing information link again so that carry out pushing away for corresponding marketing message according to the sequence of the marketing information link
It send, the sequence of significant notation node when can be established based on the portrait in existing stage, it is special to meet the user that community gradually develops
Community of seeking peace feature, thus, the push effect of these marketing messages is preferable.In addition, the marketing information link obtained has one to know
Other code, enabling the marketing information link is obtained by extracting the identification code, is pushed accordingly, convenient for managing, and institute
State NiA marketing message may be the stronger one group of marketing message of correlation with same or similar theme, such as breakfast theme
Or the marketing message etc. of summer clothing theme, it can carry out concentrating the push of theme by managing the identification code, marketing efficiency is high.
Further, the present embodiment is in the single marketing message in pushing marketing information link obtained above, also by it
As the nested execution of step 2 in above-mentioned implementation method, and record whether produce the mark that the degree of association reaches effective threshold value
The situation of label marks the identification of the marketing information link if occurring on significant notation node corresponding to the marketing message
Code and timestamp.Above step is repeated to other marketing messages on marketing information link, makes the knowledge of the marketing information link
Other code is distributed to (Ni-Mi) on a significant notation node, wherein the MiIt is reached effectively for the above-mentioned degree of association that produces does not occur
The number of the single marketing message of the label of threshold value.By above method, there is N aboveiIt is a may be effective in the community
It is further effective to verify the possibility that the basis is drawn a portrait at this stage by practical push in the marketing information link of marketing message
Marketing message actual effectiveness, to obtain (Ni-Mi) a for the certain effective marketing message of the community.
Then, above-mentioned method for pushing is repeated, K marketing information link is obtained after repeating K times, and have
A effective marketing message really, thus so that there is the identification that record has more than one marketing information link on label chained file
The single significant notation node of code.
By above step, the mapping relations between K marketing information link and several significant notation nodes become multipair
More relationships.The identification code of single marketing information link is distributed to multiple significant notation nodes, will make more significant notations
Property node carries the marketing information link with same or similar theme marketing message, represents the corresponding master of the marketing information link
Topic maintains prolonged attention rate in the community, therefore should continue to excavate the marketing information link and further push;And
Single significant notation nodes records have the corresponding identification code of several marketing information links, then represent sometime nodes records
A certain label has corresponded to the marketing message of multiple themes, and such complex subject label then can comprehensively express community user
Feature.
To which the embodiment of the present invention three additionally provides a kind of marketing message method for pushing of community, and this method is based on upper
K marketing information link stating and unevenly record has several significant notation nodes of the K marketing information link.The party
Method includes:Significant notation node with same identification code is pressed the identification by the identification code for extracting above-mentioned K marketing information link
Timestamp arrangement when code is labeled is pushed according to the order of the arrangement and is reached with the label degree of association on each significant notation node
To the marketing message of effective threshold value.By the method for pushing, acquired marketing information link is corresponded into corresponding significant notation
On property node, and the marketing message that the degree of association on the significant notation node reaches effective threshold value is directly pushed, realized identical
The cluster of marketing message pushes, and matching degree height, the marketing effectiveness of marketing message and community are good.
Various embodiments of the present invention pass through the friendship to being generated in community after push marketing message in the building process of portrait
Stream element is captured, by the carry out the matching analysis of the exchange element and marketing message corresponding label that grab, to test repeatedly
Card constitutes validity of each label for particular communities of portrait, with strong points, is suitble to community operation, and solve the prior art
Due to community, user is relatively fewer, and does not often have real consumption behavior, so that label does not have representativeness, portrait is not
Accurate problem.In addition, the embodiment of the present invention establishes a kind of battalion based on the portrait for being building up to the portrait of certain phase
Pin mechanism, the Marketing Mechanism do not carry out corresponding marketing only in accordance with the portrait in existing stage, with preferable specific aim and centainly
Foresight, marketing efficiency improve;Can also the nested building process for substituting into above-mentioned portrait, to the respective labels in existing stage into
Row verification and update form a kind of benign cycle and combination of marketing and portrait, and allow marketing management person from institute
It states and extracts valuable marketing information link identification code and portrait label in label chained file, further community marketing is goed deep into
It excavates, formulates the marketing program for being suitble to community Historical Evolution direction.
The foregoing is merely the preferred embodiment of the present invention, it is not intended to limit its scope of the claims, it is every to utilize the present invention
Equivalent transformation made by description is directly or indirectly used in other related technical areas, is similarly included in this hair
In bright scope of patent protection.
Claims (10)
1. a kind of implementation method of the neural network label of community portrait, which is characterized in that include the following steps:
Step 1:Structure portrait label corresponding with marketing message;
Step 2:Marketing message is pushed to the community;
Step 3:Capture exchange element of the community user after step 2;
Step 4:Analyze the degree of association of exchange element portrait label corresponding with the marketing message;
Step 5:The degree of association in step 4 is reached into the label of effective threshold value as same significant notation nodes records to label chain
After connecing the last significant notation node in file, and mark timestamp;
Step 6:By the degree of association in step 4 to do not reach effective threshold value label be used as contrast, by label chained file by according to
Label remove and as the last invalid flag node in same invalid flag nodes records to label chained file after,
And mark timestamp;If a significant notation node only there are one label and by according to removal, also deletes this significant notation
Property node and its timestamp;
Step 7:It repeats step 1 and arrives step 6;Form the portrait with the neural network label.
2. a kind of implementation method of the neural network label of community portrait according to claim 1, which is characterized in that described
Step 5 and the step 6 can exchange.
3. a kind of implementation method of the neural network label of community portrait according to claim 1, which is characterized in that described
Significant notation node further includes the abstract of marketing message or marketing message other than label information and timestamp.
4. a kind of implementation method of the neural network label of community portrait according to claim 1, which is characterized in that described
Invalid flag node further includes the abstract of marketing message or marketing message other than label information and timestamp.
5. a kind of implementation method of the neural network label of community portrait according to claim 1, which is characterized in that if battalion
The corresponding portrait label of pin information has been built up, and can skip step 1.
6. a kind of implementation method of the neural network label of community portrait according to claim 1 or 5, which is characterized in that
The exchange element of the community is monitored before step 1, establishes exchange element and the associated trigger of marketing message;The triggering
When device is triggered, step 2 is executed.
7. a kind of marketing message method for pushing of community, which is characterized in that the method is according in claim 1 to 6 any one
Implementation method described in, by several marketing messages according to matching strategy by its corresponding portrait label and claim 1 to 6
The label in the significant notation node in implementation method described in any one is matched;Several marketing message roots after matching
The criterion memory node sequence that is corresponding with obtained according to its matching sorts to form marketing information link, and generates and know for the marketing information link
Other code pushes to the community marketing message on the chain according to marketing information link successively.
8. a kind of marketing message method for pushing of community according to claim 7, which is characterized in that further include following step
Suddenly:The step 2 that the single marketing message of marketing information link is substituted into as step 2 in claim 1 to 6 described in any one is held
Row;And record and produce the situation of the label that the degree of association reaches effective threshold value and whether occur, in the marketing message institute if occurring
The identification code and timestamp of the marketing information link are marked on corresponding significant notation node;It repeats, makes the marketing
The identification code of information chain is distributed on several significant notation nodes.
9. a kind of marketing message method for pushing of community according to claim 8, which is characterized in that further include following step
Suddenly:Method for pushing according to any one of claims 8 is repeated, makes to record marketing message described in more than one on significant notation node
The identification code of chain.
10. a kind of marketing message method for pushing of community, which is characterized in that the method is directly extracted described in claim 9
The identification code of marketing information link, the timestamp by the significant notation node with same identification code by the identification code when labeled
Arrangement, the marketing that effective threshold value is reached according to order push and the label degree of association on each significant notation node of the arrangement are believed
Breath.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810424781.5A CN108629624A (en) | 2018-05-07 | 2018-05-07 | A kind of implementation method and its marketing message method for pushing of the neural network label of community portrait |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810424781.5A CN108629624A (en) | 2018-05-07 | 2018-05-07 | A kind of implementation method and its marketing message method for pushing of the neural network label of community portrait |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108629624A true CN108629624A (en) | 2018-10-09 |
Family
ID=63695536
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810424781.5A Pending CN108629624A (en) | 2018-05-07 | 2018-05-07 | A kind of implementation method and its marketing message method for pushing of the neural network label of community portrait |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108629624A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111784397A (en) * | 2020-06-30 | 2020-10-16 | 广东奥园奥买家电子商务有限公司 | Community marketing method and device based on user portrait and computer equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120173550A1 (en) * | 2009-09-15 | 2012-07-05 | International Business Machines Corporation | System, method and computer program product for improving messages content using user's tagging feedback |
US20150289030A1 (en) * | 2009-12-04 | 2015-10-08 | At&T Intellectual Property I, Lp | Apparatus and method for tagging media content and managing marketing |
CN106844407A (en) * | 2015-12-07 | 2017-06-13 | 华为技术有限公司 | Label network production method and system based on data set correlation |
CN107205042A (en) * | 2017-06-30 | 2017-09-26 | 宇宙世代信息技术(深圳)有限公司 | A kind of accurate information method for pushing and system |
CN107886354A (en) * | 2017-10-31 | 2018-04-06 | 广州云移信息科技有限公司 | A kind of method and system for determining marketing target colony |
-
2018
- 2018-05-07 CN CN201810424781.5A patent/CN108629624A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120173550A1 (en) * | 2009-09-15 | 2012-07-05 | International Business Machines Corporation | System, method and computer program product for improving messages content using user's tagging feedback |
US20150289030A1 (en) * | 2009-12-04 | 2015-10-08 | At&T Intellectual Property I, Lp | Apparatus and method for tagging media content and managing marketing |
CN106844407A (en) * | 2015-12-07 | 2017-06-13 | 华为技术有限公司 | Label network production method and system based on data set correlation |
CN107205042A (en) * | 2017-06-30 | 2017-09-26 | 宇宙世代信息技术(深圳)有限公司 | A kind of accurate information method for pushing and system |
CN107886354A (en) * | 2017-10-31 | 2018-04-06 | 广州云移信息科技有限公司 | A kind of method and system for determining marketing target colony |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111784397A (en) * | 2020-06-30 | 2020-10-16 | 广东奥园奥买家电子商务有限公司 | Community marketing method and device based on user portrait and computer equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR101938790B1 (en) | Auto response server device, terminal device, reply system, reply method, and program | |
TWI689874B (en) | Method and device for neural network model training and transaction behavior risk identification | |
De Meo et al. | On Facebook, most ties are weak | |
Newman | The structure and function of complex networks | |
Choi et al. | Role of network structure and network effects in diffusion of innovations | |
De Nadai et al. | Are safer looking neighborhoods more lively? A multimodal investigation into urban life | |
CN110149361B (en) | Internet business operation process backtracking method, system and device | |
US8171055B2 (en) | System and method for generating communication subscriber description information | |
CN105279535B (en) | A kind of identification code intellectual analysis processing system and its processing method | |
CN114491615A (en) | Asynchronous longitudinal federal learning fair incentive mechanism method based on block chain | |
CN113010255A (en) | Interaction method and device based on binding session group and computer equipment | |
CN112862303A (en) | Crowdsourcing quality evaluation system and method based on block chain | |
CN110019542A (en) | The method, apparatus of the generation method of business connection, generation organizational member database and identification member of the same name in enterprise's map | |
CN108629624A (en) | A kind of implementation method and its marketing message method for pushing of the neural network label of community portrait | |
Neubert | Post Pandemic Internationalization Behavior | |
Sevin et al. | A new China: Media portrayal of Chinese mega-cities | |
CN106021441B (en) | A kind of mechanism friend recommendation method and system | |
CN113377656B (en) | Public testing recommendation method based on graph neural network | |
CN105718564A (en) | Promotion behavior detection method and apparatus | |
Guo | Research on information spreading model of social network | |
CN111080463B (en) | Key communication node identification method, device and medium | |
CN107798603A (en) | Transaction data processing method and device | |
CN110895604B (en) | Correlation fusion method of virtual identity information | |
CN113313615A (en) | Method and device for quantitatively grading and grading enterprise judicial risks | |
CN112416922A (en) | Group partner association data mining method, device, equipment and storage medium |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181009 |
|
WD01 | Invention patent application deemed withdrawn after publication |