CN113222647B - Advertisement recommendation method, system and storage medium based on click rate estimation model - Google Patents
Advertisement recommendation method, system and storage medium based on click rate estimation model Download PDFInfo
- Publication number
- CN113222647B CN113222647B CN202110456025.2A CN202110456025A CN113222647B CN 113222647 B CN113222647 B CN 113222647B CN 202110456025 A CN202110456025 A CN 202110456025A CN 113222647 B CN113222647 B CN 113222647B
- Authority
- CN
- China
- Prior art keywords
- model
- advertisement
- click rate
- data
- user
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000003860 storage Methods 0.000 title claims abstract description 18
- 238000012549 training Methods 0.000 claims abstract description 34
- 238000005457 optimization Methods 0.000 claims abstract description 19
- 238000002474 experimental method Methods 0.000 claims abstract description 12
- 238000012545 processing Methods 0.000 claims abstract description 12
- 238000013480 data collection Methods 0.000 claims abstract description 7
- 238000004140 cleaning Methods 0.000 claims abstract description 6
- 230000000694 effects Effects 0.000 claims abstract description 6
- 239000000463 material Substances 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 4
- 238000009826 distribution Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 238000013461 design Methods 0.000 description 5
- 238000010801 machine learning Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
-
- 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
- G06Q30/0277—Online advertisement
Abstract
Advertisement recommendation method, system and storage medium based on click rate estimation model, the recommendation method comprises: data collection, namely cleaning service data to construct a data set; carrying out feature engineering processing on the data set to generate a training sample for model training; performing model training by adopting a Wide & Deep algorithm; adjusting model structure parameters, determining the best super-parameter combination, and performing model optimization; performing an A/B experiment on the optimization model and the online model, and replacing the old model with the increased click rate; and updating the characteristic data to the cloud storage OSS in real time, generating a GRPC interface by using Tensorflow Serving, estimating the click rate of the candidate advertisement list in batches, taking the advertisement with the highest click rate, and recommending and exposing the advertisement with the click rate more than 0.5. The invention has good effect on advertisement recommendation, improves the accuracy of click rate estimation, and is beneficial to improving the recall rate of lost users.
Description
Technical Field
The invention belongs to the field of advertisement recommendation, and particularly relates to an advertisement recommendation method, an advertisement recommendation system and a storage medium based on a click rate estimation model.
Background
Click rate estimation is one of the core basis of DSP flow distribution, and meanwhile, the accuracy of click rate estimation plays a very important role in recommending online advertisements. In recent years, click rate estimation models have been converted from traditional machine learning models to Deep learning models, and common click rate estimation models include LR, FM, wide & Deep, deep fm, and the like. Due to the natural nature of advertisement data: the data volume is large, the dimension is high, and the data is sparse, so that the click rate estimation of the advertisement has certain challenges.
The data and features determine the upper bound of machine learning, and the model and algorithm approach this upper bound only indefinitely. By combining the characteristics of service data, selecting a proper algorithm model, and performing characteristic engineering innovation treatment, network structure design and optimization of an Embedding method is a key of good model landing. The accuracy of the existing model estimation is still to be improved.
Disclosure of Invention
The invention aims to solve the problem of low model estimation accuracy of advertisement data in the prior art, and provides an advertisement recommendation method, an advertisement recommendation system and a storage medium based on a click rate estimation model, which have good effect on advertisement recommendation, improve the click rate estimation accuracy and are beneficial to improving the recall rate of lost users.
In order to achieve the above purpose, the present invention has the following technical scheme:
an advertisement recommendation method based on click rate estimation model comprises the following steps:
-data collection, cleaning the business data, constructing a data set;
-feature engineering the dataset to generate training samples for use in model training;
-model training using the Wide & Deep algorithm based on the Tensorflow framework;
-adjusting model structural parameters, determining the best super-parameter combination, and performing model optimization;
performing an A/B experiment on the optimization model and the online model, and replacing the old model with the increased click rate;
-updating the feature data to the cloud storage OSS in real time, generating a GRPC interface using Tensorflow Serving, estimating click rate of the candidate advertisement list in batches, taking the advertisement with the highest click rate, and performing recommended exposure with the click rate > 0.5.
In one embodiment of the invention, the objects of the data collection include users, advertisements, request context, media; wherein the user characteristic data comprises: user ID, user commodity preference, user tag, user history funnel depth, last 3/7/14/28 day exposure times, click times, browse times, purchase times, click rate; the advertisement feature data includes: advertisement ID, advertisement category, advertisement material ID, advertisement template ID, exposure times, click times and click rate; the request context feature data includes: time, country, city, channel, media, version size, device type, display type; the media characteristic data includes: category, ranking, PV, UV, click rate.
In one embodiment of the present invention, the specific steps of feature engineering a dataset are as follows:
step 1) sample sampling, adding positive and negative sample punishment weights;
step 2) dividing the data in the data set into continuous features and discrete features, and carrying out normalization and missing value processing;
step 3) performing feature cross-over combination.
In one embodiment of the invention, the continuous features in the user feature data comprise exposure times, click times, browsing times and click rates, the continuous features in the advertisement feature data comprise exposure times, click times and click rates, the continuous features in the media feature data comprise ranking, PV, UV and historical click rates, and the continuous features are normalized by adopting a logarithmic function; and meanwhile, performing equal-frequency barrel discretization on part of continuous features, including ranking in media feature data, purchasing times and purchasing times of user feature data.
The discrete features include user ID, user merchandise preference, user tag, user history funnel depth, advertisement ID, advertisement category, advertisement material ID, advertisement template ID, time, country, city, channel, media, plate size, device type, display type, and category, and the above discrete features are One-Hot encoded using a Hash scheme.
In one embodiment of the present invention, the missing value processing specifically includes: carrying out feature distribution statistics, analyzing the missing proportion, and discarding the feature if the missing proportion reaches more than 80%; default filling is used for discrete features and mean filling is used for continuous features.
In one embodiment of the present invention, the step 3) performs feature cross-combining, where the user features, advertisement, and media features are cross-combined, including cross-combining the user merchandise preference, user tag, user history funnel depth, and advertisement ID, advertisement material ID, advertisement category, and category.
In one embodiment of the present invention, the specific steps of model training using the Wide & Deep algorithm are as follows: inputting discrete features and cross combination features in a Wide layer, inputting continuous features in a Deep layer, and taking a historical data set as a training set and a data set of the last 1 day as a test set;
performing model tuning, adding Dropout and L2 regularization to prevent the model from being fitted excessively, and introducing Batch Normalization to accelerate the convergence of the model; compared with different learners, selecting Adm with better effect; meanwhile, tuning parameters of different Learning Rate, batch Size and Learning parameters is tried to be optimized, a training model is carried out, and finally a SavedModel format model file is generated.
In one embodiment of the present invention, the specific steps for generating a GRPC interface using Tensorflow Serving include: step 1) starting a Docker, pulling a Tensorflow Servin Docker mirror image; step 2) generating a prediction interface; step 3) online deployment, configuring a unified domain name Tservice, and generating a final scoring interface service; step 4) online reasoning, constructing request data in a batch mode, calling a prediction interface, and finally realizing that 50 advertisement prediction results can be returned within 10 ms.
The invention also provides an advertisement recommendation system based on the click rate estimation model, which comprises:
the data set construction module is used for collecting data, cleaning service data and constructing a data set;
the training sample generation module is used for carrying out characteristic engineering processing on the data set to generate a training sample;
the model training module is used for carrying out model training by adopting a Wide & Deep algorithm based on a Tensorflow framework;
the model optimization module is used for adjusting model structure parameters, determining the best super-parameter combination and carrying out model optimization; performing an A/B experiment on the optimization model and the online model, and replacing the old model with the increased click rate;
and the recommending module is used for updating the characteristic data to the cloud storage OSS in real time, generating GRPC interfaces by using Tensorflow Serving, estimating click rate of candidate advertisement lists in batches, and recommending exposure according to the click rate.
The invention also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program realizes the steps of the advertisement recommendation method based on the click rate estimation model when being executed by a processor.
Compared with the prior art, the invention has at least the following beneficial effects:
the invention has good effect on advertisement recommendation, and combines the abstract features of artificial designs such as user commodity preference, user labels, user history behavior funnel depth and the like, and the user preference features and advertisement features in a crossing way, so that the model has better memory capacity; meanwhile, a Deep model is adopted, and low-dimensional dense feature input is used by an Embedding method, so that different dimensions of feature vectors are fully crossed, the generalization capability of the model is enhanced, and the recall rate of the model is improved. Meanwhile, the feature engineering part is used for excavating and analyzing the combined features based on a large amount of business data, so that the features have stronger flexibility and the interpretation of the model is enhanced. Experiments were assessed by offline AUC and online ABTest. The AUC of the Wide & Deep model exceeds that of the original LR model. In the AB Test on-line experiment, the click rate of the Wide & Deep model is improved by 27% compared with that of the LR model.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention, and that other related drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of an advertisement recommendation method based on a click rate estimation model according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, one of ordinary skill in the art may also obtain other embodiments without undue burden.
Referring to fig. 1, an advertisement recommendation method based on click rate estimation model includes the following steps:
s1, data collection is carried out, service data are cleaned, and a data set is constructed;
the data collection objects comprise users, advertisements, request contexts and media; wherein the user characteristic data comprises: user ID, user commodity preference, user tag, user history funnel depth, last 3/7/14/28 day exposure times, click times, browse times, purchase times, click rate; the advertisement feature data includes: advertisement ID, advertisement category, advertisement material ID, advertisement template ID, exposure times, click times and click rate; the request context feature data includes: time, country, city, channel, media, version size, device type, display type; the media characteristic data includes: category, ranking, PV (Page View, i.e., page View or click-through), UV (unit View), refers to the natural person accessing and browsing the web Page through the internet, and a computer client accessing the web site is a Visitor), and click-through rate.
S2, performing feature engineering processing on the data set to generate a training sample for model training;
the specific steps of the feature engineering treatment are as follows:
step 1) sample sampling, and adding positive and negative sample punishment weights.
And 2) dividing the data in the data set into continuous features and discrete features, and carrying out normalization and missing value processing.
Continuous features:
the continuous features in the user feature data comprise exposure times, clicking times, browsing times and clicking rates, the continuous features in the advertisement feature data comprise exposure times, clicking times and clicking rates, the continuous features in the media feature data comprise ranking, PV, UV and historical clicking rates, and the continuous features are normalized by adopting a logarithmic function; and meanwhile, performing equal-frequency barrel discretization on part of continuous features, including ranking in media feature data, purchasing times and purchasing times of user feature data.
The discrete features include user ID, user merchandise preference, user tag, user history funnel depth, advertisement ID, advertisement category, advertisement material ID, advertisement template ID, time, country, city, channel, media, plate size, device type, display type, and category, and the above discrete features are One-Hot encoded using a Hash scheme.
The missing value processing specifically includes: carrying out feature distribution statistics, analyzing the missing proportion, and discarding the feature if the missing proportion reaches more than 80%; default filling is used for discrete features and mean filling is used for continuous features.
Step 3) performing feature cross-over combination.
And manually performing feature cross-combination to cross user features, advertisements and media features, including cross-combination of user commodity preference, user labels, user history funnel depth, advertisement ID, advertisement material ID, advertisement category and category.
S3, model training is carried out by adopting a Wide & Deep algorithm based on a Tensorflow framework;
the method comprises the following specific steps:
and finally generating a training set through sample construction and feature engineering to serve as a training sample of the Wide & Deep. The Wide layer inputs discrete features and cross combination features, including user ID, commodity preference, tag, history funnel depth, advertisement ID, advertisement category, advertisement material ID, advertisement template ID, request time, country, city, channel, media, position size, equipment type, presentation type, media category, and cross combination of user commodity preference, user tag, history funnel depth, advertisement ID, advertisement material ID, advertisement category, media category. The Deep layer inputs are mainly continuous features, including the number of exposures, clicks, browses, click rates of the user, the number of exposures, clicks, click rates of the advertisement, and ranking, PV, UV, historical click rates of the media, in addition to discrete features. A data set of 60 days history was used as a training set and the last 1 day was used as a test set. Model tuning, namely adding Dropout and L2 regularization to prevent the model from being over-fitted, and introducing Batch Normalization to accelerate the convergence of the model. Compared with different learners, adm with better effect is selected. Meanwhile, different Learning Rate, batch Size and coding parameters are tried to be optimized. Training the model, and finally generating a SavedModel format model file.
S4, adjusting model structure parameters, determining the best super-parameter combination, and performing model optimization;
s5, performing an A/B experiment on the optimization model and the online model, and replacing the old model with the increased click rate;
and S6, updating the characteristic data to an OSS (Object Storage Service, OSS for short), namely, massive, safe and highly reliable cloud storage service provided by the Arian cloud, using Tensorflow Serving (a flexible and high-performance application system suitable for a machine learning model is specially designed for production environment, new algorithms and experiments can be easily deployed by means of Tensorflow Serving) to generate GRPC (GRPC is an RPC framework, a high-performance, open-source and general RPC framework, facing to a service end and a mobile end and based on HTTP/2 design) interfaces, estimating click rate of a candidate advertisement list in batches, taking advertisements with the highest click rate, and carrying out recommended exposure of the advertisements, wherein the new algorithms and experiments can be easily deployed by means of Tensorflow Serving.
The specific steps for generating a GRPC interface using Tensorflow Serving are as follows:
step 1) starting a Docker, pulling a Tensorflow Serving Docker mirror image; step 2) generating a prediction interface; step 3) online deployment, configuring a unified domain name Tservice, and generating a final scoring interface service; step 4) online reasoning, constructing request data in a batch mode, calling a prediction interface, and finally realizing that 50 advertisement prediction results can be returned within 10 ms.
The click rate comparison result of the Wide & Deep model and the LR model used in the invention is shown in the following table:
in the ABTest online experiment, the click rate of the Wide & Deep model is improved by 27% compared with that of the LR model.
An advertisement recommendation system based on click-through rate estimation model, comprising:
the data set construction module is used for collecting data, cleaning service data and constructing a data set;
the training sample generation module is used for carrying out characteristic engineering processing on the data set to generate a training sample;
the model training module is used for carrying out model training by adopting a Wide & Deep algorithm based on a Tensorflow framework;
the model optimization module is used for adjusting model structure parameters, determining the best super-parameter combination and carrying out model optimization; performing an A/B experiment on the optimization model and the online model, and replacing the old model with the increased click rate;
and the recommending module is used for updating the characteristic data to the cloud storage OSS in real time, generating GRPC interfaces by using Tensorflow Serving, estimating click rate of candidate advertisement lists in batches, and recommending exposure according to the click rate.
A computer readable storage medium storing a computer program which when executed by a processor performs the steps of the click rate estimation model based advertisement recommendation method.
The computer program may be partitioned into one or more modules/units that are stored in the memory and executed by the processor to perform the click-through rate estimation model-based advertisement recommendation method.
The memory may be used to store computer programs and/or modules that implement the various functions of the present system by running or executing the computer programs and/or modules stored in the memory and invoking data stored in the memory.
The advertisement recommendation method, the advertisement recommendation system and the storage medium based on the click rate estimation model have the following advantages:
1. the click rate estimation accuracy is high, based on understanding of service data, characteristic design and characteristic combination are manually carried out, and the displayed, direct, association rule and other characteristics are used as the input of the Wide part to the model, so that the memory capacity of the model is improved. 2. The recall rate is high, namely, the method has certain generalization capability. And performing the coding optimization and the network design on different ID characteristics as Deep parts, so that the model can grasp more nonlinear characteristics and information of combined characteristics, and the expression capability is enhanced.
The foregoing description of the preferred embodiment of the present invention is not intended to limit the technical solution of the present invention in any way, and it should be understood that the technical solution can be modified and replaced in several ways without departing from the spirit and principle of the present invention, and these modifications and substitutions are also included in the protection scope of the claims.
Claims (7)
1. An advertisement recommendation method based on a click rate estimation model is characterized by comprising the following steps:
-data collection, cleaning the business data, constructing a data set;
-feature engineering the dataset to generate training samples for use in model training;
-model training using the Wide & Deep algorithm based on the Tensorflow framework;
-adjusting model structural parameters, determining the best super-parameter combination, and performing model optimization;
performing an A/B experiment on the optimization model and the online model, and replacing the old model with the increased click rate;
-updating the feature data to the cloud storage OSS in real time, generating a GRPC interface using Tensorflow Serving, estimating click rate of the candidate advertisement list in batches, taking the advertisement with the highest click rate, and recommending exposure with the click rate > 0.5;
the specific steps of the feature engineering processing of the data set are as follows:
step 1) sample sampling, adding positive and negative sample punishment weights;
step 2) dividing the data in the data set into continuous features and discrete features, and carrying out normalization and missing value processing;
step 3) performing feature cross combination;
the missing value processing specifically comprises the following steps: carrying out feature distribution statistics, analyzing the missing proportion, and discarding the feature if the missing proportion reaches more than 80%; filling by adopting a default value for discrete features and adopting a mean value for continuous features;
and 3) when the feature cross combination is carried out, the user features, advertisements and media features are crossed, including cross combination of user commodity preference, user labels, user history funnel depth, advertisement ID, advertisement material ID, advertisement category and category.
2. The advertisement recommendation method based on the click-through rate estimation model of claim 1, wherein: the data collection objects comprise users, advertisements, request contexts and media; wherein the user characteristic data comprises: user ID, user commodity preference, user tag, user history funnel depth, last 3/7/14/28 day exposure times, click times, browse times, purchase times, click rate; the advertisement feature data includes: advertisement ID, advertisement category, advertisement material ID, advertisement template ID, exposure times, click times and click rate; the request context feature data includes: time, country, city, channel, media, version size, device type, display type; the media characteristic data includes: category, ranking, PV, UV, click rate.
3. The advertisement recommendation method based on the click-through rate estimation model of claim 1, wherein:
the continuous features in the user feature data comprise exposure times, clicking times, browsing times and clicking rates, the continuous features in the advertisement feature data comprise exposure times, clicking times and clicking rates, the continuous features in the media feature data comprise ranking, PV, UV and historical clicking rates, and the continuous features are normalized by adopting a logarithmic function; meanwhile, performing equal-frequency barrel discretization on part of continuous features, including ranking in media feature data, purchasing times and purchasing times of user feature data;
the discrete features include user ID, user merchandise preference, user tag, user history funnel depth, advertisement ID, advertisement category, advertisement material ID, advertisement template ID, time, country, city, channel, media, plate size, device type, display type, and category, and the above discrete features are One-Hot encoded using a Hash scheme.
4. The advertisement recommendation method based on click rate estimation model as set forth in claim 1, wherein the specific steps of model training using the Wide & Deep algorithm are as follows: inputting discrete features and cross combination features in a Wide layer, inputting continuous features in a Deep layer, and taking a historical data set as a training set and a data set of the last 1 day as a test set;
performing model tuning, adding Dropout and L2 regularization to prevent the model from being fitted excessively, and introducing Batch Normalization to accelerate the convergence of the model; compared with different learners, selecting Adm with better effect; meanwhile, tuning parameters of different Learning Rate, batch Size and Learning parameters is tried to be optimized, a training model is carried out, and finally a SavedModel format model file is generated.
5. The advertisement recommendation method based on click through rate estimation model of claim 1, wherein the specific step of generating the GRPC interface using Tensorflow Serving comprises:
step 1) starting a Docker, pulling a Tensorflow Serving Docker mirror image; step 2) generating a prediction interface; step 3) online deployment, configuring a unified domain name Tservice, and generating a final scoring interface service; step 4) online reasoning, constructing request data in a batch mode, calling a prediction interface, and finally realizing that 50 advertisement prediction results can be returned within 10 ms.
6. An advertisement recommendation system based on click rate estimation model, which is used for implementing the advertisement recommendation method based on click rate estimation model as claimed in any one of claims 1 to 5, comprising:
the data set construction module is used for collecting data, cleaning service data and constructing a data set;
the training sample generation module is used for carrying out characteristic engineering processing on the data set to generate a training sample;
the model training module is used for carrying out model training by adopting a Wide & Deep algorithm based on a Tensorflow framework;
the model optimization module is used for adjusting model structure parameters, determining the best super-parameter combination and carrying out model optimization; performing an A/B experiment on the optimization model and the online model, and replacing the old model with the increased click rate;
and the recommending module is used for updating the characteristic data to the cloud storage OSS in real time, generating GRPC interfaces by using Tensorflow Serving, estimating click rate of candidate advertisement lists in batches, and recommending exposure according to the click rate.
7. A computer-readable storage medium storing a computer program, characterized in that: the computer program when executed by a processor performs the steps of the click rate estimation model-based advertisement recommendation method according to any one of claims 1 to 5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110456025.2A CN113222647B (en) | 2021-04-26 | 2021-04-26 | Advertisement recommendation method, system and storage medium based on click rate estimation model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110456025.2A CN113222647B (en) | 2021-04-26 | 2021-04-26 | Advertisement recommendation method, system and storage medium based on click rate estimation model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113222647A CN113222647A (en) | 2021-08-06 |
CN113222647B true CN113222647B (en) | 2023-11-28 |
Family
ID=77089472
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110456025.2A Active CN113222647B (en) | 2021-04-26 | 2021-04-26 | Advertisement recommendation method, system and storage medium based on click rate estimation model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113222647B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113704615A (en) * | 2021-08-30 | 2021-11-26 | 万汇互联(深圳)科技有限公司 | Deep interest network recommendation method based on multiple modes |
CN113837483A (en) * | 2021-09-29 | 2021-12-24 | 深圳市易平方网络科技有限公司 | Advertisement flow pre-estimation processing method and device based on wireless receiving device and terminal |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20110063160A (en) * | 2009-12-04 | 2011-06-10 | 동국대학교 산학협력단 | Apparatus and method for providing advertisement |
CN105590240A (en) * | 2015-12-30 | 2016-05-18 | 合一网络技术(北京)有限公司 | Discrete calculating method of brand advertisement effect optimization |
CN108228824A (en) * | 2017-12-29 | 2018-06-29 | 暴风集团股份有限公司 | Recommendation method, apparatus, electronic equipment, medium and the program of a kind of video |
CN108416625A (en) * | 2018-02-28 | 2018-08-17 | 阿里巴巴集团控股有限公司 | The recommendation method and apparatus of marketing product |
CN108596645A (en) * | 2018-03-13 | 2018-09-28 | 阿里巴巴集团控股有限公司 | A kind of method, apparatus and equipment of information recommendation |
WO2018212710A1 (en) * | 2017-05-19 | 2018-11-22 | National University Of Singapore | Predictive analysis methods and systems |
CN110619540A (en) * | 2019-08-13 | 2019-12-27 | 浙江工业大学 | Click stream estimation method of neural network |
CN110728541A (en) * | 2019-10-11 | 2020-01-24 | 广州市丰申网络科技有限公司 | Information stream media advertisement creative recommendation method and device |
CN110852793A (en) * | 2019-10-28 | 2020-02-28 | 北京深演智能科技股份有限公司 | Document recommendation method and device and electronic equipment |
CN111435507A (en) * | 2019-01-11 | 2020-07-21 | 腾讯科技(北京)有限公司 | Advertisement anti-cheating method and device, electronic equipment and readable storage medium |
CN112434184A (en) * | 2020-12-15 | 2021-03-02 | 四川长虹电器股份有限公司 | Deep interest network sequencing method based on historical movie posters |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190325293A1 (en) * | 2018-04-19 | 2019-10-24 | National University Of Singapore | Tree enhanced embedding model predictive analysis methods and systems |
-
2021
- 2021-04-26 CN CN202110456025.2A patent/CN113222647B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20110063160A (en) * | 2009-12-04 | 2011-06-10 | 동국대학교 산학협력단 | Apparatus and method for providing advertisement |
CN105590240A (en) * | 2015-12-30 | 2016-05-18 | 合一网络技术(北京)有限公司 | Discrete calculating method of brand advertisement effect optimization |
WO2018212710A1 (en) * | 2017-05-19 | 2018-11-22 | National University Of Singapore | Predictive analysis methods and systems |
CN108228824A (en) * | 2017-12-29 | 2018-06-29 | 暴风集团股份有限公司 | Recommendation method, apparatus, electronic equipment, medium and the program of a kind of video |
CN108416625A (en) * | 2018-02-28 | 2018-08-17 | 阿里巴巴集团控股有限公司 | The recommendation method and apparatus of marketing product |
CN108596645A (en) * | 2018-03-13 | 2018-09-28 | 阿里巴巴集团控股有限公司 | A kind of method, apparatus and equipment of information recommendation |
CN111435507A (en) * | 2019-01-11 | 2020-07-21 | 腾讯科技(北京)有限公司 | Advertisement anti-cheating method and device, electronic equipment and readable storage medium |
CN110619540A (en) * | 2019-08-13 | 2019-12-27 | 浙江工业大学 | Click stream estimation method of neural network |
CN110728541A (en) * | 2019-10-11 | 2020-01-24 | 广州市丰申网络科技有限公司 | Information stream media advertisement creative recommendation method and device |
CN110852793A (en) * | 2019-10-28 | 2020-02-28 | 北京深演智能科技股份有限公司 | Document recommendation method and device and electronic equipment |
CN112434184A (en) * | 2020-12-15 | 2021-03-02 | 四川长虹电器股份有限公司 | Deep interest network sequencing method based on historical movie posters |
Non-Patent Citations (1)
Title |
---|
基于宽深度模型的广告点击率预估方法;林启迪;华南理工大学硕士学位论文;1-68 * |
Also Published As
Publication number | Publication date |
---|---|
CN113222647A (en) | 2021-08-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Zuo et al. | Tag-aware recommender systems based on deep neural networks | |
CA2992563C (en) | Method and system for applying probabilistic topic models to content in a tax environment to improve user satisfaction with a question and answer customer support system | |
US11049149B2 (en) | Determination of targeted food recommendation | |
CN113222647B (en) | Advertisement recommendation method, system and storage medium based on click rate estimation model | |
JP2019505936A (en) | Web interface generation and testing using artificial neural networks | |
CN107908740A (en) | Information output method and device | |
CN109376844A (en) | The automatic training method of neural network and device recommended based on cloud platform and model | |
CN107111859A (en) | For the social activity scoring of network element | |
CN110140134A (en) | Recommendation is generated using deep learning model | |
CN102360386A (en) | Intelligent shopping guide system and method of electronic commerce website | |
CN108632311A (en) | Information-pushing method and device | |
Liu et al. | Fractal and multifractal analyses of bipartite networks | |
CN109087162A (en) | Data processing method, system, medium and calculating equipment | |
EP3732592A1 (en) | Intelligent routing services and systems | |
CN112085565A (en) | Deep learning-based information recommendation method, device, equipment and storage medium | |
Faisal et al. | Discovering Indonesian digital workers in online gig economy platforms | |
CN113569129A (en) | Click rate prediction model processing method, content recommendation method, device and equipment | |
CN112116426A (en) | Method and device for pushing article information | |
CN107767152A (en) | Product purchase intention analysis method and server | |
CN111814044A (en) | Recommendation method and device, terminal equipment and storage medium | |
Yue et al. | A parallel and constraint induced approach to modeling user preference from rating data | |
CN111507471B (en) | Model training method, device, equipment and storage medium | |
CN115098698A (en) | Method and device for constructing Schema model in knowledge graph | |
CN112633931A (en) | Click rate prediction method, device, electronic equipment and medium | |
CN109918570A (en) | Content recommendation method, device, storage medium and content processing method |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |