CN111062746B - Advertisement flow prediction method and device and electronic equipment - Google Patents

Abstract

The application discloses an advertisement traffic prediction method, an advertisement traffic prediction device and electronic equipment. The method comprises the steps of obtaining historical flow data of production within a preset time period according to target casting conditions; selecting the sub-dimensions based on the order of the sub-dimensions in the specified merging dimension performs the following operations: merging the historical flow data based on the selected current sub-dimension; if the data quantity difference value of the historical flow data in each two appointed unit time periods obtained based on the combination of the current sub-dimensions is smaller than a preset value, carrying out flow prediction according to a preset distributed mixed time sequence prediction model and the historical flow data in the appointed unit time periods under the current sub-dimensions to obtain a plurality of groups of predicted flow, otherwise, returning to the next sub-dimension; and selecting a group of estimated flow meeting preset conditions as an advertisement flow estimated value according to the loss value of each group of estimated flow so as to improve the accuracy of the flow estimated result.

Description

Advertisement flow estimation method, device and electronic equipment

technical field

The present application relates to the technical field of advertising, and in particular to an advertising traffic estimation method, device and electronic equipment.

Background technique

With the rapid development of Internet advertising, various forms of charging have appeared in advertising, for example, CPM (Cost Per Mille, charging based on thousand impressions), CPA (Cost Per Action, charging based on user behavior after promotion), ADX (Ad Exchange, Internet advertising trading platform), etc. Among them, taking ADX as an example, advertising traffic estimation is an indispensable key step in ADX, involving various stages in the advertising process, for example, advertising space traffic estimation in the early stage of advertising sales, traffic estimation under various fixed investment conditions ; Another example is the traffic forecast in the bidding stage, etc.

At present, in related technologies, the following methods are mainly used for advertising traffic estimation:

(1) Historical average: by counting the various dimensions of historical data that have been placed, take the average of the past 30 days;

(2) Statistics of experience value: the experience value used as a reference for advertising traffic through the statistical data of the media for a period of time;

(3) Traffic forecasting model: simulate the trend of future traffic through a single traffic forecasting model, and calculate the estimated traffic;

However, the above-mentioned advertising flow estimations all have different degrees of limitations. For example, for (1) and (2), the use of historical averages cannot take into account factors such as the periodicity and seasonality of advertising, let alone accidental The influence of factors, such as Double Eleven, Valentine's Day, etc., lead to inaccurate estimates.

For another example, for type (3), since the learning of model parameters is strongly dependent on historical data, a single time series model is easily affected by fluctuations in historical data, and cannot effectively resist fluctuations in accidental factors.

Contents of the invention

In the first aspect, the embodiment of the present application provides a method for estimating advertising traffic, including:

Obtain the historical flow data produced within the preset time period according to the target fixed investment conditions;

Selecting subdimensions based on the order of the subdimensions in the specified merged dimension performs the following operations:

Merging the historical traffic data based on the selected current sub-dimension;

If the data volume difference between the historical traffic data within two specified units of time combined based on the current sub-dimension is less than the preset value, then according to the preset distributed hybrid time series prediction model and the specified unit under the current sub-dimension The historical traffic data within the time period is used for traffic estimation to obtain multiple sets of estimated traffic, otherwise return to the next sub-dimension;

According to the loss value of each group of estimated traffic, a group of estimated traffic meeting preset conditions is selected as an estimated advertising traffic value.

Further, as an optional implementation manner, the distributed hybrid timing prediction model includes multiple different distributed timing prediction models; and

According to the preset distributed hybrid time series prediction model and the historical traffic data within the specified unit time period under the current sub-dimension, the traffic is estimated to obtain multiple sets of estimated traffic, including:

Based on the historical traffic data within the specified unit time period under the current sub-dimension, each distributed time series prediction model is used to perform traffic prediction respectively, and multiple sets of estimated traffic corresponding to each distributed time series prediction model are obtained.

Further, as an optional implementation,

If a group of estimated traffic includes an estimated traffic within a specified unit duration, correspondingly, selecting a group of estimated traffic that meets the preset conditions as an estimated advertising traffic includes: selecting the estimated traffic with the smallest loss value as a The estimated advertising traffic within the specified unit duration;

If a group of estimated traffic includes estimated traffic within at least two specified unit durations, correspondingly, selecting a group of estimated traffic that meets the preset conditions as an estimated advertising traffic includes: separately calculating each group of estimated traffic The average loss value of the estimated traffic within the at least two specified unit durations; selecting a group of estimated traffic with the smallest average loss value as the estimated advertising traffic within the at least two specified unit durations.

Further, as an optional implementation, the step of obtaining the target fixed investment conditions includes:

Displaying the selection interface of the fixed investment conditions, the selection interface of the fixed investment conditions shows that there are multiple fixed investment condition tabs;

In response to an advertiser's selection of fixed investment conditions based on the plurality of fixed investment condition tabs, the target fixed investment conditions are determined according to the selection result.

Further, as an optional implementation, the method further includes:

Before the traffic is estimated according to the preset distributed mixed time series prediction model and the historical traffic data within the specified unit time under the current sub-dimension to obtain multiple sets of estimated traffic, based on the preset Gaussian smoothing model for the current sub-dimension denoises the historical traffic data within the specified unit time period.

Further, as an optional implementation manner, the historical traffic data includes at least one or more of historical click log data, historical display log data, and historical request log data.

In the second aspect, the embodiment of the present application also provides an advertising traffic estimation device, including:

The data acquisition module is used to acquire the historical traffic data produced within the preset time period according to the target fixed investment conditions;

A data merging module, configured to select sub-dimensions based on the order of each sub-dimension in the specified merging dimension, and merge the historical traffic data based on the selected current sub-dimension;

The traffic estimation module is used to estimate the data according to the preset distributed hybrid timing model and The historical traffic data within the specified unit time period under the current sub-dimension is estimated to obtain multiple sets of estimated traffic, otherwise return to the data merging module and enter the next sub-dimension;

The estimation result determination module is configured to select a group of estimated traffic meeting preset conditions as an estimated advertising traffic value according to the loss value of each group of estimated traffic.

Further, as an optional implementation, the distributed hybrid timing estimation model includes multiple different distributed timing estimation models, and the traffic estimation module is further configured to Using the historical traffic data in the network, each distributed time series prediction model is used to perform traffic prediction, and multiple groups of estimated traffic corresponding to each distributed time series prediction model are obtained.

In a third aspect, the embodiment of the present application further provides an electronic device, including:

processor;

memory for storing said processor-executable instructions;

Wherein, the processor is configured to execute the instructions, so as to realize the above advertisement traffic estimation method.

In a fourth aspect, the embodiment of the present application further provides a computer-readable storage medium, and when instructions in the storage medium are executed by a processor in the electronic device, the electronic device can execute the above advertisement traffic estimation method.

The above at least one technical solution adopted in the embodiment of the present application can achieve the following beneficial effects:

When the historical traffic data is obtained according to the target fixed investment conditions for advertising traffic forecasting, the distributed hybrid time series forecasting model can be used for traffic forecasting to obtain multiple sets of estimated traffic, and then selected from the loss values corresponding to each group of estimated traffic The preset traffic whose loss value meets the preset conditions is used as the estimated value of advertising traffic to avoid the influence of historical data fluctuations or accidental factors on the advertising estimation results, effectively ensuring the accuracy of the advertising traffic estimation results.

The above description is only an overview of the technical solution of the application. In order to better understand the technical means of the application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the application more obvious and understandable, the following Specific embodiments of the present application are specifically cited.

Description of drawings

The drawings described here are used to provide a further understanding of the application and constitute a part of the application. The schematic embodiments and descriptions of the application are used to explain the application and do not constitute an improper limitation to the application. In the attached picture:

Fig. 1 is a schematic diagram of an application scenario of a method for estimating advertisement traffic according to an exemplary embodiment.

Fig. 2 is a flowchart of a method for estimating advertisement traffic according to an exemplary embodiment.

Fig. 3 is a schematic diagram of an interface for selecting a fixed investment condition according to an exemplary embodiment.

Fig. 4 is a schematic diagram of an advertisement scheduling interface provided according to an exemplary embodiment.

Fig. 5 is a block diagram of an advertising traffic estimation device according to an exemplary embodiment.

Fig. 6 is a block diagram of an electronic device provided according to an exemplary embodiment.

Detailed ways

In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in conjunction with specific embodiments of the present application and corresponding drawings. Apparently, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.

The technical solutions provided by various embodiments of the present application will be described in detail below in conjunction with the accompanying drawings.

Embodiment one

As shown in FIG. 1 , it is a schematic diagram of an application scenario of an advertisement traffic estimation method according to an exemplary embodiment. The application scenario includes an advertisement management platform and multiple advertisement systems, and the advertisement management platform is connected to each advertisement system respectively. In this application scenario, the media can request advertisements from various advertising systems managed by the advertising management platform, which is called traffic. Advertisers can also place orders with the advertising systems in the advertising management platform to purchase traffic under certain investment conditions. Since media request traffic is affected by multiple factors (such as holidays, media factors, etc.), the traffic received by the advertising system is constantly changing every day. In addition, advertisers will also purchase a certain amount of advertisements from the advertising management platform according to the fixed investment conditions such as regions and groups of people. Therefore, it is necessary to use traffic estimation technology to inform advertisers that under the fixed investment conditions they choose, the advertising system will How much traffic is left to buy in .

Please refer to FIG. 2 , which is a flow diagram of a method for estimating advertisement traffic according to an exemplary embodiment. Part or all of the steps in the method for estimating advertisement traffic may be performed by, but not limited to, the advertisement management shown in FIG. 1 platform execution. Referring to Fig. 2, the advertising traffic management method may include the following steps.

S1. Obtain the historical flow data produced within the preset time period according to the target fixed investment conditions.

Among them, the targeted delivery conditions refer to the various delivery conditions considered when the advertisement is placed, for example, the gender, occupation, age group, region (such as region, country) of the target, and the display platform (such as website, official account, etc.) etc. In actual implementation, the various specific conditions set by the advertiser when placing an advertisement are the target delivery conditions of the advertisement. The historical traffic data can be obtained, but not limited to, from hadoop clusters, etc., and the historical traffic data can at least include one or more of historical click log data, historical display log data, and historical request log data. The preset duration can be set according to actual needs, for example, the preset duration can be half a month, 1 month, 3 months, or a year.

For example, as a possible implementation, the step of obtaining the target fixed investment conditions may include:

S11, displaying the regular investment condition selection interface, where there are multiple regular investment condition selection tabs displayed;

S12. Responding to the selection operation of the fixed investment condition initiated by the advertiser based on the multiple fixed investment condition tabs, and determining the target fixed investment condition according to the selection result.

Please refer to Figure 3. When an advertiser needs to place an advertisement, he or she can select one or more fixed investment conditions through the fixed investment condition tab on the fixed investment condition selection interface as shown in Figure 3. The fixed investment condition selection operation initiated by the selection interface determines the target fixed investment condition. It should be understood that the fixed investment condition selection interface shown in FIG. 3 is only a schematic interface. In actual implementation, the fixed investment condition selection interface may be, but not limited to, that shown in FIG. 3 .

S2, select sub-dimensions based on the order of sub-dimensions in the specified merged dimension, and perform operations in S3 and S4

S3. Merge the historical traffic data based on the selected current sub-dimension.

S4, judging whether the data volume difference of the historical flow data in every two specified unit durations based on the combination of the current sub-dimensions is less than the preset value, if so, execute S5 and S6, otherwise, return to S2 to select the next sub-dimension dimension, and use the next sub-dimension as the current sub-dimension to perform the step of merging the historical traffic data described in S3.

Wherein, the designated merged dimension may be one or more of time, region, occupation, age, etc., assuming that the designated merged dimension is time, then the subdimensions in the designated merged dimension may be minutes, hours, days, months, etc. Alternatively, assuming that the designated merged dimension is region, then the sub-dimensions in the designated merged dimension may be district, city, province, country, etc., which is not limited in this embodiment.

It should be noted that the merging result obtained during data merging is a time series arranged in chronological order according to the generation time of historical traffic data (such as historical click log data, historical display log data, historical request log data, etc.).

Wherein, as shown in S4, the specified unit duration can be sub-dimensions such as minutes, hours, and days involved in the specified merged dimension, or other sub-dimensions (time units) that are set according to requirements. This embodiment No specific restrictions are made. For example, as a possible implementation, take the historical traffic data as the historical advertisement click log, specify the combined dimension as time, and each sub-dimension as minutes, hours, days, and months as an example. For the data involved in S2 to S4 A brief introduction to the merging process.

First of all, historical advertisement click logs can be merged according to the sub-dimension of "minute" to obtain the amount of click log data per minute, and then calculate the difference between the amount of click log data within every two minutes (specified unit duration). Difference, when the difference is less than the preset value, it proves that the result of data merging according to the sub-dimension of "point" is stable, and S5 and S6 can be executed based on this merging result; otherwise, when the difference is not less than the preset value, it proves that the merging result of data merging based on the sub-dimension of "score" is not stable.

Then, it is necessary to enter the next sub-dimension from the "minute" sub-dimension for data merging, that is to say, based on the merging result of the "minute" sub-dimension, continue to perform data merging according to the "time" sub-dimension. Get the amount of click log data per hour, when the difference between the amount of click log data in every two hours (such as 2 o'clock to 3 o'clock and 3 o'clock to 4 o'clock, etc.) is not less than the preset value, It proves that the result of data merging according to the sub-dimension of "time" is also unstable;

Then, it is necessary to enter the next sub-dimension from the "time" sub-dimension for data merging, that is to say, based on the merging results of the "time" sub-dimension, continue to perform data merging according to the "day" sub-dimension, Get the amount of click log data per day... until the result of data merging under the specified merging dimension of "time" reaches stability.

For example, assuming that the data is merged according to the sub-dimension of "day", the difference between the amount of click log data in every two days (specified unit time length) is less than the preset value, which proves that the "day" sub-dimension If the result of data merging in the sub-dimensions is stable, then the step of advertising traffic estimation in S5 can be performed based on the result, so as to ensure the accuracy of the estimation result.

It should be understood that the aforementioned difference can be set according to requirements, and the data amount difference corresponding to different sub-dimensions can be different or the same, which is not limited in this embodiment. In addition, since the merging order of the sub-dimensions in the specified merging dimension and the size of the preset value in S3 determine the accuracy of the advertising traffic estimation result, in this embodiment, when data merging is performed, the specified merging The sub-dimensions involved in the dimension are sorted, and then the advertising traffic data is merged sequentially based on the sorting results until the merged result is stable; otherwise, continue to merge based on the next sub-dimension.

S5. Perform traffic estimation according to the preset distributed hybrid time series estimation model and the historical traffic data within the specified unit time period under the current sub-dimension to obtain multiple sets of estimated traffic.

Wherein, the distributed hybrid time series prediction model may be a model trained based on historical advertisement data, and the model may be fused with multiple distributed time series prediction models, or may include multiple different distributed time series prediction models.

As a possible implementation, here, taking the distributed hybrid timing prediction model including multiple different distributed timing prediction models as an example, the distributed hybrid timing prediction model according to the preset described in S5 and The process of obtaining multiple groups of estimated traffic by performing traffic estimation on the historical traffic data within the specified unit time under the current sub-dimension is introduced. This process may include: based on the historical traffic data within the specified unit time under the current sub-dimension, respectively use Each distributed time series prediction model performs traffic prediction to obtain multiple sets of estimated traffic corresponding to each distributed time series prediction model.

Optionally, the distributed time series prediction model can be, but not limited to, the first-order exponential smoothing prediction model, Holt-Winters model, autoregressive model, etc. The model model is briefly explained.

(1) First-order exponential smoothing forecasting model: suitable for data forecasting with no obvious trend or seasonal change in time series. When there is an obvious trend change in the time series, the time series can be differentiated multiple times to remove the changing trend, and then the first-order exponential smoothing forecasting model is used for forecasting. Wherein, the first-order exponential smoothing prediction model can be shown in Formula 1.

s _i ＝α★x _i +(1-α)★s _i- 1Formula 1

Among them, s _i represents the smoothed value on time step i (such as the estimated value of advertising traffic), _xi is the time data on this time step (such as historical traffic data), and α is between 0 and 1 An arbitrary value that controls the balance between old and new information such as forecasts and historical traffic data.

(2) Holt-Winters model: suitable for time series of linear growth trends with seasonal effects, which can be divided into multiplicative, additive and non-seasonal models. The additive models in Equation 2 to Equation 5 below may be used in this application.

L _t ＝α(y _t -s _ts +(1-α)(L _t-1 +b _t-1 ) Formula 2

b _t =β(L _t- -L _t-1 )+(1-β)b _t-1Formula 3

s _t =γ(y _t -L _t )+(1-γ)s _ts Formula 4

F _t+1 ＝L _t +b _t +S _t+1-s Formula 5

Among them, b _t represents the trend, S _t represents the season, F _t+1 represents the predicted value, and α, β, γ are constants.

(3) Autoregressive model: It is suitable for describing the relationship between the current value and the historical value, and using the historical time data of the variable itself to predict itself. As shown in Equation 6, it is a p-order autoregressive model.

Among them, y _t is the current value, μ is a constant term, p is the order, _ri is the autocorrelation coefficient, _ri is greater than or equal to 0.5, and ∈ _t is the error.

Further, as a possible implementation, before performing the advertising traffic estimation step described in S4, denoising can be performed on the time series of the merged historical traffic data, so as to improve the accuracy of traffic estimation results . The actual implementation process may include: based on the preset Gaussian smoothing model, denoising the historical traffic data within the specified unit time period under the current sub-dimension, and based on the denoised historical traffic data within the specified unit time period, execute S4 The step of performing traffic forecasting according to the preset distributed hybrid time series forecasting model and the historical traffic data within the specified unit time period under the current sub-dimension to obtain multiple sets of estimated traffic.

Optionally, the preset Gaussian smoothing model is used to eliminate noise in the estimated sequence to obtain a smooth estimated sequence. In this embodiment, the Gaussian smoothing model may be as shown in formula 7.

Among them, σ is the variance of the time series of historical flow data, and x is each historical flow data in the time series.

It should be noted that when performing data denoising, the aforementioned Gaussian smoothing model can be used, but not limited to.

S6. According to the loss value of each group of estimated traffic, select a group of estimated traffic meeting a preset condition as an estimated advertising traffic value.

It should be noted that the loss value of each estimated flow can be calculated by using, but not limited to, square loss function (such as least square method, etc.), absolute value loss function, logarithmic loss function (Cross Entropy Loss, Softmax Loss), among which The specific calculation process of the loss value will not be repeated here in this embodiment.

Optionally, the preset condition may be set according to actual requirements, for example, it may be a loss value or an average loss value corresponding to the estimated flow.

As an optional implementation method, if a group of estimated traffic includes an estimated traffic within a specified unit duration, correspondingly, selecting a group of estimated traffic that meets preset conditions as an estimated advertising traffic includes: selecting The estimated traffic with the smallest loss value is used as the estimated advertising traffic within a specified unit duration.

As another optional implementation method, if a set of estimated traffic includes estimated traffic within at least two specified unit durations, correspondingly, a group of estimated traffic that meets preset conditions is selected as the advertising traffic estimate Including: separately calculating the average loss value of the estimated flow within the at least two specified unit time periods in each group of estimated flow; Advertising traffic estimates.

Among them, in the aforementioned implementation method, assuming that the specified unit duration is days, and a group of estimated traffic includes estimated traffic in the next 3 days, then it is necessary to calculate the average loss value of each group of estimated traffic in these 3 days, and then Based on the average loss value corresponding to each group of estimated traffic, the estimated traffic with the smallest average loss value is selected as the estimated advertising traffic value.

The estimated advertising traffic estimated by the advertising traffic estimation method given in S1 to S6 above can be used for pre-sales inventory query to adjust the feasibility of the current advertising contract, or determine the ranking of the advertising contract. Period priority. It should be noted that the aforementioned advertisement contract may also be a bidding advertisement, which is not limited in this embodiment. For example, FIG. 4 is a schematic diagram of an advertisement scheduling interface for contract advertisements, which is used for making pre-sale advertisement schedules and the like.

Further, based on the description of the advertising traffic estimation method given in S1 to S6 above, as a possible implementation method, when the advertiser initiates an advertisement based on the selection interface of the scheduled delivery condition shown in Figure 3 displayed on the app or web page When requesting, the ad request can be forwarded to the preset ad engine in the ad management platform through SDK (Software Development Kit) or JS (Javascript), and the ad engine queries the ad index module to obtain all the ad systems ( Such as contract system, bidding system, etc.), and then obtain the estimated advertising traffic value through the traffic estimation module preset with the advertising traffic estimation method described in S1 to S6, and determine the priority of the advertising system according to the estimated advertising traffic value level, if there is an advertising system that meets the delivery requirements set in the ad request, then the ad can be served through this advertising system.

In the advertising traffic estimation method given in the embodiment of the present application, at least the following advantages are provided:

When the historical traffic data is obtained according to the target fixed investment conditions for advertising traffic estimation, the distributed hybrid time series estimation model is used for traffic estimation to obtain multiple groups of estimated traffic, and the loss value is selected from the loss values corresponding to the multiple groups of estimated traffic The preset traffic with the smallest value or the smallest average loss value is used as the advertising traffic estimation value to avoid the influence of historical data fluctuations or accidental factors on the advertising estimation results, effectively ensuring the accuracy of the advertising traffic estimation results.

Embodiment two

Fig. 5 is a block diagram showing an advertisement traffic estimation device 100 according to an exemplary embodiment, and the advertisement traffic estimation device 100 may be applied to the advertisement management platform shown in Fig. 1 . Referring to FIG. 5 , the advertisement traffic estimation device 100 includes a data acquisition module 110 , a data combination module 120 , a traffic estimation module 130 and a prediction result determination module 140 .

The data acquisition module 110 is configured to acquire historical traffic data produced within a preset time period according to the target fixed investment condition; optionally, the historical traffic data includes at least one of historical click log data, historical display log data, historical request log data or Various.

The data merging module 120 is configured to select sub-dimensions based on the order of each sub-dimension in the designated merging dimension, and merge the historical traffic data based on the selected current sub-dimension;

The traffic estimation module 130 is configured to: if the data volume difference of the historical traffic data in every two specified unit time periods combined based on the current sub-dimensions is less than the preset value, then according to the preset distributed hybrid time series prediction model and the historical traffic data within the specified unit time period under the current sub-dimension to perform traffic estimation to obtain multiple sets of estimated traffic; otherwise, return to the data merging module 120 to enter the next sub-dimension.

Optionally, when the distributed hybrid time-series prediction model includes multiple different distributed time-series prediction models, the traffic prediction module 130 can be specifically configured to, based on the historical traffic data within a specified unit duration under the current sub-dimension, respectively Each distributed time series prediction model is used to perform traffic prediction, and multiple sets of estimated traffic corresponding to each distributed time series prediction model are obtained.

The estimation result determination module 140 is configured to select a group of estimated traffic meeting preset conditions as an estimated advertising traffic value according to the loss value of each group of estimated traffic.

Optionally, if a set of estimated traffic includes an estimated traffic within a specified unit duration, correspondingly, the estimated result determination module 140 may be configured to select the estimated traffic with the smallest loss value as an advertisement within a specified unit duration traffic estimates; or,

If a group of estimated flows includes estimated flows within at least two specified unit time periods, correspondingly, the estimated result determination module 140 can also be used to separately calculate The average loss value of the estimated traffic; select a group of estimated traffic with the smallest average loss value as the estimated advertising traffic within the corresponding number of specified unit durations.

Further, as a possible implementation, the advertising traffic estimation device may also include a module for acquiring the conditions of fixed casting, and the module of the conditions of fixed casting includes:

The interface display unit is used to display the fixed investment condition selection interface, and there are multiple fixed investment condition selection tabs displayed on the fixed investment condition selection interface.

The fixed investment condition determining unit is used to respond to the selected operation of the fixed investment condition initiated by the advertiser based on multiple fixed investment condition tabs, and determine the target fixed investment condition according to the selection result.

Further, as yet another possible implementation, the device may further include:

The data denoising module is used to denoise the historical traffic data within the specified unit time period under the current sub-dimension based on the preset Gaussian smoothing model; and call based on the historical traffic data within the specified unit time period after denoising The traffic estimation module 130 performs traffic estimation based on the preset distributed hybrid time series estimation model and the historical traffic data within the specified unit time period under the current sub-dimension to obtain multiple sets of estimated traffic, otherwise, enter the step of the next sub-dimension .

With regard to the advertising traffic estimation apparatus 100 in this embodiment, the specific manner in which each module executes operations has been described in detail in the embodiment related to the method, and will not be described in detail here.

Embodiment three

Please refer to FIG. 6 , which is a block diagram of an electronic device 10 according to an exemplary embodiment. The electronic device 10 may at least include a processor 11 and a memory 12 for storing instructions executable by the processor 11 . Wherein, the processor 11 is configured to execute instructions to implement all or part of the steps of the information interaction method in the foregoing embodiments.

The processor 11 and the memory 12 are directly or indirectly electrically connected to realize data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines.

Wherein, the processor 11 is used for reading/writing data or programs stored in the memory, and executing corresponding functions.

The memory 12 is used for storing programs or data, such as storing instructions executable by the processor 11 . This memory 12 can be, but not limited to, random access memory (Random Access Memory, RAM), read-only memory (Read Only Memory, ROM), programmable read-only memory (Programmable Read-Only Memory, PROM), erasable Read-only memory (Erasable Programmable Read-Only Memory, EPROM), Electric Erasable Programmable Read-Only Memory (EEPROM), etc.

Further, as a possible implementation manner, the electronic device 10 may further include a power supply component, a multimedia component, an audio component, an input/output (I/O) interface, a sensor component, a communication component, and the like.

The power supply component provides power to various components of the electronic device 10 . Power components may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 10 .

The multimedia component includes a screen that provides an output interface between the electronic device 10 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensor may not only sense a boundary of a touch or a swipe action, but also detect duration and pressure associated with the touch or swipe operation. In some embodiments, the multimedia component includes a front camera and/or a rear camera. When the electronic device 10 is in an operation mode, such as a shooting mode or a video mode, the front camera and/or the rear camera can receive external multimedia data. Each front camera and rear camera can be a fixed optical lens system or have focal length and optical zoom capability.

The audio component is configured to output and/or input audio signals. For example, the audio component includes a microphone (MIC), which is configured to receive external audio signals when the electronic device 10 is in operation modes, such as call mode, recording mode and voice recognition mode. The received audio signal may be further stored in memory 12 or sent via a communication component. In some embodiments, the audio component further includes a speaker for outputting audio signals.

The I/O interface provides an interface between the processing component and the peripheral interface module, and the above peripheral interface module can be a keyboard, a click wheel, a button, and the like. These buttons may include, but are not limited to: a home button, volume buttons, start button, and lock button.

The sensor assembly includes one or more sensors for providing various aspects of status assessment for electronic device 10 . For example, the sensor component can detect the open/closed state of the electronic device 10, the relative positioning of components, such as the display and keypad of the electronic device 10, and the sensor component can also detect the position change of the electronic device 10 or a component of the electronic device 10 , the existence or non-existence of the user's contact with the electronic device 10 , the orientation or acceleration/deceleration of the electronic device 10 and the temperature change of the electronic device 10 . The sensor assembly may include a proximity sensor configured to detect the presence of nearby objects in the absence of any physical contact. The sensor assembly may also include optical sensors, such as CMOS or CCD image sensors, for use in imaging applications. In some embodiments, the sensor assembly may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor or a temperature sensor.

The communication component is configured to facilitate wired or wireless communication between electronic device 10 and other devices. The electronic device 10 can access a wireless network based on a communication standard, such as WiFi, an operator network (such as 2G, 3G, 4G or 5G), or a combination thereof. In one exemplary embodiment, the communication component receives a broadcast signal or broadcast related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, the communication assembly also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, Infrared Data Association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

In an exemplary embodiment, electronic device 10 may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable A programmable gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation for performing the methods described above.

It should be understood that the structure shown in FIG. 6 is only a schematic structural diagram of the electronic device 10, and the electronic device 10 may also include more or fewer components than those shown in FIG. configuration. Each component shown in FIG. 6 may be implemented by hardware, software or a combination thereof.

Embodiment four

In an exemplary embodiment, there is also provided a non-transitory computer-readable storage medium including instructions, such as the memory 12 including instructions, which can be executed by the processor 11 of the electronic device 10 to complete the above page processing method. For example, the non-transitory computer readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like.

Those skilled in the art should understand that the embodiments of the present application may be provided as methods, systems, or computer program products. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

It should be noted that the term "comprises", "comprises" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus comprising a set of elements includes not only those elements, but also includes none other elements specifically listed, or also include elements inherent in the process, method, commodity, or apparatus. Without further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising said element.

Those skilled in the art should understand that the embodiments of the present application may be provided as methods, systems or computer program products. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The above descriptions are only examples of the present application, and are not intended to limit the present application. For those skilled in the art, various modifications and changes may occur in this application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims (10)

1. A method for estimating advertising traffic, comprising: Obtain the historical flow data produced within the preset time period according to the target fixed investment conditions; Sort the sub-dimensions involved in the specified merged dimension in advance, and select sub-dimensions based on the order of the sub-dimensions in the specified merged dimension to perform the following operations: Merging the historical traffic data based on the selected current sub-dimension; If the data volume difference between the historical traffic data within two specified units of time combined based on the current sub-dimension is less than the preset value, then according to the preset distributed hybrid time series prediction model and the specified unit under the current sub-dimension The historical traffic data within the time period is used for traffic estimation to obtain multiple sets of estimated traffic, otherwise return to the next sub-dimension; According to the loss value of each group of estimated traffic, a group of estimated traffic meeting preset conditions is selected as an estimated advertising traffic value. 2. The advertising traffic estimation method according to claim 1, wherein the distributed hybrid timing estimation model comprises a plurality of different distributed timing estimation models; and According to the preset distributed hybrid time series prediction model and the historical traffic data within the specified unit time period under the current sub-dimension, the traffic is estimated to obtain multiple sets of estimated traffic, including: Based on the historical traffic data within the specified unit time period under the current sub-dimension, each distributed time series prediction model is used to perform traffic prediction respectively, and multiple sets of estimated traffic corresponding to each distributed time series prediction model are obtained. 3. The advertisement traffic estimation method according to claim 1, characterized in that, If a group of estimated traffic includes an estimated traffic within a specified unit duration, correspondingly, selecting a group of estimated traffic that meets the preset conditions as an estimated advertising traffic includes: selecting the estimated traffic with the smallest loss value as a The estimated advertising traffic within the specified unit duration; If a group of estimated traffic includes estimated traffic within at least two specified unit durations, correspondingly, selecting a group of estimated traffic that meets the preset conditions as an estimated advertising traffic includes: separately calculating each group of estimated traffic The average loss value of the estimated traffic within the at least two specified unit durations; selecting a group of estimated traffic with the smallest average loss value as the estimated advertising traffic within the at least two specified unit durations. 4. The advertising traffic estimation method according to claim 1, wherein the step of obtaining the target fixed delivery conditions comprises: Displaying the selection interface of the fixed investment conditions, the selection interface of the fixed investment conditions shows that there are multiple fixed investment condition tabs; In response to an advertiser's selection of fixed investment conditions based on the plurality of fixed investment condition tabs, the target fixed investment conditions are determined according to the selection result. 5. The advertising traffic estimation method according to claim 1, further comprising: Before performing traffic forecasting according to the preset distributed hybrid time series forecasting model and the historical traffic data within the specified unit time under the current sub-dimension to obtain multiple sets of estimated traffic, the current sub-dimension is analyzed based on the preset Gaussian smoothing model. The historical traffic data within the specified unit time period under the dimension is denoised. 6 . The advertising traffic estimation method according to claim 1 , wherein the historical traffic data includes at least one or more of historical click log data, historical display log data, and historical request log data. 7. An advertising traffic estimation device, characterized in that it comprises: The data acquisition module is used to acquire the historical traffic data produced within the preset time period according to the target fixed investment conditions; A data merging module, configured to sort the sub-dimensions involved in the specified merging dimension in advance, select a sub-dimension based on the order of the sub-dimensions in the specified merging dimension, and merge the historical traffic data based on the selected current sub-dimension; The traffic estimation module is used to estimate the data according to the preset distributed hybrid timing model and The historical traffic data within the specified unit time period under the current sub-dimension is estimated to obtain multiple sets of estimated traffic, otherwise return to the data merging module and enter the next sub-dimension; The estimation result determination module is configured to select a group of estimated traffic meeting preset conditions as an estimated advertising traffic value according to the loss value of each group of estimated traffic. 8. The advertising traffic estimation device according to claim 7, wherein the distributed hybrid timing estimation model includes a plurality of different distributed timing estimation models, and the traffic estimation module is also used for For the historical traffic data within the specified unit time period under the current sub-dimension, each distributed time series prediction model is used to perform traffic prediction respectively, and multiple sets of estimated traffic corresponding to each distributed time series prediction model are obtained. 9. An electronic device, characterized in that it comprises: processor; memory for storing said processor-executable instructions; Wherein, the processor is configured to execute the instructions, so as to realize the advertising traffic estimation method according to any one of claims 1-6. 10. A computer-readable storage medium, characterized in that when the instructions in the storage medium are executed by the processor in the electronic device, the electronic device can execute the advertisement according to any one of claims 1-6 Flow Estimation Method.

Images