CN109711885A - Motivate video ads intelligence put-on method - Google Patents
Motivate video ads intelligence put-on method Download PDFInfo
- Publication number
- CN109711885A CN109711885A CN201811607597.0A CN201811607597A CN109711885A CN 109711885 A CN109711885 A CN 109711885A CN 201811607597 A CN201811607597 A CN 201811607597A CN 109711885 A CN109711885 A CN 109711885A
- Authority
- CN
- China
- Prior art keywords
- crowd
- user
- attribute
- advertisement
- basis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a kind of excitation video ads intelligence put-on methods, comprising: establishes customer data base, obtains user data；User data is standardized；User's crowd portrayal is portrayed, matches multiple crowd's attribute tags for each user；Target group is divided into core crowd, strong correlation crowd and radiation crowd；CTR prediction model is established, clicking rate is carried out and estimates；Basic appraisal is carried out to core crowd, strong correlation crowd and radiation crowd respectively according to the clicking rate of prediction, and is adjusted in real time according to trading situation；Evaluated using basis and carry out advertisement dispensing processing, precisely launched by lock onto target crowd, completion advertisement dispensing is shown to be handled with realization.The present invention passes through the advertisement serving policy for the customized personalization of different user, the personalization for realizing Internet advertising is precisely launched cashes with maximization flow, it avoids that no purpose is delivered to enterprise and advertiser brings waste in resource and cost, reaches optimal marketing effectiveness using a small amount of resource.
Description
Technical field
The present invention relates to a kind of advertisement placement methods, and in particular to a kind of excitation video ads intelligence put-on method.
Background technique
With the development of internet technology, big data technology is by feat of excellent data collection and analysis technology to internet
Advertisement brings innovation and progress, has more and more advertisers in Internet advertising industry at present and is sought using big data
Pin locks specific crowd by the maintenance data algorithm in advertisement is launched and crowd's directional technology, finally pushes to product and disappear
Fei Zhe, however the clicking rate of dispensing advertisement and conversion ratio be not high, is on the one hand since big data technological development is incomplete, still
It so comes with some shortcomings, such as is difficult to be purified and be stored in face of mass data, hidden property, ineffectivity and the contingency of data
It has a certain impact to the accuracy tool of finally obtained analysis data, the dynamic role of audient is caused to be underestimated.On the other hand, when
Internet user's fast transition oneself in certain period of time concern hobby when, third party technology company be difficult to accurately
Audient's hobby is held, be easy to cause certain deviation, this no purpose advertisement serving policy can bring resource to enterprise and advertiser
With the waste in cost.
Summary of the invention
Aiming at the shortcomings in the prior art, the object of the present invention is to provide a kind of excitation video ads intelligence put-on method,
By the advertisement serving policy for the customized personalization of different user, with realize the personalization of Internet advertising precisely launch with
It maximizes flow to cash, improves the clicking rate and conversion ratio of advertisement.
The purpose of the present invention is adopt the following technical solutions realization:
A kind of excitation video ads intelligence put-on method, comprising:
Customer data base is established, obtains user data in real time；
User data is standardized；
User's crowd portrayal is portrayed, is that each user matches multiple crowd's attribute tags according to crowd portrayal；
Calculate user's crowd portrayal attribute and advertising media's attribute correlation degree, according to correlation degree to target group into
Row divides；
CTR prediction model is established, clicking rate is carried out to core crowd, strong correlation crowd and radiation crowd respectively and is estimated；
Basic appraisal and real-time is carried out to core crowd, strong correlation crowd and radiation crowd respectively according to the clicking rate of prediction
Adjustment；
Evaluated according to basis and carry out advertisement bidding dispensing, completion advertisement dispensing is shown to be handled with realization.
Further, the customer data base is according to the variation real-time update of user behavior data.
Further, the user data is to record user browsing behavior using being temporarily stored in user computer for DSP acquisition
With the cookies data of browse state.
Further, described that user data is standardized, i.e., each user data is limited in by calculating
In [0,1] range, so that the user data variate-value by standardization is fluctuated around about 0, and variance is 1.
Further, user's crowd portrayal of portraying method particularly includes: acquisition user is in different periods when logging in
Between, age, gender, interest, the regional information where user and history advertisement CTR information characteristics data, analyze user in difference
The interest and behavioural characteristic of period, it is emerging according to primary attribute, media environment, user environment, tool application, commercial interest, social activity
Interest, media interests, entertainment interest, User Status, wish behavior and ten dimension of space attribute portray user's crowd portrayal.
Further, the correlation degree for calculating user's crowd portrayal attribute and advertising media's attribute, to target group
It is divided, i.e., according to the dimension registration of user's crowd portrayal attribute and advertising media's attribute, target group is divided into core
Heart crowd, strong correlation crowd and radiation crowd, specific division methods are as follows:
By user's crowd portrayal Attribute transposition be basic attribute, media environment, user environment, tool application, commercial interest,
Social interests, media interests, entertainment interest, User Status, wish behavior and space attribute totally ten one level-one dimensions；
By advertising media's Attribute transposition be essential attribute, media categories, quality type, industrial nature, user group, region,
Page type, keyword and content of pages quality totally nine level-one dimensions；
User's crowd portrayal attribute label corresponding with advertising media's each dimension of attribute is matched, each dimension setting
Corresponding weighted value is q1；
The relevance for matching label is divided into 3 grades, setting weighted value is q2；
It is weighted according to the corresponding dimension of matching label with property coefficient is associated with, obtains the percentage of dimension registration
Value；
When dimension registration is greater than 70%, belong to core crowd, 50%~70% belongs to strong correlation crowd, less than 50%
Belong to radiation crowd.
Further, the CTR prediction model is by input layer, Embedding layers, Product layers and 2 layers of hidden layer
The 5 layers of PNN network structure model constituted；
The input layer is for inputting discrete feature vector；
Described Embedding layers is obtained by whole network training, for discrete feature vector to be embedded into higher-dimension
Continuous space obtains continuous feature value vector；
Described Product layers, for carrying out characteristic crossover, is first concluded Embedding layers of continuous characteristic vector to phase
It answers classification and obtains z vector, then any two feature vector in Embedding layers is done into inner product and apposition obtains P vector, most
Afterwards by z vector sum P vector together as the input of neural network；
The hidden layer is for exporting CTR prediction result.
Further, described that basis appraisal is carried out to core crowd, strong correlation crowd and radiation crowd and is adjusted in real time
Method are as follows: using Bayes's smoothing technique preset CTR initial value, according to current click volume and light exposure to CTR initial value by
Step amendment, in conjunction with advertiser budget, KPI and advertisement position unit price, gradually correct bid results, finally obtain core crowd,
The basis appraisal of strong correlation crowd and radiation crowd.
Further, the basis appraisal of the core crowd is higher than the basis appraisal of strong correlation crowd, strong correlation crowd's
Basis appraisal is higher than the basis appraisal of radiation crowd.
Further, described that the process for carrying out advertisement bidding dispensing is evaluated according to basis are as follows: by being carried out to basis appraisal
Sequence acquisition is optimal to bid, and exports optimal corresponding advertisement of bidding, and completes advertisement dispensing and shows and cash.
The technical solution that embodiments herein provides can include the following benefits:
Intelligent put-on method proposed by the present invention based on excitation video ads, is retouched using big data technology and intelligent algorithm
User's portrait is drawn, belongs to its own personalized marketing strategy for different users is customized, and finally by personalized production
To consumer, maintenance data algorithm and crowd's directional technology lock someone in advertisement dispensing, avoid nothing for product advertisement pushing
The extra and waste that purpose is launched, is brought the saving in resource and cost to enterprise and advertiser, is reached using a small amount of resource
Optimal marketing effectiveness, in addition, excitation video ads conversion ratio it is higher than the conversion ratio of conventional ads, excitation video ads with
By means of its good user experience, keep the retention ratio of user bigger, thus clicking rate conversion ratio is bigger, it is final to realize maximized stream
Quantitative change is existing.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is excitation video ads intelligence put-on method flow chart；
Fig. 2 is CTR prediction model structural schematic diagram；
Fig. 3 is excitation video system architecture figure；
Fig. 4 is that the advertisement intelligent based on excitation video advertisement system launches flow chart；
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, technical solution of the present invention will be carried out below
Detailed description.Obviously, described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Base
Embodiment in the present invention, those of ordinary skill in the art are obtained all without making creative work
Other embodiment belongs to the range that the present invention is protected.
First preferred embodiment:
Fig. 1 is excitation video ads intelligence put-on method flow chart, as shown in Figure 1, the described method comprises the following steps:
Customer data base is established, obtains user data in real time；
User data is standardized；
User's crowd portrayal is portrayed, is that each user matches multiple crowd's attribute tags according to crowd portrayal；
Calculate user's crowd portrayal attribute and advertising media's attribute correlation degree, according to correlation degree to target group into
Row divides；
CTR prediction model is established, clicking rate is carried out to core crowd, strong correlation crowd and radiation crowd respectively and is estimated；
Basic appraisal and real-time is carried out to core crowd, strong correlation crowd and radiation crowd respectively according to the clicking rate of prediction
Adjustment；
Evaluated according to basis and carry out advertisement bidding dispensing, completion advertisement dispensing is shown to be handled with realization.
Specifically, the present invention obtains user data in real time by establishing customer data base to track user behavior, due to
The behavior at family is in constantly variation, and therefore, customer data base can be according to the variation real-time update of user behavior.User data
Acquisition is based on the collection to cookies data, and cookie is the data that server is stored temporarily in user computer, Neng Gouji
It is more to employ cookies of the family in online browsing behavior and browse state, customer data base, then can storage size it is huger
Big target group.Cookies data are obtained in several ways by DSP, such as by burying on the website of advertiser
Point puts the invisible pixel of a lxl, when user's access advertisement main web site for the first time, DSP will obtain one
Cookie, to track behavior of the user on advertisement main web site.
Specifically, after getting a large number of users data, due to the cookie data in customer data base cannot cross-domain name,
Browser-cross read-write, i.e., the cookie of same user are different on advertisement main web site, third party website and DMP, need
The user data of acquisition is standardized, makes each user that there is unified ID to indicate, standard is carried out to user data
The principle for changing processing is that each user data is limited in [0,1] range by calculating, so that the use by standardization
User data variate-value is around about 0 fluctuation, and variance is 1.
Specifically, needing user data after to user data standardization according to crowd's attribute, region, interest
Multiple dimensions such as hobby, purchase intention are classified, and portraying user's crowd portrayal is the important hand classified to user data
Section, user's crowd portrayal is the global feature for describing user as a group with general character, such as portrays user's
10-20 years old user attribute characteristic values were labeled as 3 labeled as 2,30-40 years old labeled as 1,20-30 years old by age attribute feature,
It is labeled as 4 within 40-50 years old, label is within 50 years old or more, and unascertainable can mark is.User crowd is portrayed by various dimensions
Portrait, customer data base can match multiple crowd's attribute marks according to user's representation data of these various dimensions for each user
Label.Such as crowd's label of a user may include " male " " 30 1 40 years old " " Beijing " " being keen on sports " " automobile " label etc..
User's portrait how is accurately portrayed, is the key that launch advertisement accurately to associated user crowd, draws a portrait it portraying user
Before, the detail time information and practical experience that the present invention is demonstrated and clicks according to advertisement analyze user in different periods
Interest and behavioural characteristic, through analysis find behavioural habits of the different user in different times in section be different and it is regular can
It follows.Such as many users can see news in the time on and off duty or see the entertainments such as TV play, film, for the use of shopping
Family then would generally at night or the time of festivals or holidays logs in shopping APP etc.；In addition, the user of different age group and different sexes
Hobby be also different, the dispensing of advertisement can be carried out based on the hobby of user.In the Information Statistics of user
Middle discovery, other than the feature of time and user interest hobby, the regional information and history advertisement CTR information where user are also same
Sample has researching value, such as the number of clicks of certain three or four line city advertisements is considerably less than the ad click rates of a tier 2 cities
Number, therefore can be using this type of information as user characteristics, and pass through some simple clustering methods, each country of selection and ground
Regional feature obvious user in area's carries out cluster operation.By analyzing above, the present invention divides time into 28
One week is divided into seven days, is divided into four different periods according to morning, noon, evening, morning daily by dimension, by acquiring user
In the regional information and history advertisement CTR information characteristics number where the landing time of each period, age, gender, interest, user
According to user is in the interest and behavioural characteristic of different periods for analysis, answers according to primary attribute, media environment, user environment, tool
With, commercial interest, social interests, media interests, entertainment interest, User Status, wish behavior and ten dimension of space attribute carve
Draw user's crowd portrayal.
Specifically, it is described calculate user's crowd portrayal attribute and advertising media's attribute correlation degree, to target group into
Row divides, i.e., according to the dimension registration of user's crowd portrayal attribute and advertising media's attribute, target group is divided into core
Crowd, strong correlation crowd and radiation crowd, specific division methods are as follows:
By user's crowd portrayal Attribute transposition be basic attribute, media environment, user environment, tool application, commercial interest,
Social interests, media interests, entertainment interest, User Status, wish behavior and space attribute totally ten one level-one dimensions；
By advertising media's Attribute transposition be essential attribute, media categories, quality type, industrial nature, user group, region,
Page type, keyword and content of pages quality totally nine level-one dimensions；
User's crowd portrayal attribute label corresponding with advertising media's each dimension of attribute is matched, each dimension setting
Corresponding weighted value is q1；
The relevance for matching label is divided into 3 grades, setting weighted value is q2；
It is weighted according to the corresponding dimension of matching label with property coefficient is associated with, obtains the percentage of dimension registration
Value；
When dimension registration is greater than 70%, belong to core crowd, 50%~70% belongs to strong correlation crowd, less than 50%
Belong to radiation crowd.
Specifically, the present invention is by establishing CTR prediction model shown in Fig. 2, respectively to core crowd, strong correlation crowd and
Radiation crowd carries out clicking rate and estimates, and the CTR prediction model is by input layer, Embedding layers, Product layers and hidden
Hide 5 layers of PNN network structure model that layer 1 and hidden layer 2 are constituted；The input layer is for inputting discrete feature vector；It is described
Embedding layers are obtained by whole network training, for discrete feature vector to be embedded into higher-dimension continuous space, are obtained
Continuous feature value vector；Described Product layers is used to carry out characteristic crossover, first by Embedding layers of continuous characteristic vector
It concludes respective classes and obtains z vector, then any two feature vector in Embedding layers is done into inner product and apposition obtains
P vector, finally by z vector sum P vector together as the input of neural network；The hidden layer is for exporting CTR (clicking rate)
Prediction result.The model in addition to obtaining z vector, also adds a p vector on the basis of FNN model, i.e. Product to
Amount.Product vector does inner product by the feature vector of each category regions or apposition obtains, and facilitates characteristic crossover.In addition PNN
Middle Embeding layers is no longer generated by FM, can be trained and be obtained in the entire network.The present invention is distinguished pre- using CTR prediction model
Core crowd, strong correlation crowd and the clicking rate for radiating crowd are surveyed, and then basic appraisal is carried out to above-mentioned crowd, further according to reality
Trading situation adjust in real time.During adjusting real-valued, CTR initial value is preset first with Bayes's smoothing technique,
CTR initial value is gradually corrected according to current click volume and light exposure, it is monovalent in conjunction with advertiser budget, KPI and advertisement position,
Bid results are gradually corrected, finally obtain the basis appraisal of core crowd, strong correlation crowd and radiation crowd, wherein core crowd
Basis appraisal be higher than strong correlation crowd basis appraisal, strong correlation crowd basis appraisal be higher than radiation crowd basis estimate
Valence.Finally by basis appraisal is ranked up obtain it is optimal bid, export optimal corresponding advertisement of bidding, complete advertisement and launch
It shows and cashes.
Second preferred embodiment:
The present invention realizes that the intelligence of excitation video ads is launched using excitation video advertisement system shown in Fig. 3, the excitation
Video advertisement system is the advertisement marketing cloud platform based on big data, and platform service includes both parties, is become for flow
It is existing, realize the sequencing transaction of advertisement both parties, balancing user experience, maximizing realizes that flow is cashed.The dispensing of advertisement
Process initiates dispensing ad-request to advertisement delivery system as shown in figure 4, user passes through Webpage search keyword first, works as advertisement
After jettison system receives dispensing ad-request, relevant advertisements request is initiated to operation system of bidding, it is complete by operation system of bidding
At the matching treatment of ad-request, generates relevant advertisements information and feed back to advertisement delivery system, advertisement delivery system carries out at this time
Advertisement, which is launched, handles and generates advertisement dispensing queue, finally sends advertisement from advertisement delivery system to webpage, completes advertisement and launch
It shows and is handled with realization.
User obtains virtual reward by viewing video ads, and then improves the clicking rate and conversion ratio of advertisement, realizes
Advertisement value maximizes.Avoid that no purpose is delivered to enterprise and advertiser brings waste in resource and cost, using a small amount of
Resource reaches optimal marketing effectiveness.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.Therefore, protection scope of the present invention should be based on the protection scope of the described claims.
Claims (10)
1. a kind of excitation video ads intelligence put-on method characterized by comprising
Customer data base is established, obtains user data in real time；
User data is standardized；
User's crowd portrayal is portrayed, is that each user matches multiple crowd's attribute tags according to crowd portrayal；
The correlation degree for calculating user's crowd portrayal attribute and advertising media's attribute, draws target group according to correlation degree
Point；
CTR prediction model is established, clicking rate is carried out to core crowd, strong correlation crowd and radiation crowd respectively and is estimated；
Basis appraisal is carried out to core crowd, strong correlation crowd and radiation crowd respectively according to the clicking rate of prediction and is adjusted in real time
It is whole；
Evaluated according to basis and carry out advertisement bidding dispensing, completion advertisement dispensing is shown to be handled with realization.
2. a kind of excitation video ads intelligence put-on method according to claim 1, which is characterized in that the user data
Library is according to the variation real-time update of user behavior data.
3. a kind of excitation video ads intelligence put-on method according to claim 1, which is characterized in that the user data
It is the cookies data for being temporarily stored in record user browsing behavior and browse state in user computer using DSP acquisition.
4. a kind of excitation video ads intelligence put-on method according to claim 1, which is characterized in that described to number of users
According to being standardized, i.e., each user data is limited in [0,1] range by calculating, so that by standardization
User data variate-value around about 0 fluctuation, and variance be 1.
5. a kind of excitation video ads intelligence put-on method according to claim 1, which is characterized in that described to portray user
Crowd portrayal method particularly includes: user is on the ground where the landing time of different periods, age, gender, interest, user for acquisition
Domain information and history advertisement CTR information characteristics data, analysis user belong in the interest and behavioural characteristic of different periods according to basis
Property, media environment, user environment, tool application, commercial interest, social interests, media interests, entertainment interest, User Status, meaning
Hope behavior and ten dimension of space attribute portray user's crowd portrayal.
6. a kind of excitation video ads intelligence put-on method according to claim 1, which is characterized in that the calculating user
The correlation degree of crowd portrayal attribute and advertising media's attribute, divides target group, i.e., according to user's crowd portrayal category
Property with the dimension registration of advertising media's attribute, target group is divided into core crowd, strong correlation crowd and radiation crowd, tool
Body division methods are as follows:
It is basic attribute, media environment, user environment, tool application, commercial interest, social activity by user's crowd portrayal Attribute transposition
Interest, media interests, entertainment interest, User Status, wish behavior and space attribute totally ten one level-one dimensions；
It is essential attribute, media categories, quality type, industrial nature, user group, region, the page by advertising media's Attribute transposition
Type, keyword and content of pages quality totally nine level-one dimensions；
User's crowd portrayal attribute label corresponding with advertising media's each dimension of attribute is matched, each dimension setting corresponds to
Weighted value be q1；
The relevance for matching label is divided into 3 grades, setting weighted value is q2；
It is weighted according to the corresponding dimension of matching label with property coefficient is associated with, obtains the percent value of dimension registration；
When dimension registration is greater than 70%, belong to core crowd, 50%~70% belongs to strong correlation crowd, belongs to less than 50%
Radiation crowd.
7. a kind of excitation video ads intelligence put-on method according to claim 1, which is characterized in that the CTR prediction
Model is 5 layers of PNN network structure model being made of input layer, Embedding layers, Product layers and 2 layers of hidden layer；
The input layer is for inputting discrete feature vector；
Described Embedding layers is obtained by whole network training, continuous for discrete feature vector to be embedded into higher-dimension
Space obtains continuous feature value vector；
Described Product layers, for carrying out characteristic crossover, is first concluded Embedding layers of continuous characteristic vector to respective class
Z vector is not obtained, then any two feature vector in Embedding layer is done into inner product and apposition obtains P vector, finally general
Input of the z vector sum P vector together as neural network；
The hidden layer is for exporting CTR prediction result.
8. a kind of excitation video ads intelligence put-on method according to claim 1, which is characterized in that described to core people
Group, strong correlation crowd and radiation crowd carry out the method that basis is evaluated and adjusted in real time are as follows: default using Bayes's smoothing technique
CTR initial value gradually corrects CTR initial value according to current click volume and light exposure, in conjunction with advertiser budget, KPI with
And advertisement position unit price, bid results are gradually corrected, the basis for finally obtaining core crowd, strong correlation crowd and radiation crowd is estimated
Valence.
9. a kind of excitation video ads intelligence put-on method according to claim 8, which is characterized in that the core crowd
Basis appraisal be higher than strong correlation crowd basis appraisal, strong correlation crowd basis appraisal be higher than radiation crowd basis estimate
Valence.
10. a kind of excitation video ads intelligence put-on method according to claim 1, which is characterized in that described according to base
Plinth appraisal carry out advertisement bidding dispensing process are as follows: by basis appraisal be ranked up obtain it is optimal bid, export it is optimal competing
The corresponding advertisement of valence completes advertisement and launches displaying and cash.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811607597.0A CN109711885A (en) | 2018-12-27 | 2018-12-27 | Motivate video ads intelligence put-on method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811607597.0A CN109711885A (en) | 2018-12-27 | 2018-12-27 | Motivate video ads intelligence put-on method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109711885A true CN109711885A (en) | 2019-05-03 |
Family
ID=66257831
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811607597.0A Pending CN109711885A (en) | 2018-12-27 | 2018-12-27 | Motivate video ads intelligence put-on method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109711885A (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110427578A (en) * | 2019-07-16 | 2019-11-08 | 盐城师范学院 | A kind of digital content put-on method based on depth machine learning |
| CN110490658A (en) * | 2019-08-21 | 2019-11-22 | 北京字节跳动网络技术有限公司 | User's motivational techniques, device, electronic equipment and the readable medium of application program |
| CN110704751A (en) * | 2019-10-22 | 2020-01-17 | 北京字节跳动网络技术有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN110807655A (en) * | 2019-10-15 | 2020-02-18 | 微梦创科网络科技（中国）有限公司 | Advertisement bidding method, device and equipment |
| CN110956490A (en) * | 2019-11-07 | 2020-04-03 | 微梦创科网络科技（中国）有限公司 | Method and device for improving advertisement putting completion rate |
| CN110991490A (en) * | 2019-11-11 | 2020-04-10 | 广州荔支网络技术有限公司 | Intelligent audio platform advertisement putting method |
| CN110992104A (en) * | 2019-12-11 | 2020-04-10 | 杭州古点网络科技有限公司 | Abnormal advertisement flow detection method based on user portrait |
| CN111612516A (en) * | 2020-04-11 | 2020-09-01 | 上海淇玥信息技术有限公司 | DMP platform and method based on RTA delivery and electronic equipment |
| CN111611155A (en) * | 2020-04-11 | 2020-09-01 | 上海淇玥信息技术有限公司 | RTA (remote terminal installation) DMP (distributed file protocol) releasing platform and method for realizing AB (AB test) and electronic equipment |
| CN111681057A (en) * | 2020-06-11 | 2020-09-18 | 北京深演智能科技股份有限公司 | Method and device for processing media resource for information delivery |
| CN112150176A (en) * | 2019-06-28 | 2020-12-29 | 阿里巴巴集团控股有限公司 | Evaluation method, system and readable storage medium |
| CN113282690A (en) * | 2020-02-19 | 2021-08-20 | 百度在线网络技术（北京）有限公司 | Method, device, equipment and storage medium for ordering interest point recalls |
| CN113298119A (en) * | 2021-04-28 | 2021-08-24 | 上海淇玥信息技术有限公司 | Method and device for evaluating putting strategy of machine learning model and electronic equipment |
| CN114266606A (en) * | 2022-03-03 | 2022-04-01 | 北京派瑞威行互联技术有限公司 | Short video advertisement putting optimization evaluation method and system |
| CN114581128A (en) * | 2022-02-28 | 2022-06-03 | 北京沃东天骏信息技术有限公司 | Advertisement putting method and device, electronic equipment and computer readable medium |
| CN115292528A (en) * | 2022-08-26 | 2022-11-04 | 青岛心法传媒科技有限公司 | Intelligent operation method, equipment and storage medium for new media video |
| CN115423537A (en) * | 2022-11-02 | 2022-12-02 | 北京车讯互联网股份有限公司 | Internet automobile industry accurate advertisement putting method based on artificial intelligence |
| CN115796959A (en) * | 2022-12-02 | 2023-03-14 | 嘉兴凌巨网络科技有限公司 | Advertisement putting effect detection method based on data acquisition and analysis |
| CN117010939A (en) * | 2023-06-12 | 2023-11-07 | 广州数说故事信息科技有限公司 | Crowd dynamic diagram dynamic adjustment method based on deep learning |
| CN118096266A (en) * | 2024-04-18 | 2024-05-28 | 上海海洋大学 | Intelligent advertisement marketing system and method based on Internet |
| CN118134553A (en) * | 2024-05-08 | 2024-06-04 | 深圳爱巧网络有限公司 | E-commerce and explosion type multi-platform collaborative pushing system, method, equipment and medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105574159A (en) * | 2015-12-16 | 2016-05-11 | 浙江汉鼎宇佑金融服务有限公司 | Big data-based user portrayal establishing method and user portrayal management system |
| CN106504099A (en) * | 2015-09-07 | 2017-03-15 | 国家计算机网络与信息安全管理中心 | A kind of system for building user's portrait |
| CN107093115A (en) * | 2017-05-10 | 2017-08-25 | 杭州纸箱哥文化传播有限公司 | The advertisement carton method for customizing and system of a kind of precision marketing |
| CN108122122A (en) * | 2016-11-29 | 2018-06-05 | 腾讯科技（深圳）有限公司 | Advertisement placement method and system |
-
2018
- 2018-12-27 CN CN201811607597.0A patent/CN109711885A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106504099A (en) * | 2015-09-07 | 2017-03-15 | 国家计算机网络与信息安全管理中心 | A kind of system for building user's portrait |
| CN105574159A (en) * | 2015-12-16 | 2016-05-11 | 浙江汉鼎宇佑金融服务有限公司 | Big data-based user portrayal establishing method and user portrayal management system |
| CN108122122A (en) * | 2016-11-29 | 2018-06-05 | 腾讯科技（深圳）有限公司 | Advertisement placement method and system |
| CN107093115A (en) * | 2017-05-10 | 2017-08-25 | 杭州纸箱哥文化传播有限公司 | The advertisement carton method for customizing and system of a kind of precision marketing |
Non-Patent Citations (2)
| Title |
|---|
| YANRU QU: "Product-based Neural Networks for User ResponsePrediction", 《2016 IEEE 16TH INTERNATIONAL CONFERENCE ON DATA MINING》 * |
| 鞠宏磊: "改写广告业的‘实时’与‘竞价’——实时竞价（RTB）广告产业链流程和运行机制研究", 《编辑之友》 * |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112150176A (en) * | 2019-06-28 | 2020-12-29 | 阿里巴巴集团控股有限公司 | Evaluation method, system and readable storage medium |
| CN110427578A (en) * | 2019-07-16 | 2019-11-08 | 盐城师范学院 | A kind of digital content put-on method based on depth machine learning |
| CN110490658A (en) * | 2019-08-21 | 2019-11-22 | 北京字节跳动网络技术有限公司 | User's motivational techniques, device, electronic equipment and the readable medium of application program |
| CN110807655A (en) * | 2019-10-15 | 2020-02-18 | 微梦创科网络科技（中国）有限公司 | Advertisement bidding method, device and equipment |
| CN110807655B (en) * | 2019-10-15 | 2022-05-17 | 微梦创科网络科技（中国）有限公司 | Advertisement bidding method, device and equipment |
| CN110704751A (en) * | 2019-10-22 | 2020-01-17 | 北京字节跳动网络技术有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN110956490A (en) * | 2019-11-07 | 2020-04-03 | 微梦创科网络科技（中国）有限公司 | Method and device for improving advertisement putting completion rate |
| CN110991490A (en) * | 2019-11-11 | 2020-04-10 | 广州荔支网络技术有限公司 | Intelligent audio platform advertisement putting method |
| CN110992104A (en) * | 2019-12-11 | 2020-04-10 | 杭州古点网络科技有限公司 | Abnormal advertisement flow detection method based on user portrait |
| CN113282690A (en) * | 2020-02-19 | 2021-08-20 | 百度在线网络技术（北京）有限公司 | Method, device, equipment and storage medium for ordering interest point recalls |
| CN113282690B (en) * | 2020-02-19 | 2024-04-02 | 百度在线网络技术（北京）有限公司 | Method, device, equipment and storage medium for ordering recall of interest points |
| CN111612516A (en) * | 2020-04-11 | 2020-09-01 | 上海淇玥信息技术有限公司 | DMP platform and method based on RTA delivery and electronic equipment |
| CN111611155A (en) * | 2020-04-11 | 2020-09-01 | 上海淇玥信息技术有限公司 | RTA (remote terminal installation) DMP (distributed file protocol) releasing platform and method for realizing AB (AB test) and electronic equipment |
| CN111681057A (en) * | 2020-06-11 | 2020-09-18 | 北京深演智能科技股份有限公司 | Method and device for processing media resource for information delivery |
| CN113298119A (en) * | 2021-04-28 | 2021-08-24 | 上海淇玥信息技术有限公司 | Method and device for evaluating putting strategy of machine learning model and electronic equipment |
| CN113298119B (en) * | 2021-04-28 | 2023-02-24 | 上海淇玥信息技术有限公司 | Method and device for evaluating putting strategy of machine learning model and electronic equipment |
| CN114581128A (en) * | 2022-02-28 | 2022-06-03 | 北京沃东天骏信息技术有限公司 | Advertisement putting method and device, electronic equipment and computer readable medium |
| CN114266606B (en) * | 2022-03-03 | 2022-05-17 | 北京派瑞威行互联技术有限公司 | Short video advertisement putting optimization evaluation method and system |
| CN114266606A (en) * | 2022-03-03 | 2022-04-01 | 北京派瑞威行互联技术有限公司 | Short video advertisement putting optimization evaluation method and system |
| CN115292528A (en) * | 2022-08-26 | 2022-11-04 | 青岛心法传媒科技有限公司 | Intelligent operation method, equipment and storage medium for new media video |
| CN115423537A (en) * | 2022-11-02 | 2022-12-02 | 北京车讯互联网股份有限公司 | Internet automobile industry accurate advertisement putting method based on artificial intelligence |
| CN115796959A (en) * | 2022-12-02 | 2023-03-14 | 嘉兴凌巨网络科技有限公司 | Advertisement putting effect detection method based on data acquisition and analysis |
| CN115796959B (en) * | 2022-12-02 | 2023-07-04 | 嘉兴凌巨网络科技有限公司 | Advertisement putting effect detection method based on data acquisition and analysis |
| CN117010939A (en) * | 2023-06-12 | 2023-11-07 | 广州数说故事信息科技有限公司 | Crowd dynamic diagram dynamic adjustment method based on deep learning |
| CN117010939B (en) * | 2023-06-12 | 2024-02-27 | 广州数说故事信息科技有限公司 | Crowd dynamic diagram dynamic adjustment method based on deep learning |
| CN118096266A (en) * | 2024-04-18 | 2024-05-28 | 上海海洋大学 | Intelligent advertisement marketing system and method based on Internet |
| CN118096266B (en) * | 2024-04-18 | 2024-07-16 | 上海海洋大学 | Intelligent advertisement marketing system and method based on Internet |
| CN118134553A (en) * | 2024-05-08 | 2024-06-04 | 深圳爱巧网络有限公司 | E-commerce and explosion type multi-platform collaborative pushing system, method, equipment and medium |
| CN118134553B (en) * | 2024-05-08 | 2024-07-19 | 深圳爱巧网络有限公司 | E-commerce and explosion type multi-platform collaborative pushing system, method, equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109711885A (en) | Motivate video ads intelligence put-on method | |
| Schüritz et al. | Capturing value from data: Revenue models for data-driven services | |
| Kazienko et al. | AdROSA—Adaptive personalization of web advertising | |
| US7548929B2 (en) | System and method for determining semantically related terms | |
| AU2011295936B2 (en) | Methods and apparatus to cluster user data | |
| US7921107B2 (en) | System for generating query suggestions using a network of users and advertisers | |
| US20180012257A1 (en) | System and methods for delivering targeted marketing offers to consumers via a distributed architecture system | |
| US20070027865A1 (en) | System and method for determining semantically related term | |
| US10282758B1 (en) | Pricing control in a real-time network-based bidding environment | |
| US20100023399A1 (en) | Personalized Advertising Using Lifestreaming Data | |
| CN107016049A (en) | Carry out personalized sponsored search layout using user behavior history | |
| CN102033883A (en) | Method, device and system for improving data transmission speed of website | |
| CN110489590A (en) | For the viewing time cluster of video search | |
| Sozinova et al. | Special aspects of studying the internet as a marketing communication channel of the service industry | |
| US10997634B2 (en) | Methods for determining targeting parameters and bids for online ad distribution | |
| CN103166976B (en) | A kind of synchronous cross-domain identify label mapping method of internet security | |
| CN109598544A (en) | A kind of media object matching process and system | |
| Soni | Role of temporal, demographic, and behavioral factors in customer conversion through dynamic creative optimization in the consumer-packaged goods setting | |
| Žitkienė et al. | Model of impact of social networks on internet marketing of enterprises | |
| US20150348098A1 (en) | Identifying A Product Placement Opportunity Within A Screenplay | |
| Kesa et al. | MSME Development Strategy for Providing Information through Central Banking Microsite Web Based | |
| Mocanu et al. | Digital Marketing Strategies: A Comprehensive Literature Review | |
| Kridel et al. | Recommender systems as a mobile marketing service | |
| KR20100086655A (en) | On-line advertising system capable of searching advertisement by click and method thereof | |
| KR101812161B1 (en) | System and method for processing advertisement targeting using advertisement storing service providing to users individually |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190503 |