CN107590689B - Advertisement data recommendation method and system - Google Patents
Advertisement data recommendation method and system Download PDFInfo
- Publication number
- CN107590689B CN107590689B CN201710744931.6A CN201710744931A CN107590689B CN 107590689 B CN107590689 B CN 107590689B CN 201710744931 A CN201710744931 A CN 201710744931A CN 107590689 B CN107590689 B CN 107590689B
- Authority
- CN
- China
- Prior art keywords
- advertisement data
- user
- characteristic information
- target
- target advertisement
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Transfer Between Computers (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses an advertisement data recommendation method and system. The method comprises the following steps: the method comprises the steps of receiving a browsing request of a user for target advertisement data, obtaining user characteristic information of the user, inputting the user characteristic information into a pre-established relevancy prediction model, obtaining the relevancy between each advertisement display form of the target advertisement data and the browsing request, and recommending the target advertisement data of the advertisement display form with the relevancy meeting a preset condition to the user.
Description
Technical Field
The invention relates to the technical field of information, in particular to an advertisement data recommendation method and an advertisement data recommendation system.
Background
With the development of network technology, it has become a popular product publicizing means for merchants to publish advertisements on network media for product display. The advertisement data issued on the network media can include various advertisements, such as image-text advertisements, video advertisements, advertisements combining text and video, and the like, and the user can acquire product information by viewing the advertisements displayed on the network media.
Currently, some video advertisements include multiple ad presentations with different ad creatives that produce different promotional effects for different users. In the prior art, after receiving a play instruction of a certain video advertisement, a network media randomly selects an advertisement presentation form from a plurality of advertisement presentation forms of the video advertisement, and then displays the video advertisement containing the selected advertisement presentation form on the network media.
Disclosure of Invention
In view of the above problems, the present invention has been made to provide an advertisement data recommendation method and a corresponding advertisement data recommendation system that overcome or at least partially solve the above problems.
According to an aspect of the present invention, there is provided an advertisement data recommendation method, the method including:
after a browsing request of a user for target advertisement data is received, acquiring user characteristic information of the user;
inputting the user characteristic information into a pre-established relevancy prediction model to obtain the relevancy between each advertisement display form of the target advertisement data and the browsing request;
and recommending the target advertisement data containing the advertisement display form with the relevance meeting the preset condition to the user.
Optionally, the correlation prediction model is obtained by:
extracting a first advertisement data usage log recorded in a first historical period;
searching the target advertisement data which has historical exposure amount larger than exposure amount threshold value and comprises a plurality of advertisement display forms from the first advertisement data usage log, and extracting user characteristic information of at least one user who has historically browsed the target advertisement data;
and training a machine learning relevance algorithm according to the plurality of advertisement display forms of the target advertisement data and the user characteristic information of the at least one user to obtain a relevance prediction model.
Optionally, before the obtaining of the user characteristic information of the user, the method further includes:
judging whether a correlation degree prediction model is established for the target advertisement data;
and if so, executing the step of acquiring the user characteristic information of the user.
If not, calculating the smooth click rate of each advertisement display form of the target advertisement data by using a pre-established smooth click rate statistical model, and recommending the target advertisement data containing the advertisement display forms meeting the preset click rate condition to the user.
Optionally, the calculating, by using a pre-established statistical model of smooth click-through rate, a smooth click-through rate of each advertisement presentation form of the target advertisement data includes:
extracting a second advertisement data usage log recorded in a second history period before the current date;
extracting historical exposure and historical click rate of each advertisement display form of the target advertisement data from the second advertisement data usage log;
inputting the historical exposure and the historical click rate of each advertisement display form into the pre-established smooth click rate statistical model to obtain the smooth click rate of each advertisement display form; the duration of the second history period is greater than the duration of the first history period.
Optionally, the acquiring the user characteristic information of the user includes:
extracting a log of usage of third advertisement data recorded in a third history period;
and searching user characteristic information of the user from the log of the third advertisement data.
Optionally, the recommending, to the user, target advertisement data including an advertisement presentation format that satisfies a preset relevancy condition includes:
and recommending the target advertisement data containing the maximum relevance in the advertisement display form to the user.
Optionally, the user characteristic information includes one or more of user identity information and user browsing information.
According to another aspect of the present invention, there is also provided an advertisement data recommendation system, the system including:
the request receiving module is used for receiving a browsing request of a user for the target advertisement data;
the user characteristic information acquisition module is used for acquiring the user characteristic information of the user;
a relevancy obtaining module, configured to input the user characteristic information into a pre-established relevancy prediction model to obtain relevancy between each advertisement presentation form of the target advertisement data and the browsing request;
and the target advertisement data recommending module is used for recommending the target advertisement data containing the advertisement display form with the correlation degree meeting the preset condition to the user.
Optionally, the system further comprises a model training module, the model training module comprising:
the first log extraction submodule is used for extracting a first advertisement data use log recorded in a first historical time interval;
the target advertisement data searching sub-module is used for searching the target advertisement data which is larger than an exposure threshold value in historical exposure and contains a plurality of advertisement display forms from the first advertisement data usage log;
the user characteristic information extraction submodule is used for extracting the user characteristic information of at least one user browsing the target advertisement data historically;
and the model obtaining submodule is used for training a machine learning relevance algorithm according to the plurality of advertisement display forms of the target advertisement data and the user characteristic information of the at least one user to obtain a relevance prediction model.
Optionally, the system further includes a model determining module and a smooth click rate statistical model using module:
the model judging module is used for judging whether a correlation degree prediction model is established for the target advertisement data before the user characteristic information of the user is obtained;
the user characteristic information acquisition module is specifically used for acquiring the user characteristic information of the user when the relevance prediction model corresponding to the target advertisement data is established;
and the smooth click rate statistical model using module is used for calculating the smooth click rate of each advertisement display form of the target advertisement data by using a pre-established smooth click rate statistical model when the relevance prediction model corresponding to the target advertisement data is not established, and recommending the target advertisement data containing the advertisement display forms meeting the preset click rate condition to the user.
Optionally, the module for using a statistical model of smooth click rate comprises:
the second log extraction submodule is used for extracting a second advertisement data use log recorded in a second historical time period before the current date;
a historical exposure and historical click rate extraction submodule, configured to extract historical exposure and historical click rate of each advertisement presentation form of the target advertisement data from the second advertisement data usage log;
the smooth click rate obtaining submodule is used for inputting the historical exposure and the historical click rate of each advertisement display form into the pre-established smooth click rate statistical model to obtain the smooth click rate of each advertisement display form; the duration of the second history period is greater than the duration of the first history period.
Optionally, the user feature information obtaining module includes:
the third log extraction submodule is used for extracting a log of the use of third advertisement data recorded in a third history time period;
and the user characteristic information searching submodule is used for searching the user characteristic information of the user from the log used by the third advertisement data.
Optionally, the targeted advertisement data recommending module is specifically configured to recommend targeted advertisement data in an advertisement presentation form including the maximum relevance to the user.
Optionally, the user characteristic information includes one or more of user identity information and user browsing information.
According to the embodiment of the invention, after a browsing request sent by a user to the target advertisement data is received, the relevance between each advertisement display form of the target advertisement data and the browsing request is obtained by utilizing a pre-established relevance prediction model according to the user characteristic information and the historical browsing information of the user, the advertisement display form with the relevance meeting the preset condition is extracted, and the target advertisement data containing the advertisement display form is recommended to the user.
When a relevance prediction model corresponding to target advertisement data is not established, the smooth click rate of each advertisement display form of the target advertisement data is calculated by using a pre-established smooth click rate statistical model, the advertisement display form with the smooth click rate meeting preset conditions is determined, and the target advertisement data containing the advertisement display form is recommended to a user, so that the multi-model advertisement data recommendation method is provided.
The foregoing description is only an overview of the technical solutions of the present invention, and the embodiments of the present invention are described below in order to make the technical means of the present invention more clearly understood and to make the above and other objects, features, and advantages of the present invention more clearly understandable.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
fig. 1 is a flowchart illustrating an advertisement data recommendation method according to embodiment 1 of the present invention;
FIG. 2 is a flowchart of an advertisement data recommendation method according to embodiment 2 of the present invention;
fig. 3 is a block diagram showing a configuration of an advertisement data recommendation system according to embodiment 1 of the present invention;
fig. 4 is a block diagram showing a configuration of an advertisement data recommendation system according to embodiment 2 of the present invention.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
Referring to fig. 1, a flowchart of an advertisement data recommendation method according to embodiment 1 of the present invention is shown, where the method specifically may include:
The network media is provided with an advertisement putting platform which manages advertisement positions provided by the network media. Generally, in an existing bidding advertisement delivery mode, an advertisement delivery platform obtains traffic of advertisements to be delivered of media and sends a request for delivering the advertisements to an advertisement trading platform, the request usually carries the traffic of the advertisements to be delivered of the media, the advertisement trading platform obtains bidding advertisements with appropriate traffic from a large number of advertisements from advertisement demanders providing advertisement data, sorts advertisement data fed back by the advertisement demanders, and delivers winning bidding advertisement data to the media providing advertisement delivery traffic for playing.
In the advertisement data played by the network media, part of the advertisement data comprises a plurality of advertisement display forms, and different advertisement display forms have different advertisement creatives. The advertisement trading platform can simultaneously acquire a plurality of advertisement display forms of advertisement data from the advertisement demand side and return the plurality of advertisement display forms to the network media together.
In order to improve the interest of the user in the played advertisement data and improve the user experience, after a browsing request of the user for the target advertisement data to be viewed is received, according to the relevant information of the user, the advertisement display form which is relatively matched with the user is extracted from the plurality of advertisement display forms of the target advertisement data, and the target advertisement data comprising the matched advertisement display form is displayed, so that the target advertisement data is recommended to the user for viewing.
The target advertisement data can be various, such as video advertisement data, images and the like.
The step of receiving the browsing request of the target advertisement data to be viewed by the user may include multiple steps based on factors such as the type and the playing time of the target advertisement data, for example, the first step may be receiving an indication of inter-cutting the video advertisement data in the currently played video or detecting that the playing of the first segment of video is finished when the target advertisement data is the video advertisement data and the video advertisement data is inter-cut in other videos; secondly, when the target advertisement data is the video advertisement data and needs to be clicked and played, the above steps may be receiving a click operation of the user on the target advertisement data.
In this step, after receiving a browsing request of the user for the target advertisement data, the user characteristic information of the user is acquired, so that the advertisement display form of the target advertisement data is selected and recommended according to the relevant information of the user in the following. The user characteristic information may include a plurality of types, such as user identity information, user browsing information, and the like, where the user identity information may be an internet protocol address, i.e., an ip address, of a user, a login account of the user, and the like, and the user browsing information may be data content, a data address, and the like of previous browsing data corresponding to the target advertisement data. When a user browses the network media, the network media can record and store information such as user identity information and user browsing information of the user, for example, the information is stored in a log form, and after receiving a browsing request of the user for target advertisement data, the media can be extracted from a specified storage position to obtain related information of the user.
And 102, inputting the user characteristic information into a pre-established relevancy prediction model to obtain the relevancy between each advertisement display form of the target advertisement data and the browsing request.
In the embodiment of the invention, different advertisement data correspond to different relevancy prediction models, and the relevancy prediction model is established in advance aiming at the target advertisement data and comprises parameters such as user characteristic information, advertisement display forms and the like. The media may record usage information for the advertisement data and generate a usage log. When the relevance prediction model is established, required information can be extracted from the use log of the advertisement data, and the extracted required information is used for training a relevance algorithm to obtain the relevance prediction model.
The embodiment of the invention inputs the acquired user characteristic information into a pre-established relevance prediction model to obtain the relevance between each advertisement display form of the target advertisement data and the browsing request. The greater the degree of correlation, the more interested the user is in the corresponding advertisement presentation form, and the greater the recommendation value of the corresponding advertisement presentation form.
And 103, recommending the target advertisement data containing the advertisement display form with the relevance meeting the preset condition to the user.
The embodiment of the invention presets the preset condition of the relevancy, determines the advertisement showing form with the relevancy meeting the preset condition after calculating the relevancy of each advertisement showing form of the target advertisement data, and further recommends the target advertisement data containing the advertisement showing form meeting the preset condition to the user for showing on the media. The preset conditions may be various, such as the correlation is maximum, the correlation is greater than a threshold, and the like, and may be set according to the actual conditions.
Since the relevance of the target advertisement data recommended to the user and the browsing request of the user is large, the user is easy to generate interest in the target advertisement data, thereby improving the user experience. For the advertisement data needing to be clicked and played, the video image displayed before the click is the image which is interested by the user, so the method of the embodiment of the invention improves the click rate of the advertisement data and reduces the ineffective exposure with low click rate of the advertisement data.
According to the embodiment of the invention, after a browsing request sent by a user to the target advertisement data is received, the relevance between each advertisement display form of the target advertisement data and the browsing request is obtained by utilizing a pre-established relevance prediction model according to the user characteristic information and the historical browsing information of the user, the advertisement display form with the relevance meeting the preset condition is extracted, and the target advertisement data containing the advertisement display form is recommended to the user.
Referring to fig. 2, a flowchart of an advertisement data recommendation method according to embodiment 2 of the present invention is shown, where the method specifically may include:
In the embodiment of the invention, the target advertisement data is the advertisement data to be browsed by the user. The targeted advertising data may be video data, picture data, etc., including a plurality of advertising presentation formats, with different advertising creatives for different advertising presentation formats. The media receives the targeted advertising data simultaneously in a plurality of advertising presentation formats of the targeted advertising data.
The target advertisement data may be video data that needs to be played only by a click operation in the same page, may be video data inserted in other videos, and the like, and the step of receiving a browsing request of the target advertisement data to be viewed by a user may include multiple steps, for example, receiving a click operation of the user on the target advertisement data, or receiving an indication of inserting video advertisement data in a currently played video, or detecting that the playing of the first segment of video is finished, based on factors such as the type of the target advertisement data, the playing opportunity, and the like.
After receiving a browsing request of a user for target advertisement data, the embodiment of the invention needs to judge whether a correlation degree prediction model is established for the target advertisement data, if so, the embodiment of the invention can execute the subsequent steps, and according to the user characteristic information of the user, the correlation degree between each advertisement display form of the target advertisement data and the browsing request is calculated by using the correlation degree prediction model.
The relevancy prediction model corresponding to the target advertisement data is established in advance, different relevancy prediction models correspond to different advertisement data, and the relevancy prediction model can be obtained in the following mode:
first, a first advertisement data usage log recorded in a first history period is extracted.
The media records the using condition of the advertisement data in the historical time period to form a first advertisement data using log, and the advertisement data using log can comprise various information, such as exposure times, click times and the like. The duration of the first history period may be set as desired, such as a seven day period.
Secondly, searching target advertisement data which are larger than the exposure threshold value in the historical exposure and comprise a plurality of advertisement display forms from the first advertisement data usage log, and extracting user characteristic information of at least one user browsing the target advertisement data in the historical mode.
For advertisement data for which an advertisement fee is charged in accordance with the number of times of advertisement display, the number of times of display is increased by one exposure. The method comprises the steps of searching one or more advertisement data containing a plurality of advertisement display forms from a first advertisement data use log, judging whether the historical exposure amount of each advertisement data in a first historical time period is larger than an exposure amount threshold value, if so, indicating that the display times of the advertisement data in the historical time period are more, determining to establish a correlation prediction model for the advertisement data, and if not, calculating the correlation by using a pre-established smooth click rate (smoothctr) statistical model.
In the embodiment of the invention, the target advertisement data is advertisement data which contains a plurality of advertisement display forms and has historical exposure greater than an exposure threshold. After the target advertisement data meeting the conditions are determined, user characteristic information, such as user identity information, user browsing information and the like, of at least one user browsing the target advertisement data in a first historical time period is extracted from the first advertisement data use log.
And finally, training a machine learning relevance algorithm according to a plurality of advertisement display forms of the target advertisement data and user characteristic information of at least one user to obtain a relevance prediction model.
After target advertisement data meeting the conditions are found out and user characteristic information of at least one user browsing the target advertisement data historically is extracted, a relevance prediction model is obtained by training a machine learning relevance algorithm according to a plurality of advertisement display forms of the target advertisement data and the user characteristic information of the at least one user. The relevancy prediction model can comprise parameter information such as user characteristic information, a plurality of advertisement display forms and the like. The machine learning relevance algorithm may include a variety of algorithms, such as a factorization machine algorithm, i.e., an FM (factorization machine) algorithm.
The embodiment of the invention uses the machine learning method to better recommend the advertisement data according to the user characteristic information and the advertisement characteristics of the user, improves the matching degree of the advertisement display form of the recommended advertisement data and the user, and improves the user experience.
And if the relevance prediction model corresponding to the target advertisement data is judged to be established, acquiring the user characteristic information of the user. The user characteristic information may include a plurality of types, such as one or more of user identity information and user browsing information, where the user identity information may be an internet protocol address, i.e., an ip address, of the user, a login account of the user, and the like, and the user browsing information may be data content, a data address, and the like of previous browsing data relative to the targeted advertisement data.
The user characteristic information of the user may be obtained in various ways, such as extracting a log of usage of the third advertisement data recorded in the third history period, and finding the user characteristic information of the user from the log of usage of the third advertisement data. The duration of the third history period may be set according to practice. The log of the usage of the third advertisement data may be a usage log of advertisement data records for all historical usages, or may be a usage log of advertisement data records used by a certain user in a historical period, and may be set according to actual situations.
And 204, inputting the user characteristic information into a pre-established relevancy prediction model to obtain the relevancy between each advertisement display form of the target advertisement data and the browsing request.
The embodiment of the invention inputs the acquired user characteristic information into the relevance prediction model, and obtains the relevance between each advertisement display form of the target advertisement data and the browsing request through model calculation.
And recommending the target advertisement data containing the advertisement display forms with the relevance meeting the preset conditions to the user after calculating the relevance between each advertisement display form of the target advertisement data and the browsing request. The preset conditions may be various, for example, the target advertisement data including the advertisement presentation format with the maximum relevance is recommended to the user, and the target advertisement data including the advertisement presentation format with the relevance greater than the preset threshold is recommended to the user, and may be set according to the actual situation. Accordingly, the step of recommending the target advertisement data containing the advertisement presentation format satisfying the preset relevancy condition to the user may include: and recommending the target advertisement data containing the maximum relevance in the advertisement display form to the user.
And step 206, if the relevance prediction model corresponding to the target advertisement data is not established, calculating the smooth click rate of each advertisement display form of the target advertisement data by using a pre-established smooth click rate statistical model, and recommending the target advertisement data containing the advertisement display forms meeting the preset click rate condition to the user.
The embodiment of the invention establishes a smooth click rate calculation model in advance, if the relevant correlation prediction model is judged not to be established aiming at the target advertisement data, the smooth click rate of each advertisement display form of the target advertisement data is calculated by using the pre-established smooth click rate statistical model, and the target advertisement data containing the advertisement display forms meeting the preset click rate condition is recommended to the user. The preset click rate conditions can be various, such as the maximum click rate, the click rate greater than a preset threshold value, and the like.
The step of calculating the smooth click rate of each advertisement presentation form of the target advertisement data using the pre-established smooth click rate statistical model may include:
first, a second advertisement data usage log recorded in a second history period before the current date is extracted.
The media records the using condition of the advertisement data in a second historical time period before the current date to form a second advertisement data using log, and the second advertisement data using log can comprise various information, such as exposure times, click times and the like. The duration of the second history period can be set according to needs, such as fourteen days, specifically, the number of the second history period is 20 today, and the second history period is 6 days to 19 days.
In the embodiment of the invention, the duration of the second historical period used by the smooth click rate statistical model is longer than the duration of the first historical period used by the correlation prediction model. When exposure of certain advertisement data is sufficient in short historical time, the relevance calculation is carried out by using the relevance prediction model, and when exposure of certain advertisement data is insufficient in short historical time, the relevance calculation is carried out by using the smooth click rate statistical model with long historical time, so that the problems of few browsing times and cold start of an advertisement display form are solved.
And if one advertisement display form of the target advertisement data is a new display form, calculating the relevancy of the target advertisement data by using the smooth click rate statistical model, and taking the relevancy of the target advertisement data as the relevancy of the new advertisement display form. In the formula of the smooth click rate statistical model described below, each parameter is a parameter of the target advertisement data.
And if the target advertisement data is new advertisement data, randomly assigning values to a plurality of advertisement display forms of the new target advertisement data, such as 0-1, sequencing the advertisement display forms based on the assignments, and preferentially recommending the advertisement display forms in the front sequence.
Next, from the second advertisement data usage log, the historical exposure amount and the historical click rate of each advertisement presentation form of the target advertisement data are extracted.
The smooth click rate statistical model includes various parameters such as historical exposure and historical click rate, so after the second advertisement data usage log is extracted, the historical exposure and the historical click rate of each advertisement presentation form of the target advertisement data are extracted from the first advertisement data usage log.
And finally, inputting the historical exposure and the historical click rate of each advertisement display form into a pre-established smooth click rate statistical model to obtain the smooth click rate of each advertisement display form.
After the historical exposure and the historical click rate of each advertisement display form of the target advertisement data are extracted, the historical exposure and the historical click rate are input into a pre-established smooth click rate statistical model to obtain the smooth click rate of each advertisement display form.
The model formula of the statistical model of the smooth click rate used in the embodiment of the invention can be as follows:
wherein L is the smooth click rate of the advertisement display form, C is the historical click rate of the advertisement display form in the second historical period, C is the current click rate of the advertisement display form in the current date, D is the historical exposure of the advertisement display form in the second historical period, D is the current exposure of the advertisement display form in the current date, and alpha is a parameter.
As can be seen from the above formula, the above-mentioned historical exposure amount extracted from the second advertisement data usage log includes the historical exposure amount in the second historical period and the current exposure amount in the current date, and the extracted historical click rate includes the historical click rate in the second historical period and the current exposure amount in the current date.
As can be seen from the above description, the embodiment of the present invention provides a multi-model advertisement data recommendation method, and for advertisement data with sufficient exposure, the correlation degree of each advertisement presentation form of the advertisement data may be calculated by using a correlation degree prediction model, and for advertisement data with insufficient exposure, the correlation degree of each advertisement presentation form of the advertisement data may be calculated by using a smooth click rate statistical model, so as to implement recommendation of the advertisement data.
According to the embodiment of the invention, after a browsing request sent by a user to the target advertisement data is received, the relevance between each advertisement display form of the target advertisement data and the browsing request is obtained by utilizing a pre-established relevance prediction model according to the user characteristic information and the historical browsing information of the user, the advertisement display form with the relevance meeting the preset condition is extracted, and the target advertisement data containing the advertisement display form is recommended to the user.
When a relevance prediction model corresponding to target advertisement data is not established, the smooth click rate of each advertisement display form of the target advertisement data is calculated by using a pre-established smooth click rate statistical model, the advertisement display form with the smooth click rate meeting preset conditions is determined, and the target advertisement data containing the advertisement display form is recommended to a user, so that the multi-model advertisement data recommendation method is provided.
Based on the description of the method embodiment, the invention also provides a corresponding advertisement data recommendation system embodiment to implement the content described in the method embodiment.
Referring to fig. 3, a block diagram of an advertisement data recommendation system according to embodiment 1 of the present invention is shown, where the advertisement data recommendation system may include:
a request receiving module 301, configured to receive a browsing request of a user for the target advertisement data.
A user characteristic information obtaining module 302, configured to obtain user characteristic information of the user.
A relevancy obtaining module 303, configured to input the user characteristic information into a pre-established relevancy prediction model, so as to obtain relevancy between each advertisement presentation form of the target advertisement data and the browsing request.
And the targeted advertisement data recommending module 304 is configured to recommend targeted advertisement data containing an advertisement presentation format with a relevance meeting a preset condition to the user.
According to the embodiment of the invention, after a browsing request sent by a user to the target advertisement data is received, the relevance between each advertisement display form of the target advertisement data and the browsing request is obtained by utilizing a pre-established relevance prediction model according to the user characteristic information and the historical browsing information of the user, the advertisement display form with the relevance meeting the preset condition is extracted, and the target advertisement data containing the advertisement display form is recommended to the user.
Referring to fig. 4, a block diagram of an advertisement data recommendation system according to embodiment 2 of the present invention is shown, where the advertisement data recommendation system may include:
a request receiving module 401, configured to receive a browsing request of a user for the target advertisement data.
A model determining module 402, configured to determine whether a relevance prediction model is established for the target advertisement data.
A user characteristic information obtaining module 403, configured to obtain user characteristic information of the user when a relevance prediction model corresponding to the target advertisement data is established.
A relevancy obtaining module 404, configured to input the user characteristic information into a pre-established relevancy prediction model, so as to obtain relevancy between each advertisement presentation form of the target advertisement data and the browsing request.
And a target advertisement data recommending module 405, configured to recommend target advertisement data in an advertisement presentation form with a relevance meeting a preset condition to the user.
And a smooth click rate statistical model using module 406, configured to calculate a smooth click rate of each advertisement presentation form of the target advertisement data by using a pre-established smooth click rate statistical model when a correlation prediction model corresponding to the target advertisement data is not established, and recommend the target advertisement data including the advertisement presentation form satisfying a preset click rate condition to the user.
In the embodiment of the present invention, preferably, the system further includes a model training module, and the model training module includes:
the first log extraction submodule is used for extracting a first advertisement data use log recorded in a first historical time interval;
the target advertisement data searching sub-module is used for searching the target advertisement data which is larger than an exposure threshold value in historical exposure and contains a plurality of advertisement display forms from the first advertisement data usage log;
the user characteristic information extraction submodule is used for extracting the user characteristic information of at least one user browsing the target advertisement data historically;
and the model obtaining submodule is used for training a machine learning relevance algorithm according to the plurality of advertisement display forms of the target advertisement data and the user characteristic information of the at least one user to obtain a relevance prediction model.
In the embodiment of the present invention, preferably, the module 406 for using a smooth click rate statistical model includes:
the second log extraction submodule is used for extracting a second advertisement data use log recorded in a second historical time period before the current date;
a historical exposure and historical click rate extraction submodule, configured to extract historical exposure and historical click rate of each advertisement presentation form of the target advertisement data from the second advertisement data usage log;
the smooth click rate obtaining submodule is used for inputting the historical exposure and the historical click rate of each advertisement display form into the pre-established smooth click rate statistical model to obtain the smooth click rate of each advertisement display form; the duration of the second history period is greater than the duration of the first history period.
In the embodiment of the present invention, preferably, the user characteristic information obtaining module 403 includes:
the third log extraction submodule is used for extracting a log of the use of third advertisement data recorded in a third history time period;
and the user characteristic information searching submodule is used for searching the user characteristic information of the user from the log used by the third advertisement data.
In the embodiment of the present invention, preferably, the targeted advertisement data recommending module 405 is specifically configured to recommend targeted advertisement data in an advertisement presentation format including the maximum relevance to the user.
In the embodiment of the present invention, preferably, the user characteristic information includes one or more of user identity information and user browsing information.
According to the embodiment of the invention, after a browsing request sent by a user to the target advertisement data is received, the relevance between each advertisement display form of the target advertisement data and the browsing request is obtained by utilizing a pre-established relevance prediction model according to the user characteristic information and the historical browsing information of the user, the advertisement display form with the relevance meeting the preset condition is extracted, and the target advertisement data containing the advertisement display form is recommended to the user.
When a relevance prediction model corresponding to target advertisement data is not established, the smooth click rate of each advertisement display form of the target advertisement data is calculated by using a pre-established smooth click rate statistical model, the advertisement display form with the smooth click rate meeting preset conditions is determined, and the target advertisement data containing the advertisement display form is recommended to a user, so that the multi-model advertisement data recommendation method is provided.
For the embodiment of the encoding mode determining apparatus, since it is basically similar to the embodiment of the method, the description is relatively simple, and the relevant points can be referred to the partial description of the embodiment of the method shown in fig. 1-2.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
As is readily imaginable to the person skilled in the art: any combination of the above embodiments is possible, and thus any combination between the above embodiments is an embodiment of the present invention, but the present disclosure is not necessarily detailed herein for reasons of space.
The algorithms and displays presented herein are not inherently related to any particular computer, virtual machine, or other apparatus. Various general purpose systems may also be used with the teachings herein. The required structure for constructing such a system will be apparent from the description above. Moreover, the present invention is not directed to any particular programming language. It is appreciated that a variety of programming languages may be used to implement the teachings of the present invention as described herein, and any descriptions of specific languages are provided above to disclose the best mode of the invention.
In the description provided herein, numerous specific details are set forth. It is understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be interpreted as reflecting an intention that: that the invention as claimed requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.
Those skilled in the art will appreciate that the modules in the device in an embodiment may be adaptively changed and disposed in one or more devices different from the embodiment. The modules or units or components of the embodiments may be combined into one module or unit or component, and furthermore they may be divided into a plurality of sub-modules or sub-units or sub-components. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or elements of any method or apparatus so disclosed, may be combined in any combination, except combinations where at least some of such features and/or processes or elements are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.
Furthermore, those skilled in the art will appreciate that while some embodiments described herein include some features included in other embodiments, rather than other features, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the following claims, any of the claimed embodiments may be used in any combination.
The various component embodiments of the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that a microprocessor or Digital Signal Processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components of an advertising data recommendation system according to embodiments of the present invention. The present invention may also be embodied as apparatus or device programs (e.g., computer programs and computer program products) for performing a portion or all of the methods described herein. Such programs implementing the present invention may be stored on computer-readable media or may be in the form of one or more signals. Such a signal may be downloaded from an internet website or provided on a carrier signal or in any other form.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The usage of the words first, second and third, etcetera do not indicate any ordering. These words may be interpreted as names.
Claims (12)
1. An advertisement data recommendation method, characterized in that the method comprises:
after a browsing request of a user for target advertisement data is received, acquiring user characteristic information of the user;
inputting the user characteristic information into a pre-established relevancy prediction model to obtain the relevancy between each advertisement display form of the target advertisement data and the browsing request, wherein the target advertisement data corresponds to the relevancy prediction model; different advertisement data correspond to different relevancy prediction models, the relevancy prediction models are established aiming at the target advertisement data in advance, and the relevancy prediction models comprise different advertisement display forms of the target advertisement data;
recommending target advertisement data containing an advertisement display form with the correlation degree meeting a preset condition to the user;
the correlation prediction model is obtained by the following method:
extracting a first advertisement data usage log recorded in a first historical period;
searching the target advertisement data which has historical exposure amount larger than exposure amount threshold value and comprises a plurality of advertisement display forms from the first advertisement data usage log, and extracting user characteristic information of at least one user who has historically browsed the target advertisement data;
and training a machine learning relevance algorithm according to the plurality of advertisement display forms of the target advertisement data and the user characteristic information of the at least one user to obtain a relevance prediction model.
2. The method of claim 1, wherein prior to said obtaining user characteristic information of said user, said method further comprises:
judging whether a correlation degree prediction model is established for the target advertisement data;
if yes, executing the step of acquiring the user characteristic information of the user;
if not, calculating the smooth click rate of each advertisement display form of the target advertisement data by using a pre-established smooth click rate statistical model, and recommending the target advertisement data containing the advertisement display forms meeting the preset click rate condition to the user.
3. The method of claim 2, wherein calculating the smooth click-through rate for each advertisement presentation form of the targeted advertisement data using a pre-established smooth click-through rate statistical model comprises:
extracting a second advertisement data usage log recorded in a second history period before the current date;
extracting historical exposure and historical click rate of each advertisement display form of the target advertisement data from the second advertisement data usage log;
inputting the historical exposure and the historical click rate of each advertisement display form into the pre-established smooth click rate statistical model to obtain the smooth click rate of each advertisement display form; the duration of the second history period is greater than the duration of the first history period.
4. The method of claim 1, wherein the obtaining user characteristic information of the user comprises:
extracting a log of usage of third advertisement data recorded in a third history period;
and searching user characteristic information of the user from the log of the third advertisement data.
5. The method of claim 1, wherein recommending targeted advertisement data containing advertisement presentations satisfying a preset relevancy condition to the user comprises:
and recommending the target advertisement data containing the maximum relevance in the advertisement display form to the user.
6. The method of claim 1, wherein the user characteristic information comprises one or more of user identity information and user browsing information.
7. An advertisement data recommendation system, characterized in that the system comprises:
the request receiving module is used for receiving a browsing request of a user for the target advertisement data;
the user characteristic information acquisition module is used for acquiring the user characteristic information of the user;
a relevancy obtaining module, configured to input the user feature information into a pre-established relevancy prediction model to obtain relevancy between each advertisement presentation form of the target advertisement data and the browsing request, where the target advertisement data corresponds to the relevancy prediction model; different advertisement data correspond to different relevancy prediction models, the relevancy prediction models are established aiming at the target advertisement data in advance, and the relevancy prediction models comprise different advertisement display forms of the target advertisement data;
the target advertisement data recommending module is used for recommending target advertisement data containing an advertisement display form with the correlation degree meeting a preset condition to the user;
the system further includes a model training module, the model training module including:
the first log extraction submodule is used for extracting a first advertisement data use log recorded in a first historical time interval;
the target advertisement data searching sub-module is used for searching the target advertisement data which is larger than an exposure threshold value in historical exposure and contains a plurality of advertisement display forms from the first advertisement data usage log;
the user characteristic information extraction submodule is used for extracting the user characteristic information of at least one user browsing the target advertisement data historically;
and the model obtaining submodule is used for training a machine learning relevance algorithm according to the plurality of advertisement display forms of the target advertisement data and the user characteristic information of the at least one user to obtain a relevance prediction model.
8. The system of claim 7, further comprising a model determination module and a smooth click rate statistics model usage module:
the model judging module is used for judging whether a correlation degree prediction model is established for the target advertisement data before the user characteristic information of the user is obtained;
the user characteristic information acquisition module is specifically used for acquiring the user characteristic information of the user when the relevance prediction model corresponding to the target advertisement data is established;
and the smooth click rate statistical model using module is used for calculating the smooth click rate of each advertisement display form of the target advertisement data by using a pre-established smooth click rate statistical model when the relevance prediction model corresponding to the target advertisement data is not established, and recommending the target advertisement data containing the advertisement display forms meeting the preset click rate condition to the user.
9. The system of claim 8, wherein the smooth click rate statistical model usage module comprises:
the second log extraction submodule is used for extracting a second advertisement data use log recorded in a second historical time period before the current date;
a historical exposure and historical click rate extraction submodule, configured to extract historical exposure and historical click rate of each advertisement presentation form of the target advertisement data from the second advertisement data usage log;
the smooth click rate obtaining submodule is used for inputting the historical exposure and the historical click rate of each advertisement display form into the pre-established smooth click rate statistical model to obtain the smooth click rate of each advertisement display form; the duration of the second history period is greater than the duration of the first history period.
10. The system of claim 7, wherein the user characteristic information obtaining module comprises:
the third log extraction submodule is used for extracting a log of the use of third advertisement data recorded in a third history time period;
and the user characteristic information searching submodule is used for searching the user characteristic information of the user from the log used by the third advertisement data.
11. The system of claim 7, wherein:
the target advertisement data recommending module is specifically used for recommending the target advertisement data containing the advertisement display form with the maximum relevance degree to the user.
12. The system of claim 7, wherein the user characteristic information comprises one or more of user identity information and user browsing information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710744931.6A CN107590689B (en) | 2017-08-25 | 2017-08-25 | Advertisement data recommendation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710744931.6A CN107590689B (en) | 2017-08-25 | 2017-08-25 | Advertisement data recommendation method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107590689A CN107590689A (en) | 2018-01-16 |
| CN107590689B true CN107590689B (en) | 2021-01-08 |
Family
ID=61043026
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710744931.6A Active CN107590689B (en) | 2017-08-25 | 2017-08-25 | Advertisement data recommendation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107590689B (en) |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108416610B (en) * | 2018-01-19 | 2022-03-01 | 创新先进技术有限公司 | User history feedback information forming method and advertisement putting frequency control method |
| CN113298510B (en) * | 2018-07-10 | 2022-06-17 | 马上消费金融股份有限公司 | Deduction instruction initiating method and device |
| CN110796465B (en) * | 2018-08-03 | 2023-08-01 | 阿里巴巴集团控股有限公司 | Resource display equipment determining method, advertisement display screen determining method and device |
| CN111626805B (en) * | 2019-02-28 | 2023-11-03 | 北京沃东天骏信息技术有限公司 | Information display method and device |
| CN110059254A (en) * | 2019-04-24 | 2019-07-26 | 秒针信息技术有限公司 | A kind of message push method and device |
| CN110381377B (en) * | 2019-07-23 | 2021-07-27 | 秒针信息技术有限公司 | Detection method and device for information playing, electronic equipment and storage medium |
| CN110555736B (en) * | 2019-09-05 | 2022-06-03 | 广东聚宝屋传媒有限公司 | Control method and device for advertising based on published demand information |
| CN110992096B (en) * | 2019-12-03 | 2023-08-29 | 秒针信息技术有限公司 | Prediction model training method and device and media identification prediction method and device |
| CN111144937B (en) * | 2019-12-20 | 2024-05-14 | 北京达佳互联信息技术有限公司 | Advertisement material determining method, device, equipment and storage medium |
| CN112435065B (en) * | 2020-11-27 | 2024-03-01 | 杭州推啊网络科技有限公司 | Mobile internet non-standard position interactive advertisement putting method and system |
| CN112767024B (en) * | 2021-01-15 | 2024-02-20 | 北京搜狗科技发展有限公司 | Advertisement creative generation method and related device |
| CN112581195B (en) * | 2021-02-25 | 2021-05-28 | 武汉卓尔数字传媒科技有限公司 | Advertisement pushing method and device and electronic equipment |
| CN113129070B (en) * | 2021-04-28 | 2023-09-01 | 北京奇艺世纪科技有限公司 | Message delivery method, device, server and storage medium |
| CN113487368A (en) * | 2021-07-29 | 2021-10-08 | 宿迁市盛邦信息科技有限公司 | Scientific and technological information push service system based on big data |
| CN114820060A (en) * | 2022-04-26 | 2022-07-29 | 北京有竹居网络技术有限公司 | Advertisement recommendation method, device, equipment and medium |
| CN117495459B (en) * | 2024-01-02 | 2024-03-19 | 蓝色火焰科技成都有限公司 | Man-machine interaction advertisement method, device, equipment and storage medium based on big data |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102567902A (en) * | 2012-02-10 | 2012-07-11 | 亿赞普(北京)科技有限公司 | Network advertisement dynamic release method and system thereof |
| CN104008109B (en) * | 2013-02-26 | 2017-11-14 | 南京邮电大学 | Web information Push Service system based on user interest |
| CN105812844B (en) * | 2014-12-29 | 2019-02-26 | 深圳市Tcl高新技术开发有限公司 | A kind of the user advertising method for pushing and system of TV |
| CN105160548A (en) * | 2015-08-20 | 2015-12-16 | 北京奇虎科技有限公司 | Method and apparatus for predicting advertisement click-through rate |
| CN105139237A (en) * | 2015-09-25 | 2015-12-09 | 百度在线网络技术（北京）有限公司 | Information push method and apparatus |
| CN106997549A (en) * | 2017-02-14 | 2017-08-01 | 火烈鸟网络(广州)股份有限公司 | The method for pushing and system of a kind of advertising message |
-
2017
- 2017-08-25 CN CN201710744931.6A patent/CN107590689B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN107590689A (en) | 2018-01-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107590689B (en) | Advertisement data recommendation method and system | |
| CN110570232A (en) | Internet advertisement putting method, device, server and storage medium | |
| JP5905449B2 (en) | System and method for directing content to users of a social networking engine | |
| CN110278466B (en) | Short video advertisement putting method, device and equipment | |
| US20100153197A1 (en) | Method and apparatus for providing advertising service | |
| US20100034466A1 (en) | Object Identification in Images | |
| CN106202393B (en) | Media information pushing method and device | |
| KR20190075141A (en) | Multi-dimensional advertisement bidding | |
| CN108076353A (en) | Business object recommends method, apparatus, storage medium and electronic equipment | |
| CN111885401B (en) | Method and device for providing data object and method and device for releasing data object | |
| US20200175543A1 (en) | Methodology for ensuring viewability of advertisements in a flip-based environment | |
| US9842354B2 (en) | Simulator for a real-time bidding system | |
| CN109345303B (en) | Rich media advertisement putting method and device | |
| US20110231261A1 (en) | Voice customization for voice-enabled text advertisements | |
| CN110264268B (en) | Advertisement putting device, method, equipment and storage medium thereof | |
| US20110231252A1 (en) | Voice-enabled text advertisements | |
| JP2011243125A (en) | Advertisement distribution system | |
| WO2021038800A1 (en) | Advertisement viewing information output method, advertisement viewing information output program, and information processing device | |
| JP2006237893A (en) | Advertisement charge setting system and advertisement charge setting program | |
| CN111582899A (en) | Method, device and system for quantizing display link effect and storage medium | |
| WO2016125166A1 (en) | Systems and methods for analyzing video and making recommendations | |
| US20200218740A1 (en) | Data prioritization through relationship analysis mapping | |
| CN113793181B (en) | Bid method, system and device for advertisement space | |
| JP2002298015A (en) | Advertisement evaluation system, advertisement evaluation method, and advertisement evaluation program | |
| US20180040031A1 (en) | Facilitating detection of bid data misuse |
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 |