CN106372959A - Internet-based user access behavior digital marketing system and method - Google Patents
Internet-based user access behavior digital marketing system and method Download PDFInfo
- Publication number
- CN106372959A CN106372959A CN201610700319.4A CN201610700319A CN106372959A CN 106372959 A CN106372959 A CN 106372959A CN 201610700319 A CN201610700319 A CN 201610700319A CN 106372959 A CN106372959 A CN 106372959A
- Authority
- CN
- China
- Prior art keywords
- user
- advertisement
- channel
- advertisement putting
- stream
- 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
- G06Q30/0241—Advertisements
- G06Q30/0242—Determining effectiveness of advertisements
- G06Q30/0245—Surveys
-
- 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
- G06Q30/0242—Determining effectiveness of advertisements
- G06Q30/0244—Optimization
-
- 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
- G06Q30/0253—During e-commerce, i.e. online transactions
-
- 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
- G06Q30/0277—Online advertisement
Abstract
The invention discloses an internet-based user access behavior digital marketing system and method. The digital marketing system comprises an advertisement cheating flow screening module, an advertisement user classification identification module and an advertisement release channel optimization module. According to the system and the method, real flow is identified by analyzing access behaviors of users, index values reflecting the user behaviors are classified and analyzed to establish a classification rule, and a classified user group is subjected to tag identification, so that really valuable customers are found and the misjudgment of the user behaviors in a redirected release process is avoided. Due to effective assessment of release channels, advertisements can be released to potential customers with high efficiency and low cost, and the problem that the release cannot be realized due to price fluctuation during real-time bidding of the redirected release process is avoided.
Description
Technical field
The present invention relates to user accesses big data technology, Internet advertising input and digital marketing field and in particular to one
Plant the user access activity digital marketing system based on the Internet and method.
Background technology
Current advertisement putting is mainly based upon what dsp (demand-side platform, i.e. party in request's platform) was carried out.
Dsp originates from the flourishing America and Europe of the web advertisement, is accompanied by the web advertisement developing rapidly new rise of the Internet and advertising
Field.It rises abruptly in the U.S. rapidly together with ad exchange and rtb, has covered within 2011 American-European, Asia-Pacific and Australia
Continent.In world's network display advertisement field, dsp is in the ascendant.The incoming China of dsp, becomes upsurge at home rapidly, becomes promotion
One of fast-developing power in Chinese network display advertisement rtb market, dsp will become sem after an advertisement mode.
Rtb (real time bidding) real time bid, is a kind of utilization third party technology in millions of websites
Or mobile terminal is directed to the technology of bidding that each user shows that behavior is estimated and bids.Throw in the frequency with making a big purchase in large quantities not
With real time bid has been evaded invalid audient and reached, and is bought for significant user.
Redirect advertisement (retargeting) and give advertisement, advertisement of marketing again, frequent customer's advertisement etc. for change also referred to as visitor.
, by using Internet technology, for website caller, before showing it, the advertisement of once browsed commodity or service, passes through for it
Twice, the repeated exposure of three times even n time and continuous prompting, realize final transaction.(programmatic is bought in sequencing
Direct buying) it is current digital advertisement marketing domain concept the most popular, redirect skill in particular with the Internet crowd
The development of art, by the mobile Internet behavior analysiss to user, thus giving for change to old user, buys by sequencing
The redirection advertisement (retargeting) that technology is realized is that Internet advertising throws in mould from traditional display advertisement to accurate advertisement
The transformation of formula.Redirecting advertisement (retargeting) is exactly briefly a kind of directional technology of web advertisement, that is, be directed to advertisement
Certain attribute of audient (audience), in same advertisement position, pushes advertisement for his customization to his page.Redirect wide
Announcement is according to certain behavior (action is often recorded in your cookie it is also possible to not be) before you, this row
By the specific advertisement being triggered, again it is pushed in face of you.
Although redirecting advertisement to be widely applied in electric business field, the E-commerce Marketing for enterprise is overall
For effect, can not fully meet them to digital marketing demand, the following is redirection advertisement technology and carrying out digital marketing
The insurmountable problem of institute in the middle of process:
(1) on the Internet, false flow and false click are intertwined with true demand volume, redirect advertisement skill
Art cannot certain behavior (action) of being formed every real traffic and false flow of effective district, the process now with advertisement putting works as
In be to truly being thrown in together with false flow it is therefore necessary to separate to authentic and valid flow.
(2) some real user's requests, are that the agent of channel under line helps client to carry out generation operation, by weight
Targeted ads technology we cannot find really valuable user, and it is carried out with the marketing of correlation.Because these agents
The mobile Internet behavior (action) of member can be redirected the user that advertisement technology is mistakenly considered authentic and valid.
(3) redirect advertisement technology and use the pattern of rtb (realtime bidding) real time bid and carry out advertisement
Push, in the middle of the process carrying out marketing advertisement pushing, user is carried out occurring in the middle of the process of real time bid and bid
Low user cannot bid successfully, or the problem of the high cost bidded.The user to some high values so can be led to no
Method carries out the problem of digital marketing.
In sum, in the existing dsp advertisement delivery system redirecting in advertisement based on mobile Internet behavioral techniquess
Advertisement putting can only be carried out in specific business, and advertising user is wished by the electric business user access activity to own client
Analysis, realizes advertisement putting under different business scene for the potential customers, but there also is not the solution of this respect at present.
Content of the invention
The invention provides a kind of user access activity digital marketing system based on the Internet and method, by electric business
The access behavior of platform user is effectively analyzed, and can filter out valuable potential user, and can carry out advertisement putting canal
The optimization in road.This digital marketing system can support cps or cpa clearing advertising expense.
The invention provides a kind of user access activity digital marketing system based on the Internet, comprising:
Advertisement cheating stream screening module, advertising user Classification and Identification module and advertisement putting channel preferred module, wherein,
Described advertisement cheating stream screening module is used for collecting the access information of user, and the behavior of output reflection user property refers to
Mark, and real traffic is screened out by setting target threshold value；
Described advertising user Classification and Identification module is used for the user's row by stream screening module output that described advertisement is practised fraud
For index, advertisement putting colony is classified, to determine advertisement putting object；
Described advertisement putting channel preferred module is used for the input channel of each advertisement putting object is estimated, and selects
The advertisement putting channel going out optimum is thrown in described advertisement putting object.
Preferably, described advertisement cheating stream screening module is received by adding one section of javascript on web or wap
Collection user sets up user behavior data storehouse in the access information of current page, and carries out data with the Subscriber Management System of advertiser
Docking, realizes associating of user access activity data and UAD.
Preferably, described advertisement cheating stream screening module is surveyed to each webpage using the behavior characteristicss index of user
Examination, in cheating stream from the different values of each characteristic index of real information stream, determines the threshold value screening cheating stream.
Preferably, described advertisement cheating stream screening module chooses certain sample from collected user access information
As training sample, remaining sample size, as test sample, then extracts user from the Subscriber Management System of advertiser to this amount
Attribute and the index of correlation accessing behavior, and all indexs are standardized process, using the characteristic equation of normalized matrix
Determine described characteristic index.
Preferably, described advertising user Classification and Identification module determines potential advertisement putting pair using the access information of user
The feature of elephant, and buy rule to find out the criterion of potential advertisement putting object by formulating, exist according to this criterion
In the test sample before randomly choosing, inspection user has played the data of order, when under the user in inspection data, forms data is being just
Really rate meets during certain value it is determined that being advertisement putting object.
Preferably, described advertisement putting channel preferred module described advertising user Classification and Identification module is determined each
The input channel of advertisement putting object is estimated, and rate of return on investment highest channel is thrown in object as respective advertisement
Throw in channel.
Preferably, the described input channel that described advertisement putting channel preferred module determines include note, edm, dsp,
Microblogging and call center, and the channel of correlation is automatically abandoned when user's channel attribute data lacks.
Present invention also offers a kind of user access activity digital marketing method based on the Internet, methods described include with
Lower step:
(1) practised fraud by advertisement and flow the access information that screening module collects user, the behavior of output reflection user property refers to
Mark, and real traffic is screened out by setting target threshold value；Described advertisement cheating stream screening module is passed through on web or wap
Add one section of javascript and set up user behavior data storehouse in the access information of current page collecting user, and and advertiser
Subscriber Management System carry out data docking, realize associating of user access activity data and UAD；Described advertisement
Cheating stream screening module is tested to each webpage using the behavior characteristicss index of user, in cheating stream and each spy of real information stream
Levy in the different values of index, determine the threshold value screening cheating stream；
(2) pass through the user behavior index to described advertisement cheating stream screening module output for the advertising user Classification and Identification module
Carry out the classification of advertisement putting colony, to determine advertisement putting object；Described advertising user Classification and Identification module utilizes user's
Access information determines the feature of potential advertisement putting object, and buys rule to find out potential advertisement putting object by formulating
Criterion, has checked under user the data of order according to this criterion in the test sample randomly choosing, has worked as check number
According in user under the accuracy of forms data meet during certain value it is determined that being advertisement putting object；
(3) by advertisement putting channel preferred module, the input channel of each advertisement putting object is estimated, and selects
Go out rate of return on investment highest input channel described advertisement putting object is thrown in.
Preferably, described advertisement cheating stream screening module chooses certain sample from collected user access information
As training sample, remaining sample size, as test sample, then extracts user from the Subscriber Management System of advertiser to this amount
Attribute and the index of correlation accessing behavior, and all indexs are standardized process, using the characteristic equation of normalized matrix
Determine described characteristic index.
Preferably, also including step (4): in each input channel setting unique monitoring chain determined by step (3)
Connect with picture the effective monitoring it is ensured that to data, after preferential channels advertisement putting certain time, user did not react, will be from
The dynamic next input channel proceeding to correlation.
The present invention provide a kind of user access activity digital marketing system based on the Internet and method, acquirement beneficial
Technique effect includes:
(1) present invention screens out real flow by analyzing the access behavior of user, and each to reflection user behavior
Desired value carries out classification analysises and sets up classifying ruless, carries out tag identifier to the user group sorting out, thus find really having
The client being worth, it is to avoid redirect the erroneous judgement to user behavior in launch process.The effective assessment throwing in channel to each can be high
Every potential customers are arrived in the advertisement putting that makes of effect low cost, it is to avoid during the real time bid of redirection launch process, because of price fluctuation
The problem that cannot realize throwing in.
(2) the system employs the launch process of first customer analysis-excavation-input-optimization, changes passing dsp and throws in
The process of the input-optimization of system, thus be substantially improved the efficiency of input.
(3) the system particularly throws in efficiency to the user of the electric business website of single product 50% lifting, solves to work as
Front each jettison system cannot service very well single product input efficiency, and can effectively be lifted old user repeat marketing effect
The problem of rate, the conversion ratio of potential user can rise to 10% about from traditional 5%.
Brief description
Fig. 1 is the schematic diagram of conventional dsp.
Fig. 2 is the digital marketing system structure chart based on the behavior of access of the present invention.
Fig. 3 is the digital marketing method flow diagram based on the behavior of access of the present invention.
Specific embodiment
For making technical scheme disclosed by the invention more specific, below in conjunction with the accompanying drawing in embodiment, to the present invention's
Technical scheme is clearly and completely described, and described embodiment is only the preferred embodiment for the present invention, rather than all
Embodiment.Based on the embodiment in the present invention, those skilled in the art are obtained under the premise of not making creative work
Every other embodiment, all should belong to the scope of protection of the invention.
To describe embodiments of the invention below in conjunction with the accompanying drawings in detail.Referring to Fig. 1-2, the invention discloses a kind of be based on mutually
On-line customer access behavior digital marketing system, include advertisement cheating flow screening module, advertising user Classification and Identification module and
Advertisement putting channel preferred module, above three module is concentrated by algorithm management module and is managed.
Distinguish effective discharge and false flow in order to be able to effective, need the access behavior of user is effectively analyzed,
Thus find out to separate out true and false user, excavate potential user and carry out digital marketing.Wherein, described advertisement cheating stream screening mould
Block is used for collecting the access information of user, the behavioral indicator of output reflection user property, and is screened by setting target threshold value
Go out real traffic.Preferably completed by one section of javascript is added on web or wap, this section of javascript can collect
User, in the access information of current page, is subsequently sent to get server ready.Sdk bag can be embedded in app, be received by sdk bag
Collection user is in the access information of current page.
Digital marketing system includes described access information is set up user behavior data storehouse, the index of correlation bag of user behavior
Include but be not limited to user ip address, city selected by visitor, user's access times, whether login user, each average access duration
The last access time of (minute), distance is spaced, whether bought product, places an order apart from last time buying interval, visitor
Number of times, backward chaining type, stand outer promote, land page, the user of the product page browses quantity, client age, Ke Huxing
... do not wait numerous indexs.These index reflections behavior property of client, can automatically periodically import for two hours for a cycle
To in the data base of digital marketing system.User behavior data storehouse carries out data with the Subscriber Management System of advertiser and docks, real
Existing user access activity data is associated with UAD.
Described advertisement cheating stream screening module can filter out from above-mentioned user behavior index and can represent client properties
" characteristic index ".Concrete grammar can the user behavior index that each two hour is collected be randomly selected, preferably
, as training sample, remaining 25% as test sample in sample 75%.Obviously, aforementioned proportion can be by people in the art
Member adjusts accordingly as needed.Visitor's attribute and visitor is extracted from the user behavior data storehouse of described digital marketing system
The index of correlation of behavior, is designated asWherein vijRepresent j-th index of i-th visitor, m, n are respectively
For visitor's number and index number, and all of index is standardized processWherein Then, to normalized matrix zijSeek correlation matrix, and seek the characteristic equation of this matrix, obtain p
Individual characteristic root, adds up, according to front c main constituent eigenvalue, the rule that variance contribution ratio is more than 85%, determines main constituent number c, this
A little main constituents are " characteristic index ".Then, by the user behavior characteristic index of above-mentioned determination, each webpage is tested, due to
The value of the characteristic index of cheating stream and real information stream is different, and those skilled in the art can be according to the difference selection of industry
Screen the concrete threshold value of cheating stream, in such as hotel industry, 100 users need to be not less than 5 orders, are otherwise just judged as
False flow.
Described advertising user Classification and Identification module is used for determining the object of advertisement putting, thus realizing precisely throwing in.Can make
With methods such as cluster, Multivariate logistic regression analysis, decision tree analysis, Bayes's classification, neutral nets, by arbitrary two
Kind or two or more algorithm combination find out the feature of potential user, and using 25% test sample, confirm potential user's
Verity.Such as, index number c of above-mentioned confirmation can be designated as xi={ x1,x2,…xn, and carry out cluster analyses.First basis
Sample point and center of mass point square distance and judging that k value is k species in distance function judgement group, then carry out k-means cluster
Find out the obvious user of purchase intention, its flow process is as follows: determine k value, each object represents the initial mean value of a cluster, to residue
Each object, according to its distance with each cluster average, it is assigned to most like cluster.Then calculate the newly equal of each cluster
Value, this process is repeated continuously, until criterion function convergence.Using square error criterionThis
In jcM () is the synthesis of the square error of all objects in data base, xiIt is the point in space, represent given data object, zj
It is cluster cjMeansigma methodss (xiAnd zjIt is all multidimensional)；Finally, select high conversion rate in population mean 5% from this k class user
Crowd, be defined as potential advertisement putting object.
Described advertising user Classification and Identification module can also use Multivariate logistic regression analysis, decision tree analysis, pattra leaves
The methods such as this classification, neutral net are regular to assist judgement to buy, and determine the criterion of potential advertisement putting object.According to
The criterion of the potential advertisement putting object of above-mentioned determination, has checked under user in the test sample randomly choosing before and has ordered
Single data, when accuracy preferably greater than 80% of forms data under the user in inspection data, then the choosing of advertisement putting object
Standard of selecting is set up, and otherwise re-starts data model and calculates it is clear that other values of this accuracy can be by those skilled in the art
It is determined as needed.Normal model discrimination index can be recorded by described digital marketing system, every two hours timing
Automatic sieve selects the data of correlation.
Described digital marketing system includes but is not limited to note, edm, dsp, microblogging, five kinds of advertisement puttings of call center
Channel, realizes the covering that user advertising is thrown in by above-mentioned channel.Described advertisement putting channel preferred module is to above-mentioned determination
The input channel of each advertisement putting object be estimated, and rate of return on investment highest channel is thrown in as respective advertisement
The input channel of object.Formula specifically can be utilizedMake yiThe channel j obtaining corresponding to maximum is then excellent
First throw in channel, wherein βjFor the weight of channel j, piFor the unit price of product i, cjCost for channel j.And the weight beta of channel jjIt is
Calculated using the roi that product respectively throws in channel,Wherein roiijThe roi, p of the channel j for product iiFor
The unit price of product i, covijFor product i channel j conversion ratio, then it follows that
Calculate the weight of each channel followed by Information Entropy, then screen optimal input channel mode.When user's channel
During attribute data disappearance, automatically abandon the channel of correlation.As phone number disappearance, automatically abandoned note channel.
Preferably, input strategy and monitoring selecting module can also be set, channel is thrown in each unique prison is set
Control links with picture the effective monitoring it is ensured that to data, because behavioral data has time inefficacy decreasing effect, when preferential channels are wide
After announcement two hour users of input do not react, system is automatically proceeded to the next of correlation and is thrown in channel.And specify preferably two
The input of this advertisement must be completed in it.
Present invention also offers a kind of digital marketing method based on the behavior of access, comprise the following steps:
Step 1: the access information of the stream screening module collection user that practised fraud by described advertisement, output reflection user property
Behavioral indicator, and real traffic is screened out by setting target threshold value.
Add one section of javascript to complete, this section of javascript can collect user and work as preferably on web or wap
The access information of the front page, is subsequently sent to get server ready.Sdk bag can be embedded in app, user be collected by sdk bag and exists
The access information of current page.Then described access information is set up with user behavior data storehouse, the index of correlation bag of user behavior
Include but be not limited to user ip address, city selected by visitor, user's access times, whether login user, each average access duration
The last access time of (minute), distance is spaced, whether bought product, places an order apart from last time buying interval, visitor
Number of times, backward chaining type, stand outer promote, land page, the user of the product page browses quantity, client age, Ke Huxing
... do not wait numerous indexs.These index reflections behavior property of client, can automatically periodically import for two hours for a cycle
To in the data base of digital marketing system.User behavior data storehouse carries out data with the Subscriber Management System of advertiser and docks, real
Existing user access activity data is associated with UAD.
Then, screen user behavior characteristic index.Concrete grammar can be the user behavior that each two hour is collected
Index is randomly selected, and, as training sample, remaining 25% as test sample for preferably 75% in sample.Obviously, above-mentioned
Ratio can be adjusted accordingly as needed by those skilled in the art.User behavior data from described digital marketing system
Extract the index of correlation of visitor's attribute and visitor's behavior in storehouse, be designated asWherein vijRepresent i-th visit
J-th index of visitor, n, m are respectively visitor's number and index number, and all of index are standardized process
Then, to normalized matrix zijSeek correlation matrix, and seek the characteristic equation of this matrix, obtain p characteristic root, according to front c
Main constituent eigenvalue adds up the rule that variance contribution ratio is more than 85%, determines main constituent number c, and i.e. " feature refers to for these main constituents
Mark ".
Finally, by the user behavior characteristic index of above-mentioned determination, each webpage is tested, due to cheating stream with truly
The value of the characteristic index of flow of information is different, and those skilled in the art can choose the concrete threshold value screening cheating stream.
Step 2: the user behavior of stream screening module output of described advertisement being practised fraud by advertising user Classification and Identification module
Index carries out the classification of advertisement putting colony, to determine advertisement putting object.Can be divided using cluster, multiple logistic regression
The methods such as analysis, decision tree analysis, Bayes's classification, neutral net, are looked for by two or more algorithm combination arbitrary
Go out the feature of potential user, and using 25% test sample, confirm the verity of potential user.Such as, can be by above-mentioned confirmation
Index number c, be designated as xi={ x1,x2,…xn, and carry out cluster analyses.First according to sample point in distance function judgement group with
Center of mass point square distance and judging that k value is k species, then carry out k-means cluster and find out the obvious user of purchase intention, its
Flow process is as follows: determines k value, each object represents the initial mean value of a cluster, to each object remaining, according to itself and each cluster
The distance of average, it is assigned to most like cluster.Then calculate the new average of each cluster, this process is repeated continuously, directly
To criterion function convergence.Using square error criterionHere jcM () is all right in data base
The synthesis of the square error of elephant, xiIt is the point in space, represent given data object, zjIt is cluster cjMeansigma methodss (xiAnd zjAll
It is multidimensional)；Finally, select high conversion rate in the crowd of population mean 5% from this k class user, be defined as potential advertisement
Throw in object.Described advertising user Classification and Identification module can also use Multivariate logistic regression analysis, decision tree analysis, pattra leaves
The methods such as this classification, neutral net are regular to assist judgement to buy, and determine the criterion of potential advertisement putting object.According to
The criterion of the potential advertisement putting object of above-mentioned determination, has checked under user in the test sample randomly choosing before and has ordered
Single data, when accuracy preferably greater than 80% of forms data under the user in inspection data, then the choosing of advertisement putting object
Standard of selecting is set up, and otherwise re-starts data model and calculates it is clear that other values of this accuracy can be by those skilled in the art
It is determined as needed.Normal model discrimination index can be recorded by described digital marketing system, every two hours timing
Automatic sieve selects the data of correlation.
Step 3: by advertisement putting channel preferred module, the input channel of each advertisement putting object is estimated, and
Select rate of return on investment highest input channel described advertisement putting object is thrown in.Described digital marketing system include but
It is not limited to note, edm, dsp, microblogging, five kinds of advertisement putting channels of call center, realized by above-mentioned channel wide to user
Accuse the covering thrown in.The input channel of each advertisement putting object to above-mentioned determination for the described advertisement putting channel preferred module enters
Row assessment, and rate of return on investment highest channel is thrown in the input channel of object as respective advertisement.Formula specifically can be utilizedMake yiObtaining the channel j corresponding to maximum is then preferentially to throw in channel, wherein βjFor the weight of channel j,
piFor the unit price of product i, cjCost for channel j.And the weight beta of channel jjIt is to be calculated using the roi that product respectively throws in channel
Go out,Wherein roiijThe roi, p of the channel j for product iiFor the unit price of product i, covijFor product i in canal
The conversion ratio of road j, then it follows that
Calculate the weight of each channel followed by Information Entropy, then screen optimal input channel mode.When user's channel
During attribute data disappearance, automatically abandon the channel of correlation.As phone number disappearance, automatically abandoned note channel.
Preferably, also include step (4) after step (3): throw in channel at each and unique monitoring link is set
With picture it is ensured that effective monitoring to data, because behavioral data has time inefficacy decreasing effect, when preferential channels advertisement putting
After two hour users do not react, system is automatically proceeded to the next of correlation and is thrown in channel.And specify must in preferably two days
The input of this advertisement must be completed.
The announcement of book and teaching according to the above description, those skilled in the art in the invention can also be to above-mentioned embodiment party
Formula is changed and is changed.Therefore, the invention is not limited in specific embodiment disclosed and described above, to invention
In the scope of the claims that a little modifications and changes should also be as fall into the present invention.Although additionally, employing in this specification
Some specific terms, but these terms are merely for convenience of description, do not constitute any restriction to invention.
Claims (10)
1. a kind of user access activity digital marketing system based on the Internet is it is characterised in that include: advertisement cheating stream screening
Module, advertising user Classification and Identification module and advertisement putting channel preferred module, wherein,
Described advertisement cheating stream screening module is used for collecting the access information of user, and output reflects the behavioral indicator of user property,
And real traffic is screened out by setting target threshold value；
Described advertising user Classification and Identification module is used for referring to by the user behavior of stream screening module output that described advertisement is practised fraud
Mark is classified to advertisement putting colony, to determine advertisement putting object；
Described advertisement putting channel preferred module is used for the input channel of each advertisement putting object is estimated, and selects
Excellent advertisement putting channel is thrown in described advertisement putting object.
2. digital marketing system according to claim 1 is it is characterised in that described advertisement cheating stream screening module is passed through
One section of javascript is added on web or wap and sets up user behavior data storehouse in the access information of current page collecting user,
And carry out data with the Subscriber Management System of advertiser and dock, realize the pass of user access activity data and UAD
Connection.
3. digital marketing system according to claim 1 is it is characterised in that described advertisement cheating stream screening module is using use
The behavior characteristicss index at family is tested to each webpage, cheating stream from the different values of each characteristic index of real information stream,
Determine the threshold value screening cheating stream.
4. digital marketing system according to claim 3 is it is characterised in that described advertisement cheating stream screening module is from being received
Choose certain sample size in the user access information collecting as training sample, remaining sample size as test sample, then
The index of correlation extracted user property from the Subscriber Management System of advertiser and access behavior, and standard is carried out to all indexs
Change is processed, and determines described characteristic index using the characteristic equation of normalized matrix.
5. digital marketing system according to claim 4 is it is characterised in that described advertising user Classification and Identification module utilizes
The access information of user determines the feature of potential advertisement putting object, and finds out potential advertisement putting by formulating purchase rule
The criterion of object, has checked under user the number of order according to this criterion in the test sample randomly choosing before
According to when the accuracy of forms data under the user in inspection data meets certain value it is determined that being advertisement putting object.
6. digital marketing system according to claim 1 is it is characterised in that described advertisement putting channel preferred module is to institute
The input channel of each advertisement putting object stating the determination of advertising user Classification and Identification module is estimated, and by rate of return on investment
Highest channel throws in the input channel of object as respective advertisement.
7. digital marketing system according to claim 1 is it is characterised in that described advertisement putting channel preferred module determines
Described input channel at least include note, edm, dsp, microblogging and call center, and when user's channel attribute data lack
Automatically the channel of correlation is abandoned during mistake.
8. a kind of user access activity digital marketing method based on the Internet is it is characterised in that methods described includes following step
Rapid:
(1) practised fraud by advertisement and flow the access information that screening module collects user, the behavioral indicator of output reflection user property, and
Real traffic is screened out by setting target threshold value；Described advertisement cheating stream screening module passes through to add one on web or wap
Section javascript sets up user behavior data storehouse, and the user with advertiser in the access information of current page collecting user
Management system carries out data docking, realizes associating of user access activity data and UAD；Described advertisement cheating stream
Screening module is tested to each webpage using the behavior characteristicss index of user, in cheating stream and each characteristic index of real information stream
Different values in, determine screen cheating stream threshold value；
(2) by advertising user Classification and Identification module, the user behavior index of described advertisement cheating stream screening module output is carried out
The classification of advertisement putting colony, to determine advertisement putting object；Described advertising user Classification and Identification module utilizes the access of user
Information determines the feature of potential advertisement putting object, and buys rule to find out the index of potential advertisement putting object by formulating
Standard, has checked under user the data of order according to this criterion, when in inspection data in the test sample randomly choosing
User under the accuracy of forms data meet during certain value it is determined that being advertisement putting object；
(3) by advertisement putting channel preferred module, the input channel of each advertisement putting object is estimated, and selects throwing
Money return rate highest is thrown in channel and described advertisement putting object is thrown in.
9. digital marketing method according to claim 8 is it is characterised in that described advertisement cheating stream screening module is from being received
Choose certain sample size in the user access information collecting as training sample, remaining sample size as test sample, then
The index of correlation extracted user property from the Subscriber Management System of advertiser and access behavior, and standard is carried out to all indexs
Change is processed, and determines described characteristic index using the characteristic equation of normalized matrix.
10. digital marketing method according to claim 8 is it is characterised in that also include step (4): in step (3) institute really
Fixed each is thrown in the unique monitoring of channel setting and is linked with picture the effective monitoring it is ensured that to data, when preferential channels advertisement
Throw in after certain time and user do not react, automatically proceeded to the next of correlation and throw in channel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610700319.4A CN106372959A (en) | 2016-08-22 | 2016-08-22 | Internet-based user access behavior digital marketing system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610700319.4A CN106372959A (en) | 2016-08-22 | 2016-08-22 | Internet-based user access behavior digital marketing system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106372959A true CN106372959A (en) | 2017-02-01 |
Family
ID=57878716
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610700319.4A Pending CN106372959A (en) | 2016-08-22 | 2016-08-22 | Internet-based user access behavior digital marketing system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106372959A (en) |
Cited By (32)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107194532A (en) * | 2017-04-07 | 2017-09-22 | 广东精点数据科技股份有限公司 | Insurance business analysis method based on big data |
| CN107609899A (en) * | 2017-08-24 | 2018-01-19 | 天脉聚源（北京）科技有限公司 | A kind of intelligence sends the method and device of advertisement |
| CN107679915A (en) * | 2017-09-30 | 2018-02-09 | 平安科技（深圳）有限公司 | Management method, device, equipment and the computer-readable storage medium of marketing activity |
| CN107742253A (en) * | 2017-09-29 | 2018-02-27 | 搜易贷（北京）金融信息服务有限公司 | A kind of automation method for running based on user behavior analysis |
| WO2018145652A1 (en) * | 2017-02-09 | 2018-08-16 | 腾讯科技（深圳）有限公司 | Media data processing method, server and storage medium |
| CN108830650A (en) * | 2018-06-11 | 2018-11-16 | 玩咖欢聚文化传媒（北京）有限公司 | Flow quantity intelligent is layered management-control method and system |
| CN108985810A (en) * | 2017-06-02 | 2018-12-11 | 北京京东尚科信息技术有限公司 | A kind of method and apparatus that party in request's platform carries out advertisement dispensing |
| CN109034888A (en) * | 2018-07-18 | 2018-12-18 | 天津璧合信息技术有限公司 | A kind of data put-on method and device |
| CN109492102A (en) * | 2018-11-08 | 2019-03-19 | 中国联合网络通信集团有限公司 | User data processing method, device, equipment and readable storage medium storing program for executing |
| CN109559149A (en) * | 2018-10-17 | 2019-04-02 | 杭州家娱互动网络科技有限公司 | A kind of flow identifying processing method and device |
| CN109558396A (en) * | 2018-10-24 | 2019-04-02 | 深圳市万屏时代科技有限公司 | A kind of user demand data cleaning method and system |
| CN109586990A (en) * | 2017-09-29 | 2019-04-05 | 北京国双科技有限公司 | A kind of method and device of identification cheating flow |
| CN109615442A (en) * | 2019-01-23 | 2019-04-12 | 上海旺翔文化传媒股份有限公司 | RTB real time bid method based on excitation video ads |
| CN109978594A (en) * | 2017-12-28 | 2019-07-05 | 北京京东尚科信息技术有限公司 | Order processing method, apparatus and medium |
| CN110097066A (en) * | 2018-01-31 | 2019-08-06 | 阿里巴巴集团控股有限公司 | A kind of user classification method, device and electronic equipment |
| CN110348745A (en) * | 2019-07-12 | 2019-10-18 | 西安点告网络科技有限公司 | The ranking method and device of advertising channel |
| CN110400095A (en) * | 2019-08-02 | 2019-11-01 | 秒针信息技术有限公司 | A kind of information resources distribution method and device |
| CN110503486A (en) * | 2019-08-28 | 2019-11-26 | 北京深演智能科技股份有限公司 | A kind of screening technique and device of advertising strategy |
| CN110544114A (en) * | 2019-08-16 | 2019-12-06 | 北京市天元网络技术股份有限公司 | Method and device for identifying and rating user group aiming at marketing preference |
| WO2019242144A1 (en) * | 2018-06-19 | 2019-12-26 | 平安科技（深圳）有限公司 | Electronic device, preference tendency prediction method and computer readable storage medium |
| CN110659943A (en) * | 2019-09-27 | 2020-01-07 | 上海淇玥信息技术有限公司 | Method and device for adjusting delivery channel based on user structure characteristics and electronic equipment |
| CN111160991A (en) * | 2019-12-31 | 2020-05-15 | 恩亿科(北京)数据科技有限公司 | PDB advertisement traffic optimization method, device, storage medium and electronic equipment |
| CN111401995A (en) * | 2020-03-09 | 2020-07-10 | 成都欧魅时尚科技有限责任公司 | System for realizing automatic material preparation by utilizing internet advertisement |
| WO2020248508A1 (en) * | 2019-06-14 | 2020-12-17 | 苏州开心盒子软件有限公司 | Method, device and apparatus for evaluating advertising channel, and storage medium |
| CN112275438A (en) * | 2020-10-13 | 2021-01-29 | 成都智叟智能科技有限公司 | Dry and wet garbage separation and crushing control method and system based on data analysis |
| CN112308584A (en) * | 2019-07-26 | 2021-02-02 | 中国移动通信集团湖南有限公司 | Marketing model training method and electronic equipment |
| CN112465545A (en) * | 2020-11-26 | 2021-03-09 | 上海移卓网络科技有限公司 | Method and device for confirming advertisement delivery abnormal channel and computer equipment |
| CN112465549A (en) * | 2020-11-30 | 2021-03-09 | 上海酷量信息技术有限公司 | System and method for identifying channel cheating |
| CN112506907A (en) * | 2020-12-10 | 2021-03-16 | 北谷电子有限公司 | Engineering machinery marketing strategy pushing method, system and device based on big data |
| CN113449107A (en) * | 2021-06-29 | 2021-09-28 | 金陵科技学院 | Distributed self-adaptive access control method facing to geographic big data |
| CN116308536A (en) * | 2023-02-17 | 2023-06-23 | 欢聚时代文化传媒（北京）有限公司 | Advertisement flow data acquisition and analysis system |
| CN116385052A (en) * | 2023-02-07 | 2023-07-04 | 网舟联合科技(北京)有限公司 | Method for analyzing potential user characteristics based on Bayesian prediction embedded point document value |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104050178A (en) * | 2013-03-13 | 2014-09-17 | 北京思博途信息技术有限公司 | Internet monitoring anti-spamming method and device |
| CN104700289A (en) * | 2015-03-17 | 2015-06-10 | 中国联合网络通信集团有限公司 | Advertising method and device |
| CN105069654A (en) * | 2015-08-07 | 2015-11-18 | 新一站保险代理有限公司 | User identification based website real-time/non-real-time marketing investment method and system |
| CN105741133A (en) * | 2016-01-26 | 2016-07-06 | 电子科技大学 | Stepping management method for online advertisement delivery |
-
2016
- 2016-08-22 CN CN201610700319.4A patent/CN106372959A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104050178A (en) * | 2013-03-13 | 2014-09-17 | 北京思博途信息技术有限公司 | Internet monitoring anti-spamming method and device |
| CN104700289A (en) * | 2015-03-17 | 2015-06-10 | 中国联合网络通信集团有限公司 | Advertising method and device |
| CN105069654A (en) * | 2015-08-07 | 2015-11-18 | 新一站保险代理有限公司 | User identification based website real-time/non-real-time marketing investment method and system |
| CN105741133A (en) * | 2016-01-26 | 2016-07-06 | 电子科技大学 | Stepping management method for online advertisement delivery |
Cited By (44)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108415908B (en) * | 2017-02-09 | 2021-12-10 | 腾讯科技（北京）有限公司 | Multimedia data processing method and server |
| WO2018145652A1 (en) * | 2017-02-09 | 2018-08-16 | 腾讯科技（深圳）有限公司 | Media data processing method, server and storage medium |
| CN108415908A (en) * | 2017-02-09 | 2018-08-17 | 腾讯科技（北京）有限公司 | A kind of processing method and server of multi-medium data |
| CN107194532A (en) * | 2017-04-07 | 2017-09-22 | 广东精点数据科技股份有限公司 | Insurance business analysis method based on big data |
| CN108985810A (en) * | 2017-06-02 | 2018-12-11 | 北京京东尚科信息技术有限公司 | A kind of method and apparatus that party in request's platform carries out advertisement dispensing |
| CN107609899A (en) * | 2017-08-24 | 2018-01-19 | 天脉聚源（北京）科技有限公司 | A kind of intelligence sends the method and device of advertisement |
| CN109586990A (en) * | 2017-09-29 | 2019-04-05 | 北京国双科技有限公司 | A kind of method and device of identification cheating flow |
| CN107742253A (en) * | 2017-09-29 | 2018-02-27 | 搜易贷（北京）金融信息服务有限公司 | A kind of automation method for running based on user behavior analysis |
| CN109586990B (en) * | 2017-09-29 | 2021-11-02 | 北京国双科技有限公司 | Method and device for identifying cheating flow |
| CN107679915A (en) * | 2017-09-30 | 2018-02-09 | 平安科技（深圳）有限公司 | Management method, device, equipment and the computer-readable storage medium of marketing activity |
| CN109978594A (en) * | 2017-12-28 | 2019-07-05 | 北京京东尚科信息技术有限公司 | Order processing method, apparatus and medium |
| CN109978594B (en) * | 2017-12-28 | 2021-11-26 | 北京京东尚科信息技术有限公司 | Order processing method, device and medium |
| CN110097066B (en) * | 2018-01-31 | 2024-01-05 | 阿里巴巴集团控股有限公司 | User classification method and device and electronic equipment |
| CN110097066A (en) * | 2018-01-31 | 2019-08-06 | 阿里巴巴集团控股有限公司 | A kind of user classification method, device and electronic equipment |
| CN108830650A (en) * | 2018-06-11 | 2018-11-16 | 玩咖欢聚文化传媒（北京）有限公司 | Flow quantity intelligent is layered management-control method and system |
| CN108830650B (en) * | 2018-06-11 | 2021-07-27 | 玩咖欢聚文化传媒（北京）有限公司 | Intelligent flow layered management and control method and system |
| WO2019242144A1 (en) * | 2018-06-19 | 2019-12-26 | 平安科技（深圳）有限公司 | Electronic device, preference tendency prediction method and computer readable storage medium |
| CN109034888A (en) * | 2018-07-18 | 2018-12-18 | 天津璧合信息技术有限公司 | A kind of data put-on method and device |
| CN109559149A (en) * | 2018-10-17 | 2019-04-02 | 杭州家娱互动网络科技有限公司 | A kind of flow identifying processing method and device |
| CN109558396A (en) * | 2018-10-24 | 2019-04-02 | 深圳市万屏时代科技有限公司 | A kind of user demand data cleaning method and system |
| CN109492102A (en) * | 2018-11-08 | 2019-03-19 | 中国联合网络通信集团有限公司 | User data processing method, device, equipment and readable storage medium storing program for executing |
| CN109615442A (en) * | 2019-01-23 | 2019-04-12 | 上海旺翔文化传媒股份有限公司 | RTB real time bid method based on excitation video ads |
| WO2020248508A1 (en) * | 2019-06-14 | 2020-12-17 | 苏州开心盒子软件有限公司 | Method, device and apparatus for evaluating advertising channel, and storage medium |
| CN110348745A (en) * | 2019-07-12 | 2019-10-18 | 西安点告网络科技有限公司 | The ranking method and device of advertising channel |
| CN112308584A (en) * | 2019-07-26 | 2021-02-02 | 中国移动通信集团湖南有限公司 | Marketing model training method and electronic equipment |
| CN110400095A (en) * | 2019-08-02 | 2019-11-01 | 秒针信息技术有限公司 | A kind of information resources distribution method and device |
| CN110544114A (en) * | 2019-08-16 | 2019-12-06 | 北京市天元网络技术股份有限公司 | Method and device for identifying and rating user group aiming at marketing preference |
| CN110503486A (en) * | 2019-08-28 | 2019-11-26 | 北京深演智能科技股份有限公司 | A kind of screening technique and device of advertising strategy |
| CN110659943B (en) * | 2019-09-27 | 2023-03-31 | 上海淇玥信息技术有限公司 | Method and device for adjusting delivery channel based on user structure characteristics and electronic equipment |
| CN110659943A (en) * | 2019-09-27 | 2020-01-07 | 上海淇玥信息技术有限公司 | Method and device for adjusting delivery channel based on user structure characteristics and electronic equipment |
| CN111160991B (en) * | 2019-12-31 | 2023-08-29 | 恩亿科(北京)数据科技有限公司 | PDB advertisement flow optimization method and device, storage medium and electronic equipment |
| CN111160991A (en) * | 2019-12-31 | 2020-05-15 | 恩亿科(北京)数据科技有限公司 | PDB advertisement traffic optimization method, device, storage medium and electronic equipment |
| CN111401995A (en) * | 2020-03-09 | 2020-07-10 | 成都欧魅时尚科技有限责任公司 | System for realizing automatic material preparation by utilizing internet advertisement |
| CN112275438A (en) * | 2020-10-13 | 2021-01-29 | 成都智叟智能科技有限公司 | Dry and wet garbage separation and crushing control method and system based on data analysis |
| CN112275438B (en) * | 2020-10-13 | 2022-03-01 | 成都智叟智能科技有限公司 | Dry and wet garbage separation and crushing control method and system based on data analysis |
| CN112465545A (en) * | 2020-11-26 | 2021-03-09 | 上海移卓网络科技有限公司 | Method and device for confirming advertisement delivery abnormal channel and computer equipment |
| CN112465545B (en) * | 2020-11-26 | 2022-11-22 | 上海移卓网络科技有限公司 | Method and device for confirming advertisement delivery abnormal channel and computer equipment |
| CN112465549A (en) * | 2020-11-30 | 2021-03-09 | 上海酷量信息技术有限公司 | System and method for identifying channel cheating |
| CN112506907A (en) * | 2020-12-10 | 2021-03-16 | 北谷电子有限公司 | Engineering machinery marketing strategy pushing method, system and device based on big data |
| CN113449107B (en) * | 2021-06-29 | 2022-03-18 | 金陵科技学院 | Distributed self-adaptive access control method facing to geographic big data |
| CN113449107A (en) * | 2021-06-29 | 2021-09-28 | 金陵科技学院 | Distributed self-adaptive access control method facing to geographic big data |
| CN116385052A (en) * | 2023-02-07 | 2023-07-04 | 网舟联合科技(北京)有限公司 | Method for analyzing potential user characteristics based on Bayesian prediction embedded point document value |
| CN116308536A (en) * | 2023-02-17 | 2023-06-23 | 欢聚时代文化传媒（北京）有限公司 | Advertisement flow data acquisition and analysis system |
| CN116308536B (en) * | 2023-02-17 | 2023-09-26 | 欢聚时代文化传媒（北京）有限公司 | Advertisement flow data acquisition and analysis system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106372959A (en) | Internet-based user access behavior digital marketing system and method | |
| Xu et al. | Does access to internet promote innovation? A look at the US broadband industry | |
| US20180293612A1 (en) | Digital information gathering and analyzing method and apparatus | |
| CN104063801B (en) | A kind of moving advertising recommend method based on cluster | |
| US20160048934A1 (en) | Property Scoring System & Method | |
| TWI391867B (en) | Method for scoring user click and click traffic scoring system thereof | |
| US20160092959A1 (en) | Tag Based Property Platform & Method | |
| CN101802856A (en) | Measuring a location based advertising campaign | |
| CN108154379B (en) | Media information publishing method and device | |
| JP2015011712A (en) | Digital information gathering and analyzing method and apparatus | |
| CN103489117A (en) | Method and system for information releasing | |
| CN103426098B (en) | A kind of content promotion method and content Extension Software Platform | |
| CN112307329A (en) | Resource recommendation method and device, equipment and storage medium | |
| CN108011740A (en) | A kind of media flow data processing method and device | |
| CN108306973A (en) | House property medium services platform user stage division and its platform | |
| CN107077498A (en) | The presentation-entity relation in online advertisement | |
| CN108429776A (en) | Method for pushing, device, client, interactive device and the system of network object | |
| CN105260913A (en) | CTR estimation method and system, and DSP server used for Internet advertisement putting | |
| CN106408325A (en) | User consumption behavior prediction analysis method based on user payment information and system | |
| CN106130756A (en) | A kind of method and device predicting access content clicking rate | |
| CN103761228A (en) | Ranking threshold determination method and ranking threshold determination system for application program | |
| CN109598543A (en) | A kind of method and system obtaining result from a variety of dimension conditions | |
| CN105978729B (en) | A kind of cellphone information supplying system and method based on user's internet log and position | |
| CN107104875B (en) | Information pushing method and device | |
| CN113362095A (en) | Information delivery method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170201 |