WO2024087796A1 - 一种视频广告投放方法、装置、计算机设备及存储介质 - Google Patents

一种视频广告投放方法、装置、计算机设备及存储介质 Download PDF

Info

Publication number
WO2024087796A1
WO2024087796A1 PCT/CN2023/111672 CN2023111672W WO2024087796A1 WO 2024087796 A1 WO2024087796 A1 WO 2024087796A1 CN 2023111672 W CN2023111672 W CN 2023111672W WO 2024087796 A1 WO2024087796 A1 WO 2024087796A1
Authority
WO
WIPO (PCT)
Prior art keywords
video
video advertisement
parameters
custom
advertisement
Prior art date
Application number
PCT/CN2023/111672
Other languages
English (en)
French (fr)
Inventor
谢朴锐
Original Assignee
深圳新度博望科技有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 深圳新度博望科技有限公司 filed Critical 深圳新度博望科技有限公司
Publication of WO2024087796A1 publication Critical patent/WO2024087796A1/zh

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q30/00Commerce
    • G06Q30/02Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241Advertisements
    • G06Q30/0251Targeted advertisements
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q30/00Commerce
    • G06Q30/02Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241Advertisements
    • G06Q30/0273Determination of fees for advertising

Definitions

  • the present application relates to the field of artificial intelligence, and in particular to a method, device, computer and storage medium for delivering video advertisements.
  • the current video advertising delivery algorithm usually uses the results of the recommended advertising ranking to display them to users in order.
  • the recommended advertising ranking algorithm is often based on the revenue obtained from each thousand times the advertisement is displayed.
  • Some algorithms also refer to the completeness of the video advertisement playback on this basis.
  • video advertisements have many different characteristics from other advertisements.
  • the technical problem to be solved by the embodiments of the present application is to provide a video advertisement delivery method, device, computer equipment and storage medium.
  • the embodiment of the present application provides a video advertisement delivery method, including:
  • custom parameters of the video advertisement include at least one of the following parameters: interaction parameters, advertisement information, network information, and device information;
  • video advertisements are delivered to users in sequence.
  • the method after setting one or more custom parameters of the video advertisement, the method also includes:
  • the importance of the custom parameter is determined according to the parameter weight.
  • the method before calculating the ranking of the video advertisements according to the custom parameters, the method further includes:
  • the custom parameters and parameter weights of the custom parameters are iteratively adjusted according to the parameter log.
  • the real-time acquisition of the custom parameter includes:
  • the interaction parameter includes information generated by all interaction behaviors between the user and the video advertisement.
  • performing a ranking calculation on the video advertisements according to the custom parameters to calculate a ranking result of the video advertisements includes:
  • the video advertisements are sorted according to the average value to obtain a sorting result of the video advertisements.
  • the ranking result of the video advertisements includes:
  • the ranking results of the video advertisements in different time periods are different.
  • the embodiment of the present application further provides a video advertisement delivery device, including:
  • a parameter setting module used to set one or more custom parameters of the video advertisement, wherein the custom parameters include interactive parameters, advertisement information, network information, and device information;
  • a sorting module used to perform sorting calculation on the video advertisement according to the custom parameters, and calculate the sorting result of the video advertisement
  • the delivery module delivers video advertisements to users in sequence according to the ranking results.
  • an embodiment of the present application further provides a computer device, including a memory, a transceiver, a processor, and a bus system, characterized in that it includes:
  • the memory is used to store programs
  • the processor is used to execute the program stored in the memory.
  • the processor executes the program stored in the memory, the processor is used to execute the steps of implementing the above-mentioned video advertisement delivery method.
  • an embodiment of the present application also provides a computer-readable storage medium, including instructions, which, when executed on a computer, enable the computer to execute the steps of implementing the above-mentioned video advertisement delivery method.
  • the present application sets one or more custom parameters of the video advertisement, the custom parameters include at least one of the following parameters: interactive parameters, advertisement information, network information, device information; the video advertisement is sorted and calculated according to the custom parameters, and the sorting result of the video advertisement is calculated; according to the sorting result, the video advertisement is delivered to the user in sequence. Since multiple custom parameters are set, and the custom parameters include interactive parameters, the characteristics of the video advertisement can be reflected, making the sorting calculation process more personalized and scenario-based, and making the final sorting result more accurate and reasonable.
  • FIG1 is a schematic diagram of an application environment of a method for delivering video advertisements in an embodiment of the present application
  • FIG2 is a schematic flow chart of a method for delivering video advertisements in one embodiment of the present application.
  • FIG3 is a principle block diagram of a video advertisement delivery device in an embodiment of the present application.
  • FIG. 4 is a schematic diagram of a computer device in an embodiment of the present application.
  • the video advertisement delivery method provided in the embodiment of the present application can be applied in the application environment as shown in FIG1 , wherein the terminal device communicates with the server through the network, and the terminal device can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and portable wearable devices.
  • the server can be implemented as an independent server or a server cluster composed of multiple servers.
  • a video advertisement delivery method is provided, which is described by taking the method applied to the server in FIG. 1 as an example, and includes the following steps:
  • S10 Setting one or more custom parameters of the video advertisement, wherein the custom parameters include at least one of the following parameters: interaction parameters, advertisement information, network information, and device information.
  • Custom parameters include but are not limited to interactive parameters, advertisement information, network information, device information, etc.
  • Interaction refers to all interactive behaviors between users and video advertisements, including but not limited to users clicking on advertisements, users closing advertisements, etc.
  • Interactive parameters refer to the information generated by all interactive behaviors between users and the video advertisements, including but not limited to the length of time users click on advertisements, the speed at which users close advertisements, etc.
  • Advertisement information refers to the information of the current video advertisement, including but not limited to advertisement pricing, advertisement classification, etc.
  • Network information refers to the network information of the client that plays the video advertisement, including but not limited to the network download speed of the client, the upload speed of the client network, etc.
  • Device information refers to the device information of the client that plays the video advertisement, including but not limited to the price of the client device, the system environment of the client device, etc.
  • video ads are different from short videos and text ads.
  • Short videos and text ads need to refer to parameters such as the number of readings, the number of shares, and the number of forwardings, but video ads do not have sharing and forwarding. Therefore, this embodiment introduces interactive parameters, that is, the relevant parameters generated by the user's own operating actions on the video ads, to ensure the accuracy and rationality of the subsequent video ad sorting.
  • the video ads are ranked and calculated according to the set custom parameters to obtain the recommended ranking results of the video ads.
  • the ranking calculation methods include but are not limited to LTR (learning to rank, abbreviated as LTR or L2R) and other methods.
  • the sorting can be better targeted at user actions and ad parameters.
  • this embodiment weakens the user's own attributes during the sorting process.
  • the custom parameters are biased towards the parameters related to interactive operations to ensure that video ads can be delivered more appropriately and reduce users' resistance to video ads.
  • the video ads are delivered to the user in order according to the ranking results.
  • the ranking results are arranged from high to low according to the user's interest in the video ads, so the order of delivering the video ads will also be based on the ranking results.
  • one or more custom parameters of the video advertisement are set, and the custom parameters include at least one of the following parameters: interactive parameters, advertisement information, network information, and device information; the video advertisement is sorted and calculated according to the custom parameters to calculate the sorting result of the video advertisement; and the video advertisement is delivered to the user in sequence according to the sorting result. Since multiple custom parameters are set, corresponding parameter weights are set according to different custom parameters.
  • the added custom parameters include interactive parameters, which enables the characteristics of the video advertisement to be reflected, making the sorting calculation process more personalized and scenario-based, and making the final sorting result more accurate and reasonable.
  • step S20 that is, before calculating the ranking of the video advertisements according to the custom parameters, the following steps are also included:
  • S41 Define one or more time periods.
  • steps S41-S42 a plurality of time periods are defined, and according to the user interaction behaviors prevalent in each time period, the parameter weights of the custom parameters in each time period are set.
  • a day is divided into three time periods: morning, afternoon, and evening.
  • morning, afternoon, and evening Considering that users watch fewer videos during work hours, user interaction behaviors will be significantly different due to different time periods. Therefore, when setting parameter weights, the evening has the largest weight, while the morning and afternoon have smaller weights.
  • sorting results will be generated according to time periods. For example, in the above example, the time periods are divided into morning, afternoon, and evening, and the subsequent sorting results will generate the best sorting in the morning, the best sorting in the afternoon, and the best sorting in the evening. When the user is in the evening, the advertisements will be delivered to the user in sequence according to the best sorting results in the evening.
  • steps S41-S42 it is considered that the user interaction behaviors are obviously different in different time periods, which leads to different interaction parameters in different time periods. Therefore, in order to accurately deliver video ads, the time periods with obviously different user interaction behaviors are sorted separately, so that the final sorting result is more reasonable and the delivered video ads are more accurate. More personalized and scenario-based.
  • step S42 that is, after setting the parameter weight of the custom parameter, the following steps are also included:
  • S52 Acquire the custom parameters in real time, and record the custom parameters acquired each time in a parameter log.
  • a default initial value is first given to the custom parameter to ensure the initial operation of the program.
  • real-time custom parameters are continuously obtained and recorded in the parameter log.
  • machine learning training is performed on the custom parameters, and the training results are set as custom parameters.
  • sorting calculations are performed based on the processed custom parameters to calculate the sorting results.
  • the custom parameters will be reinitialized to the custom parameters after the last processing, and then subsequent sorting calculations will be performed.
  • the initialization and training process is a cyclic iterative adjustment of the custom parameters and parameter weights, which realizes the dynamic adjustment of the custom parameters.
  • the above-mentioned methods for machine learning training of custom parameters include but are not limited to gradient training, regression models, etc.
  • the video ad playback duration staytime in the custom parameter is initialized to 1 minute, and 1 minute is used as the initial value for training staytime.
  • the training results show that staytime is 50 seconds and the weight is 50.
  • the staytime of 50 seconds and the weight of 50 are substituted into the subsequent sorting calculation.
  • staytime is initialized to 50 seconds, and staytime is trained. This cycle is repeated and iterated repeatedly to finally obtain the optimal value of staytime.
  • the custom parameters are repeatedly trained using a machine learning training method to obtain the custom parameters of the optimal solution, which effectively avoids manual parameter adjustment, solves the problem caused by manual parameter adjustment, reduces the labor cost of sorting calculation, and improves the calculation efficiency and rationality of sorting.
  • This embodiment effectively integrates multiple custom parameters through the above-mentioned method of automatically adjusting custom parameters, avoiding overfitting of subsequent sorting calculation formulas.
  • step S52 that is, obtaining the custom parameters in real time, specifically includes the following steps:
  • S527 Obtaining the duration from when the close button of the video advertisement is displayed to when it is closed;
  • step S521 the price of the video advertisement, that is, the pricing of the video advertisement, is obtained, and the obtained pricing is set as the advertisement price parameter in the custom parameter.
  • the advertisement price parameter price in the custom parameter is set to 10 yuan.
  • step S522 the price of the playback device where the video advertisement client is located is obtained.
  • the video advertisement is played on a mobile phone, the price of the mobile phone is obtained to be 1,000 yuan, and the device price phoneprice in the custom parameter is set to 1,000 yuan.
  • step S524 the network speed of the video advertisement client is obtained. For example, if the video advertisement is played on a mobile phone and the network download speed of the mobile phone is 2MB/s, the download rate in the custom parameter is set to 2MB/s.
  • step S525-S528 and step S523 the number of times the video ad is delivered, the duration from the start of the video ad to the jump to the ad page, the number of times the video ad is closed, the duration from the display of the close button of the video ad to the closing button, and the playing time of the video ad are obtained.
  • the interaction parameters in the custom parameters can be obtained. For example, the number of times the video ad is delivered is 100 times, and the number of times it is closed is 60 times. The calculated proportion of video ad closed is closerate That is 60%, then set the video closing ratio in the custom parameters to 60%.
  • steps S521-S528, various custom parameters are obtained, including but not limited to interactive parameters, network information, advertising information, and device information.
  • the setting of these custom parameters, especially interactive parameters better reflects the different characteristics of video ads compared to other types of ads. For example, the playing time of video ads is a unique feature of video ads, which other types of ads do not have. Therefore, the setting of these parameters well supports the personalization and scenario-based delivery of video ads.
  • step S23 the custom parameters are input into the ranking learning model to obtain the ranking result of the video advertisement, which specifically includes the following steps:
  • S231 Calculate the video advertisements one by one according to the custom parameters and the parameter weights of the custom parameters to obtain scores of each custom parameter in the video advertisement.
  • S233 Sort the video advertisements according to the average value to obtain a sorting result of the video advertisements.
  • a score ranking algorithm is used to calculate the score of each video advertisement.
  • the specific calculation formula is as follows:
  • p is a custom parameter
  • w is the parameter weight of the custom parameter
  • i is the array index of the custom parameter
  • N is the number of parameters
  • Score is the video ad score
  • step S233 the video ads are sorted according to the calculation results of the above S231-S232.
  • the sorting method can be from high score to low score, or from low score to high score, which is not limited here.
  • the video ads that the user is more interested in are arranged at the front, that is, the video ads are sorted in the order from large to small according to the above parameter scores. According to the sorting results, the video ads are delivered to the user in order.
  • the device price parameter price in the custom parameters of the video ad is 1,000 yuan
  • the parameter weight is 20
  • the video ad is closed for 50%
  • the parameter weight is 50
  • the video ad stay time is 60s
  • the parameter weight is 30, then according to the above formula, the following formula is obtained:
  • the score of the video ad is 65.86. Assuming that the video ad is A, and there is also a video ad B with a score of 82 and a video ad C with a score of 12, then the video ads will be delivered in descending order according to the scores, that is, B ad will be delivered to the user first, then A ad, and finally C ad.
  • the ranking formula in steps S231-S233 scores the video ads by calculating the average of all parameters and their weights, and then considers the partial order between the scores of the custom parameters to obtain the final ranking result. This allows the ranking result to take into account the different weights of different custom parameters, making the final video ad delivery more scientific, reasonable and accurate.
  • step S30 namely the ranking result of the video advertisements, is specifically the ranking result of the video advertisements in different time periods.
  • the time periods are divided into morning, afternoon, and evening.
  • the custom parameter staytime of video ad A is 50s
  • the weight in the morning is 20
  • the weight in the afternoon is 30, and the weight in the evening is 60
  • the custom parameter staytime of video ad B is 60s
  • the weight in the morning is 50
  • the weight in the afternoon is 10
  • the weight in the evening is 20
  • the calculated scores of video ad A in the morning are 33.98, 50.97 in the afternoon, and 101.94 in the evening
  • the scores of video ad B in the morning are 88.91, 17.78 in the afternoon, and 35.56 in the evening.
  • the final sorting results will be divided into three time periods, with B>A in the morning, and B>A in the afternoon.
  • the result is A>B in the morning, and the result is A>B in the evening. Therefore, if you need to display ads to users in the morning, B ads will be prioritized. If you need to display ads to users in the afternoon or evening, A ads will be prioritized.
  • different custom parameter weights are set according to different time periods to obtain different ranking results for different time periods. Differentiating different time periods can effectively distinguish the user's interaction behaviors that differ greatly in different time periods, so as to obtain the user's optimal video advertising delivery plan in different time periods.
  • a video advertisement delivery device which corresponds to the video advertisement delivery method in the above embodiment.
  • the video advertisement delivery device includes a parameter setting module, a sorting module and a delivery module.
  • Each functional module is described in detail as follows:
  • a parameter setting module used to set one or more custom parameters of the video advertisement, wherein the custom parameters include interactive parameters, advertisement information, network information, and device information;
  • a sorting module used to perform sorting calculation on the video advertisement according to the custom parameters, and calculate the sorting result of the video advertisement
  • the delivery module delivers video advertisements to users in sequence according to the ranking results.
  • Each module in the above-mentioned video advertisement delivery device can be implemented in whole or in part by software, hardware and a combination thereof.
  • the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
  • a computer device which may be a server, and its internal structure diagram may be shown in FIG4.
  • the computer device includes a processor, a memory, a network interface, and a database connected via a system bus.
  • the processor of the computer device is used to provide computing and control capabilities.
  • the memory of the computer device includes a non-volatile storage medium and an internal memory.
  • the non-volatile storage medium stores an operating system, computer-readable instructions, and a database.
  • the internal memory provides an environment for the operation of the operating system and computer-readable instructions in the non-volatile storage medium.
  • the network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer-readable instructions are executed by the processor, a method for delivering video advertisements is implemented.
  • a computer device including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, wherein the processor executes the computer-readable instructions to implement the following steps:
  • custom parameters of the video advertisement include at least one of the following parameters: interaction parameters, advertisement information, network information, and device information;
  • video advertisements are delivered to users in sequence.
  • a computer-readable storage medium on which computer-readable instructions are stored, and when the computer-readable instructions are executed by a processor, the following steps are implemented:
  • custom parameters of the video advertisement include at least one of the following parameters: interaction parameters, advertisement information, network information, and device information;
  • video advertisements are delivered to users in sequence.
  • Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory.
  • Volatile memory may include random access memory (RAM) or external cache memory.
  • RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Strategic Management (AREA)
  • Finance (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本申请涉及人工智能领域,公开了一种视频广告投放方法、装置、计算机及存储介质,引入用户与视频广告的互动参数,以解决视频广告的投放缺乏准确性的问题。所述方法包括:设定视频广告的一个或多个自定义参数,自定义参数至少包括以下一种参数:互动参数、广告信息、网络信息、设备信息;根据自定义参数对视频广告进行排序计算,计算出视频广告的排序结果;根据排序结果,依序向用户投放视频广告。

Description

一种视频广告投放方法、装置、计算机设备及存储介质
本申请以2022年10月27日提交的申请号为“202211340039.9”,名称为“一种视频广告投放方法、装置、计算机设备及存储介质”的中国发明专利申请为基础,并要求其优先权。
技术领域
本申请涉及人工智能领域,尤其涉及一种视频广告投放方法、装置、计算机及存储介质。
背景技术
随着视频技术愈发成熟,视频广告越来越频繁地出现在人们的视野中。而为合适的人群投放合适的广告,使广告能被推送到最需要它的人面前,因此视频广告投放算法成为了视频广告中最为重要的技术之一。
当前视频广告投放算法通常采用推荐广告排序的结果,按次序向用户展示。而推荐广告排序的算法往往是根据广告每展示一千次所获得的收益进行排序的,还有部分的算法会在此基础上参考视频广告播放的完整度。而实际上,视频广告与其他广告有诸多不同特征,为了保证视频广告投放的准确性,需要另外引入与视频广告相关的参数。但上述排序的算法难以引入更多的参数,这使得视频广告的投放缺乏准确性。
发明内容
本申请实施例所要解决的技术问题在于,提供一种视频广告投放方法、装置、计算机设备及存储介质。可以在。
为了解决上述技术问题,本申请实施例提供了一种视频广告投放方法,包括:
设定视频广告的一个或多个自定义参数,所述自定义参数至少包括以下一种参数:互动参数、广告信息、网络信息、设备信息;
根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果;
根据所述排序结果,依序向用户投放视频广告。
在一种可能的实现中,所述设定视频广告的一个或多个自定义参数之后,所述方法 还包括:
划定一个或多个时间段;
在每一个所述时间段内,分别设定所述自定义参数的参数权重;
根据所述参数权重,确定所述自定义参数的重要程度。
在一种可能的实现中,所述根据所述自定义参数对所述视频广告进行排序计算之前,所述方法还包括:
初始化所述自定义参数;
实时获取所述自定义参数,将每次获取到的所述自定义参数都记录在参数日志中;
根据所述参数日志,迭代调整所述自定义参数和所述自定义参数的参数权重。
在一种可能的实现中,所述实时获取所述自定义参数,包括:
获取所述视频广告的价格;
获取所述视频广告播放设备的价格;
获取所述视频广告的投放次数;
获取所述视频广告播放网络的速度;
获取所述视频广告从开始到跳转进入广告页面的时长;
获取所述视频广告被关闭的次数;
获取所述视频广告的关闭按钮从显示到被关闭的时长;
获取所述视频广告的播放时长。
在一种可能的实现中,所述互动参数包括用户与所述视频广告的所有交互行为所产生的信息。
在一种可能的实现中,所述根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果,包括:
根据所述自定义参数及所述自定义参数的参数权重,对所述视频广告逐一进行计算,得到所述视频广告中各项所述自定义参数的分数;
分别计算每个所述视频广告中所有所述分数的平均值;
根据所述平均值,对所述视频广告进行排序,得到所述视频广告的排序结果。
在一种可能的实现中,所述视频广告的排序结果,包括:
所述视频广告在不同时间段的排序结果。
相应的,本申请实施例还提供了一种视频广告投放装置,包括:
参数设定模块,用于设定视频广告的一个或多个自定义参数,所述自定义参数包括互动参数、广告信息、网络信息、设备信息;
排序模块,用于根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果;
投放模块,根据所述排序结果,依序向用户投放视频广告。
相应的,本申请实施例还提供了一种计算机设备,包括存储器、收发器、处理器以及总线系统,其特征在于,包括:
所述存储器,用于存储程序;
所述处理器,用于执行所述存储器存储的程序,当所述处理器执行所述存储器存储的程序时,所述处理器用于执行实现上述视频广告投放方法的步骤。
相应的,本申请实施例还提供了一种计算机可读存储介质,,包括指令,当其在计算机上运行时,使得计算机执行实现上述视频广告投放方法的步骤。
上述视频广告投放方法、装置、计算机设备及存储介质,本申请通过设定视频广告的一个或多个自定义参数,所述自定义参数至少包括以下一种参数:互动参数、广告信息、网络信息、设备信息;根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果;根据所述排序结果,依序向用户投放视频广告。由于设定多个自定义参数,而自定义参数中包含互动参数,这使得使视频广告的特征得以体现,使得排序计算的过程更加个性化、场景化,令最终排序的结果更加准确、合理。
本申请的一个或多个实施例的细节在下面的附图和描述中提出,本申请的其他特征和优点将从说明书、附图以及权利要求变得明显。
附图说明
为了更清楚地说明本申请实施例的技术方案,下面将对本申请实施例的描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动性的前提下,还可以根据这些附图获得其他的附图。
图1是本申请一实施例中视频广告投放方法的一应用环境示意图;
图2是本申请一实施例中视频广告投放方法的一示意流程图;
图3是本申请一实施例中视频广告投放装置的一原理框图;
图4是本申请一实施例中计算机设备的一示意图。
具体实施方式
下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整 地描述,显然,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
本申请实施例提供的视频广告投放方法,可应用在如图1的应用环境中,其中,终端设备通过网络与服务器进行通信,终端设备可以但不限于各种个人计算机、笔记本电脑、智能手机、平板电脑和便携式可穿戴设备。服务器可以用独立的服务器或者是多个服务器组成的服务器集群来实现。
在一实施例中,如图2所示,提供一种视频广告投放方法,以该方法应用在图1中的服务器为例进行说明,包括如下步骤:
S10:设定视频广告的一个或多个自定义参数,所述自定义参数至少包括以下一种参数:互动参数、广告信息、网络信息、设备信息。
本实施例中,首先需要设定视频广告排名需要参考的参数,即自定义参数。自定义参数包括但不仅限于互动参数、广告信息、网络信息、设备信息等。其中互动指用户与视频广告的所有交互行为,包括但不仅限于用户点开广告、用户关闭广告等,互动参数指用户与所述视频广告的所有交互行为所产生的信息,包括但不仅限于用户点击广告的时长、用户关闭广告的速度等。广告信息指当前视频广告的信息,包括但不仅限于广告定价,广告分类等。网络信息指播放视频广告的客户端网络信息,包括但不仅限于客户端的网络下载速度、客户端网络的上传速度等。设备信息指播放视频广告的客户端设备信息,包括但不仅限于客户端设备的价格,客户端设备的系统环境等。
需要说明的是,视频广告不同于短视频和文字广告,短视频和文字广告更多的需要参考阅读量、分享数量和转发数量等参数,但视频广告则不存在分享与转发,因此本实施例引入了互动参数,即用户本身对视频广告的操作动作产生的相关参数,以确保后续视频广告排序的准确性与合理性。
S20:根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果。
在设定了视频广告的自定义参数之后,根据所设定的自定义参数对视频广告进行排序计算,得出视频广告的推荐排序结果。其中排序计算的方法包括但不仅限于LTR(排序学习,全称为Learning to rank,简称LTR或L2R)等方法。
通过自定义参数对视频广告进行排序,使排序可以更好地针对用户操作动作及广告本身参数进行,并且为保证用户隐私,在排序的过程中,本实施例弱化了用户本身的属 性,对自定义参数偏重于互动操作的相关参数,以保证视频广告能更恰当地投放,减少用户对视频广告的抗拒心理。
S30:根据所述排序结果,依序向用户投放视频广告。
在计算出视频广告的排序结果后,依照排序结果按顺序向用户投放视频广告。在本实施例中,排序结果将按照用户对视频广告的兴趣程度由大到小排列,因此投放视频广告的顺序也将依照此排序结果,依次投放。
在步骤S10-S30中,通过设定视频广告的一个或多个自定义参数,所述自定义参数至少包括以下一种参数:互动参数、广告信息、网络信息、设备信息;根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果;根据所述排序结果,依序向用户投放视频广告。由于设定多个自定义参数,并根据不同的自定义参数设定了对应的参数权重。其中增加的自定义参数中包含互动参数,这使得使视频广告的特征得以体现,使得排序计算的过程更加个性化、场景化,令最终排序的结果更加准确、合理。
在一实施例中,步骤S20之前,即根据所述自定义参数对所述视频广告进行排序计算之前,还包括如下步骤:
S41:划定一个或多个时间段。
S42:在每一个所述时间段内,分别设定所述自定义参数的参数权重。
在步骤S41-S42中,划定多个时间段,并根据每个时间段普遍存在的用户交互行为,设定各个时间段内自定义参数的参数权重。
例如,将一天的时间进行划分,从6:00到12:00划为上午,从12:00到18:00划为下午,其余时间为晚上,即划定三个时间段:上午、下午、晚上。考虑到在上班的时间段里,用户观看视频的较少,用户的交互行为由于时间段的不同,将产生明显的区别。因此设置参数权重时,晚上的权重最大,上午和下午的权重较小。
在后续的排序过程中,将依照时间段,分别生成排序结果。例如,上述事例中,将时间段分为上午、下午、晚上,而后续排序结果也将生成上午的最优排序、下午的最优排序和晚上的最优排序,当用户所处的时间是晚上时,则根据晚上的最优排序结果,依次对用户投放广告。
在步骤S41-S42中,考虑到用户交互行为在不同的时间段里有着明显的不同,即导致了不同时间段内交互参数的不同。因此为了投放视频广告的准确性,对用户交互行为有着明显不同的时间段进行分别排序,使最终排序的结果更加合理,让投放的视频广告 更具个性化、场景化。
在一实施例中,步骤S42之后,即设定所述自定义参数的参数权重之后,还包括如下步骤:
S51:初始化所述自定义参数。
S52:实时获取所述自定义参数,将每次获取到的所述自定义参数都记录在参数日志中。
S53:根据所述参数日志,迭代调整所述自定义参数和所述自定义参数的参数权重。
在本实施例中,先给自定义参数一个默认的初始值,确保程序的初始运行。在设定自定义参数的过程中,不断获取实时的自定义参数并记录在参数日志中。结合参数日志与当前设定的自定义参数,对自定义参数进行机器学习训练,将训练的结果设置为自定义参数,后续根据处理过的自定义参数进行排序计算,计算出排序结果。在之后的运算中,自定义参数将被重新初始化为上次处理后的自定义参数,再进行后续的排序计算。这其中初始化和训练的过程是对自定义参数及参数权重的循环迭代调整,实现了自定义参数的动态调整。上述对自定义参数机器学习训练的方式包括但不仅限于梯度训练、回归模型等。
例如,初始化自定义参数中视频广告播放时长staytime为1min,将1min作为初始值对staytime进行训练,训练得出staytime为50s,权重为50。将staytime为50s,权重为50代入后续的排序计算。下一轮运算时,初始化staytime为50s,对staytime进行训练,周而复始,反复迭代,最终得到staytime的最优值。
在步骤S51-S53中,采用机器学习训练的方法对自定义参数进行反复的训练,得到最优解的自定义参数,有效避免了人工手动调参,解决了手动调参导致的,降低排序计算的人力成本,提高了计算效率和排序的合理性。本实施例通过上述自动调节自定义参数的方法,有效了融合了多个自定义参数,避免后续排序计算公式过拟合。
在一实施例中,步骤S52中,即实时获取所述自定义参数,具体包括如下步骤:
S521:获取所述视频广告的价格;
S522:获取所述视频广告播放设备的价格;
S523:获取所述视频广告的投放次数;
S524:获取所述视频广告播放网络的速度;
S525:获取所述视频广告从开始到跳转进入广告页面的时长;
S526:获取所述视频广告被关闭的次数;
S527:获取所述视频广告的关闭按钮从显示到被关闭的时长;
S528:获取所述视频广告的播放时长。
在步骤S521中,获取视频广告的价格,即视频广告的定价,将获取到的定价设置为自定义参数中的广告价格参数。例如,设置自定义参数中广告价格参数price为10元。
在步骤S522中,获取视频广告客户端所处的播放设备的价格,例如,视频广告在手机上播放,获取该手机的价格为1000元,设置自定义参数中设备价格phoneprice为1000元。
在步骤S524中,获取视频广告客户端所处的网络速度,例如视频广告在手机上播放,获取该手机所处的网络下载速度为2MB/s,则设置自定义参数中下载速度downloadrate为2MB/s。
在步骤S525-S528,以及步骤S523中,获取了视频广告的投放次数、视频广告从开始到跳转进入广告页面的时长、视频广告被关闭的次数、视频广告的关闭按钮从显示到被关闭的时长以及视频广告的播放时长。从获取到的数据中,可以得出自定义参数中的交互参数。例如,获取到视频广告的投放次数为100次,被关闭的次数为60次,计算出视频广告被关闭的比例closerate为即60%,则设置自定义参数中视频关闭的比例为60%。
在步骤S521-S528中,获取了各类自定义参数,包括但不仅限于互动参数、网络信息、广告信息和设备信息。这些自定义参数的设定,尤其是互动参数,更好地体现了视频广告相较于其他类型广告不一样的特征,例如视频广告的播放时长就是视频广告特有的特征,其他类型的广告并不具备。因此这些参数的设定很好地支持了视频广告投放中的个性化、场景化。
在一实施例中,步骤S23中,即将所述自定义参数输入所述排序学习模型,得出所述视频广告的排序结果,具体包括如下步骤:
S231:根据所述自定义参数及所述自定义参数的参数权重,对所述视频广告逐一进行计算,得到所述视频广告中各项所述自定义参数的分数。
S232:分别计算每个所述视频广告中所有所述分数的平均值。
S233:根据所述平均值,对所述视频广告进行排序,得到所述视频广告的排序结果。
在步骤S231-S232中,采用得分排序算法,对每个视频广告进行分数计算,具体的计算公式如下:
其中,p为自定义参数,w为自定义参数的参数权重,i为自定义参数的数组下标,N为参数个数,Score为视频广告得分。
在步骤S233中,根据上述S231-S232的计算结果对视频广告进行排序。排序方式可以由分数高排到分数低,也可以由分数低排列到分数高,此处不作限制。本实施例中,将用户更感兴趣的视频广告排列靠前,也就是说,将上述参数得分依照从大到小的顺序对视频广告进行排序。依照排序结果,将视频广告依序向用户投放。
例如,视频广告的自定义参数中设备价格参数price为1000元,该参数权重为20,视频广告被关闭占比为50%,该参数权重为50,视频广告停留时长staytime为60s,该参数权重为30,则依据上述公式,得到下面的算式:
根据上述算式,得出该视频广告的得分Score为65.86,假设该视频广告为A,另外还有视频广告B得分为82,视频广告C得分为12,那么依照得分由大到依次进行视频广告投放,则会先向用户投放B广告,然后投放A广告,最后投放C广告。
在步骤S231-S233中的排序公式,通过计算所有参数及其权重的平均值,给视频广告进行评分,继而考虑自定义参数得分间的偏序,从而得出最终的排序结果。这使得排序结果考虑了不同自定义参数的不同权重,让最终的视频广告投放更加科学合理,更具精准度。
在一实施例中,步骤S30中,即视频广告的排序结果,具体为视频广告在不同时间段的排序结果。
在不同时间段中,根据不同的时间段设定自定义参数的不同参数权重,继而分别计算不同时间段的排序结果。最终根据当前时间选择对应的排序结果,进行视频广告的投放。
例如,如上述步骤S41-S42的示例,将时间段分割为上午、下午、晚上。假设视频广告A的自定义参数staytime为50s,在上午的权重为20,下午的权重为30,晚上的权重为60,而视频广告B的自定义参数staytime为60s,在上午的权重为50,下午的权重为10,晚上的权重为20,则计算得出视频广告A在上午的得分为33.98,下午的得分为50.97,晚上的得分为101.94,视频广告B在上午的得分为88.91,下午的得分为17.78,晚上的得分为35.56。因此最终的排序结果将分为三个时间段,上午的结果为B>A,下午 的结果为A>B,晚上的结果为A>B。故此若在上午需要向用户投放广告,则优先投放B广告,若是下午或晚间需要向用户投放广告,则优先投放A广告。
在本实施例中,依照不同的时间段,设置不同的自定义参数权重,以便得到不同时间段的不同排序结果。区分不同时间段能够有效地区别用户在不同时间段区别较大的交互行为,以便得到用户在不同时间段的最优视频广告投放方案。
应理解,上述实施例中各步骤的序号的大小并不意味着执行顺序的先后,各过程的执行顺序应以其功能和内在逻辑确定,而不应对本申请实施例的实施过程构成任何限定。
在一实施例中,提供一种视频广告投放装置,该视频广告投放装置与上述实施例中视频广告投放方法一一对应。如图3所示,该视频广告投放装置包括参数设定模块、排序模块和投放模块。各功能模块详细说明如下:
参数设定模块,用于设定视频广告的一个或多个自定义参数,所述自定义参数包括互动参数、广告信息、网络信息、设备信息;
排序模块,用于根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果;
投放模块,根据所述排序结果,依序向用户投放视频广告。
关于视频广告投放装置的具体限定可以参见上文中对于视频广告投放方法的限定,在此不再赘述。上述视频广告投放装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中,也可以以软件形式存储于计算机设备中的存储器中,以便于处理器调用执行以上各个模块对应的操作。
在一个实施例中,提供了一种计算机设备,该计算机设备可以是服务器,其内部结构图可以如图4所示。该计算机设备包括通过系统总线连接的处理器、存储器、网络接口和数据库。其中,该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性存储介质、内存储器。该非易失性存储介质存储有操作系统、计算机可读指令和数据库。该内存储器为非易失性存储介质中的操作系统和计算机可读指令的运行提供环境。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机可读指令被处理器执行时以实现一种视频广告投放的方法。
在一个实施例中,提供了一种计算机设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机可读指令,处理器执行计算机可读指令时实现以下步骤:
设定视频广告的一个或多个自定义参数,所述自定义参数至少包括以下一种参数:互动参数、广告信息、网络信息、设备信息;
根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果;
根据所述排序结果,依序向用户投放视频广告。
在一个实施例中,提供了一种计算机可读存储介质,其上存储有计算机可读指令,计算机可读指令被处理器执行时实现以下步骤:
设定视频广告的一个或多个自定义参数,所述自定义参数至少包括以下一种参数:互动参数、广告信息、网络信息、设备信息;
根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果;
根据所述排序结果,依序向用户投放视频广告。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机可读指令来指令相关的硬件来完成,所述的计算机可读指令可存储于一非易失 性计算机可读取存储介质中,该计算机可读指令在执行时,可包括如上述各方法的实施例的流程。其中,本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用,均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限,RAM以多种形式可得,诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchlink)DRAM(SLDRAM)、存储器总线(Rambus)直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。
所属领域的技术人员可以清楚地了解到,为了描述的方便和简洁,仅以上述各功能单元、模块的划分进行举例说明,实际应用中,可以根据需要而将上述功能分配由不同的功能单元、模块完成,即将所述装置的内部结构划分成不同的功能单元或模块,以完成以上描述的全部或者部分功能。
以上所述实施例仅用以说明本申请的技术方案,而非对其限制;尽管参照前述实施例对本申请进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围,均应包含在本申请的保护范围之内。

Claims (10)

  1. 一种视频广告投放方法,包括:
    设定视频广告的一个或多个自定义参数,所述自定义参数至少包括以下一种参数:互动参数、广告信息、网络信息、设备信息;
    根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果;
    根据所述排序结果,依序向用户投放视频广告。
  2. 如权利要求1所述的视频广告投放方法,所述根据所述自定义参数对所述视频广告进行排序计算之前,所述方法还包括:
    划定一个或多个时间段;
    在每一个所述时间段内,分别设定所述自定义参数的参数权重。
  3. 如权利要求2所述的视频广告投放方法,所述设定所述自定义参数的参数权重之后,所述方法还包括:
    初始化所述自定义参数;
    实时获取所述自定义参数,将每次获取到的所述自定义参数都记录在参数日志中;
    根据所述参数日志,迭代调整所述自定义参数和所述自定义参数的参数权重。
  4. 如权利要求3所述的视频广告投放方法,所述实时获取所述自定义参数,包括:
    获取所述视频广告的价格;
    获取所述视频广告播放设备的价格;
    获取所述视频广告的投放次数;
    获取所述视频广告播放网络的速度;
    获取所述视频广告从开始到跳转进入广告页面的时长;
    获取所述视频广告被关闭的次数;
    获取所述视频广告的关闭按钮从显示到被关闭的时长;
    获取所述视频广告的播放时长。
  5. 如权利要求1所述的视频广告投放方法,所述互动参数包括用户与所述视频广告的所有交互行为所产生的信息。
  6. 如权利要求1所述的视频广告投放方法,所述根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果,包括:
    根据所述自定义参数及所述自定义参数的参数权重,对所述视频广告逐一进行计算,得到所述视频广告中各项所述自定义参数的分数;
    分别计算每个所述视频广告中所有所述分数的平均值;
    根据所述平均值,对所述视频广告进行排序,得到所述视频广告的排序结果。
  7. 如权利要求1至6任一项所述的视频广告投放方法,所述视频广告的排序结果,包括:
    所述视频广告在不同时间段的排序结果。
  8. 一种视频广告投放装置,包括:
    参数设定模块,用于设定视频广告的一个或多个自定义参数,所述自定义参数包括互动参数、广告信息、网络信息、设备信息;
    排序模块,用于根据所述自定义参数对所述视频广告进行排序计算,计算出所述视频广告的排序结果;
    投放模块,根据所述排序结果,依序向用户投放视频广告。
  9. 一种计算机设备,包括存储器、收发器、处理器以及总线系统,包括:
    所述存储器,用于存储程序;
    所述处理器,用于执行所述存储器存储的程序,当所述处理器执行所述存储器存储的程序时,所述处理器用于执行权利要求1至7中任一项所述的方法。
  10. 一种计算机可读存储介质,包括指令,当其在计算机上运行时,使得计算机执行如权利要求1至7中任一项所述的方法。
PCT/CN2023/111672 2022-10-27 2023-08-08 一种视频广告投放方法、装置、计算机设备及存储介质 WO2024087796A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211340039.9 2022-10-27
CN202211340039.9A CN115689649A (zh) 2022-10-27 2022-10-27 一种视频广告投放方法、装置、计算机设备及存储介质

Publications (1)

Publication Number Publication Date
WO2024087796A1 true WO2024087796A1 (zh) 2024-05-02

Family

ID=85046411

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/111672 WO2024087796A1 (zh) 2022-10-27 2023-08-08 一种视频广告投放方法、装置、计算机设备及存储介质

Country Status (2)

Country Link
CN (1) CN115689649A (zh)
WO (1) WO2024087796A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115689649A (zh) * 2022-10-27 2023-02-03 深圳新度博望科技有限公司 一种视频广告投放方法、装置、计算机设备及存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8296179B1 (en) * 2007-05-02 2012-10-23 Monster Worldwide, Inc. Targeted advertisement placement based on explicit and implicit criteria matching
CN104732413A (zh) * 2013-12-20 2015-06-24 中国科学院声学研究所 一种智能个性化视频广告推送方法及系统
CN110310148A (zh) * 2019-06-05 2019-10-08 上海易点时空网络有限公司 基于大数据和机器学习的广告精准投放方法
CN111967901A (zh) * 2020-07-31 2020-11-20 深圳市彬讯科技有限公司 广告商家排序方法、装置、计算机设备及存储介质
CN113506123A (zh) * 2021-05-18 2021-10-15 太龙智显科技(深圳)有限公司 一种广告显示控制方法、装置、计算机设备以及存储介质
CN115689649A (zh) * 2022-10-27 2023-02-03 深圳新度博望科技有限公司 一种视频广告投放方法、装置、计算机设备及存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8296179B1 (en) * 2007-05-02 2012-10-23 Monster Worldwide, Inc. Targeted advertisement placement based on explicit and implicit criteria matching
CN104732413A (zh) * 2013-12-20 2015-06-24 中国科学院声学研究所 一种智能个性化视频广告推送方法及系统
CN110310148A (zh) * 2019-06-05 2019-10-08 上海易点时空网络有限公司 基于大数据和机器学习的广告精准投放方法
CN111967901A (zh) * 2020-07-31 2020-11-20 深圳市彬讯科技有限公司 广告商家排序方法、装置、计算机设备及存储介质
CN113506123A (zh) * 2021-05-18 2021-10-15 太龙智显科技(深圳)有限公司 一种广告显示控制方法、装置、计算机设备以及存储介质
CN115689649A (zh) * 2022-10-27 2023-02-03 深圳新度博望科技有限公司 一种视频广告投放方法、装置、计算机设备及存储介质

Also Published As

Publication number Publication date
CN115689649A (zh) 2023-02-03

Similar Documents

Publication Publication Date Title
CN109783730A (zh) 产品推荐方法、装置、计算机设备和存储介质
US20220198289A1 (en) Recommendation model training method, selection probability prediction method, and apparatus
CN109543111B (zh) 推荐信息筛选方法、装置、存储介质及服务器
US20230088171A1 (en) Method and apparatus for training search recommendation model, and method and apparatus for sorting search results
CN110476182B (zh) 分享图像的自动建议
US20170372329A1 (en) Predictive analytics for providing targeted information
CN107273436A (zh) 一种推荐模型的训练方法和训练装置
US11663250B2 (en) Method for capturing and updating database entries of CRM system based on voice commands
WO2019080662A1 (zh) 信息推荐方法及装置、设备
WO2024087796A1 (zh) 一种视频广告投放方法、装置、计算机设备及存储介质
WO2020244152A1 (zh) 数据推送方法、装置、计算机设备和存储介质
CN110413867B (zh) 用于内容推荐的方法及系统
US11361152B2 (en) System and method for automated content labeling
WO2021114936A1 (zh) 信息推荐方法、装置、电子设备及计算机可读存储介质
US20200349415A1 (en) Method for capturing and storing contact information from a physical medium using machine learning
WO2020253360A1 (zh) 应用程序的内容展示方法、装置、存储介质和计算机设备
CN111967914A (zh) 基于用户画像的推荐方法、装置、计算机设备和存储介质
CN106846191A (zh) 一种课程列表的编排方法、系统及服务器
CN113015010A (zh) 推送参数确定方法、装置、设备及计算机可读存储介质
CN112581226A (zh) 产品推荐方法及系统
CN108509041A (zh) 用于执行操作的方法和装置
WO2022095661A1 (zh) 推荐模型的更新方法、装置、计算机设备和存储介质
CN113850416A (zh) 广告推广合作对象确定方法和装置
CN111738754A (zh) 对象推荐方法及装置、存储介质、计算机设备
CN117057855A (zh) 一种数据处理方法及相关装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23881382

Country of ref document: EP

Kind code of ref document: A1