Data processing method, device, electronic device and storage medium

Abstract

The present disclosure relates to a data processing method, device, electronic device and storage medium. After obtaining the delivery data of each object to be delivered, if the delivery data of the object to be delivered does not meet the preset conditions, the delivery data of the object to be delivered is numerically corrected to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty; according to the target delivery data of the object to be delivered, the viewing probability of the object to be delivered is determined, and according to the viewing probability of each object to be delivered, the target object is determined from all the objects to be delivered. By numerically correcting the delivery data of the object to be delivered, the viewing probability of the object to be delivered whose delivery data value is empty can be determined. In the process of determining the target object based on the viewing probability, the object to be delivered whose delivery data value is empty will be used as a candidate object, so that the object to be delivered whose delivery data value is empty has the possibility of being used as the target object.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to a data processing method, device, electronic device and storage medium. Background Art

[0002] In related technologies, with the rapid development of the Internet, the objects to be delivered can be displayed with the help of the Internet, and the corresponding delivery scenarios of the objects to be delivered have also developed from one delivery scenario to multiple delivery scenarios. For example, advertisements can be delivered in different delivery scenarios.

[0003] For different delivery scenarios, not only are there differences in user distribution, but also in the styles of delivery objects. It is necessary to determine the target objects that can be delivered to the Internet from the objects to be delivered. For example, if the target object is determined based on the historical delivery data of the objects to be delivered in all delivery scenarios, if the value of the delivery data of the object to be delivered is empty, it cannot be determined as the target object through relevant technologies. Summary of the invention

[0004] The present disclosure provides a data processing method, device, electronic device and storage medium to at least solve the problem in the related art that when the value of the delivery data of the delivery object is empty, it cannot be determined as the target object. The technical solution of the present disclosure is as follows:

[0005] According to a first aspect of an embodiment of the present disclosure, there is provided a data processing method, including:

[0006] Obtain the delivery data of each object to be delivered;

[0007] If the delivery data of the object to be delivered does not meet the preset conditions, the delivery data of the object to be delivered is numerically corrected to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty;

[0008] Determining the viewing probability of the object to be delivered according to the target delivery data of the object to be delivered, wherein the viewing probability is used to indicate the probability of the object to be delivered being viewed by the user from delivery;

[0009] According to the viewing probability of each object to be delivered, the target object is determined from all the objects to be delivered.

[0010] Optionally, determining the viewing probability of the object to be delivered according to the target delivery data of the object to be delivered includes:

[0011] According to the distribution of the target delivery data of the object to be delivered, the current delivery trend of the object to be delivered is estimated to obtain the estimation result of the object to be delivered;

[0012] The viewing probability of the object to be delivered is determined according to the estimation result of the object to be delivered.

[0013] Optionally, the estimation result of the object to be delivered includes at least one of a target cover click-through rate, a target video click-through rate, and a target conversion rate;

[0014] The target cover click rate estimation process includes: constructing a cover distribution function of the target cover click amount and the target cover exposure amount according to the target cover click amount in the target delivery data and the target cover exposure amount in the target delivery data; determining the target cover click rate of the object to be delivered according to the cover distribution function;

[0015] The target video click rate estimation process includes: constructing a video distribution function of the target video click amount and the target video exposure amount according to the target video click amount in the target delivery data and the target video exposure amount in the target delivery data; determining the target video click rate of the object to be delivered according to the video distribution function;

[0016] The target conversion rate estimation process includes: constructing a conversion distribution function of the target conversion number and the target video click volume according to the target conversion number and the target video click volume in the target delivery data; and determining the target conversion rate of the object to be delivered according to the conversion distribution function.

[0017] Optionally, the method further includes: if the delivery data of the object to be delivered meets a preset condition, determining the viewing probability of the object to be delivered according to the delivery data of the object to be delivered.

[0018] Optionally, the method further includes: acquiring a first number of target objects determined according to the viewing probability obtained from the target delivery data and a second number of target objects determined according to the viewing probability obtained from the delivery data;

[0019] If the quantitative relationship between the first quantity and the second quantity does not satisfy a preset quantitative relationship, the preset condition is adjusted.

[0020] Optionally, determining the target object from all the objects to be delivered according to the viewing probability of each object to be delivered includes:

[0021] According to the viewing probability obtained through the target delivery data and the first preset number, determining a first preset number of target objects from the objects to be delivered whose viewing probability is determined according to the target delivery data;

[0022] According to the viewing probability obtained through the delivery data and the second preset number, a second preset number of target objects are determined from the objects to be delivered whose viewing probability is determined according to the delivery data.

[0023] Optionally, the delivery data includes: cover clicks, cover exposures, video clicks, video exposures and conversions;

[0024] The preset conditions include: the cover click volume, cover exposure volume, video click volume, video exposure volume and conversion number are respectively greater than their respective thresholds.

[0025] Optionally, obtaining the delivery data of each to-be-delivered object includes:

[0026] Obtain the real-time delivery data of each object to be delivered in the current time period;

[0027] Obtain the historical delivery data of each object to be delivered;

[0028] The historical delivery data of each object to be delivered and the real-time delivery data of each object to be delivered are aggregated to obtain the delivery data of each object to be delivered.

[0029] Optionally, the method further includes: dividing the delivery data of each to-be-delivered object according to the delivery scenario to obtain a delivery data set corresponding to each delivery scenario;

[0030] Determining the target object from all the objects to be delivered according to the viewing probability of each object to be delivered includes: determining the target object belonging to each delivery scene according to the viewing probability of each object to be delivered in each delivery data set.

[0031] According to a second aspect of an embodiment of the present disclosure, there is provided a data processing device, including:

[0032] An acquisition unit is configured to acquire the delivery data of each object to be delivered;

[0033] A correction unit is configured to perform a numerical correction on the delivery data of the object to be delivered if the delivery data of the object to be delivered does not meet a preset condition, so as to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty;

[0034] A first determining unit is configured to determine the viewing probability of the object to be delivered according to the target delivery data of the object to be delivered, wherein the viewing probability is used to indicate the probability of the object to be delivered being viewed by the user from delivery;

[0035] The second determining unit is configured to determine the target object from all the objects to be delivered according to the viewing probability of each object to be delivered.

[0036] Optionally, the first determination unit is configured to estimate the current delivery trend of the object to be delivered according to the distribution of the target delivery data of the object to be delivered, and obtain the estimated result of the object to be delivered; and determine the viewing probability of the object to be delivered according to the estimated result of the object to be delivered.

[0037] Optionally, the estimation result of the object to be delivered includes at least one of a target cover click-through rate, a target video click-through rate, and a target conversion rate;

[0038] The first determining unit is configured to construct a cover distribution function of the target cover click volume and the target cover exposure volume according to the target cover click volume in the target delivery data and the target cover exposure volume in the target delivery data; and determine the target cover click rate of the to-be-delivered object according to the cover distribution function;

[0039] The first determining unit is configured to construct a video distribution function of the target video click volume and the target video exposure volume according to the target video click volume in the target delivery data and the target video exposure volume in the target delivery data; and determine the target video click rate of the object to be delivered according to the video distribution function;

[0040] The first determination unit is configured to execute a conversion distribution function of the target conversion number and the target video click volume according to the target conversion number and the target video click volume in the target delivery data; and determine the target conversion rate of the object to be delivered according to the conversion distribution function.

[0041] Optionally, the device further includes: a third determining unit configured to determine the viewing probability of the object to be delivered according to the delivery data of the object to be delivered if the delivery data of the object to be delivered meets a preset condition.

[0042] Optionally, the acquisition unit is configured to further acquire a first number of target objects determined according to the viewing probability obtained from the target delivery data and a second number of target objects determined according to the viewing probability obtained from the delivery data;

[0043] The device further includes: an adjustment unit configured to adjust the preset condition if the quantity relationship between the first quantity and the second quantity does not satisfy a preset quantity relationship.

[0044] Optionally, the second determination unit is configured to determine a first preset number of target objects from the objects to be delivered that have obtained the target delivery data based on the viewing probability and the first preset number obtained through the target delivery data; and to determine a second preset number of target objects from the objects to be delivered that have obtained the delivery data based on the viewing probability and the second preset number obtained through the delivery data.

[0045] Optionally, the delivery data includes: cover clicks, cover exposures, video clicks, video exposures and conversions;

[0046] The preset conditions include: the cover click volume, cover exposure volume, video click volume, video exposure volume and conversion number are respectively greater than their respective thresholds.

[0047] Optionally, the acquisition unit is configured to execute acquisition of real-time delivery data of each object to be delivered in the current time period, and acquisition of historical delivery data of each object to be delivered; aggregate the historical delivery data of each object to be delivered and the real-time delivery data of each object to be delivered to obtain the delivery data of each object to be delivered.

[0048] Optionally, the device further comprises: a division unit configured to divide the delivery data of each to-be-delivered object according to the delivery scenario to obtain a delivery data set corresponding to each delivery scenario;

[0049] The second determination unit is configured to determine the target object belonging to each delivery scene according to the viewing probability of each to-be-delivered object in each delivery data set.

[0050] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device, including: a processor;

[0051] a memory for storing instructions executable by the processor;

[0052] The processor is configured to execute the instructions to implement the above data processing method.

[0053] According to a fourth aspect of an embodiment of the present disclosure, a storage medium is provided, in which instructions are stored. When the instructions in the storage medium are executed by a processor, the processor is enabled to execute the above-mentioned data processing method.

[0054] According to a fifth aspect of an embodiment of the present disclosure, there is provided a computer program product which, when executed on an electronic device, is suitable for executing a program initialized with the steps of the above-mentioned image processing method.

[0055] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:

[0056] After obtaining the delivery data of each object to be delivered, if the delivery data of the object to be delivered does not meet the preset conditions, the delivery data of the object to be delivered is numerically corrected to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty; according to the target delivery data of the object to be delivered, the viewing probability of the object to be delivered is determined, and according to the viewing probability of each object to be delivered, the target object is determined from all the objects to be delivered. For the object to be delivered, the delivery data of the object to be delivered can be numerically corrected so that the value of the target delivery data of the object to be delivered is not empty, so that the viewing probability of the object to be delivered with an empty delivery data value can be determined, so that in the process of determining the target object based on the viewing probability, the object to be delivered with an empty delivery data value will be used as a candidate object, so that the object to be delivered with an empty delivery data value has the possibility of being used as a target object, for example, for a currently generated object to be delivered that has not been used as a target object or even has not been viewed, its delivery data will have an empty value, and by numerically correcting the delivery data, this type of object to be delivered may be used as a target object, so that its viewing probability can be determined according to the corrected target delivery data, thereby improving the accuracy of the target object determination.

[0057] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.

[0059] Figure 1 is a flow chart of a data processing method according to an exemplary embodiment;

[0060] Figure 2 is a flow chart of another data processing method according to an exemplary embodiment;

[0061] Figure 3 is a flow chart of yet another data processing method according to an exemplary embodiment;

[0062] Figure 4 is a block diagram of a data processing device according to an exemplary embodiment;

[0063] Figure 5 is a block diagram of another data processing device according to an exemplary embodiment. DETAILED DESCRIPTION

[0064] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.

[0065] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0066] Figure 1 is a flow chart of a data processing method according to an exemplary embodiment, which may include the following steps:

[0067] In step S11, the delivery data of each object to be delivered is obtained. The delivery data can indicate the viewing situation of the object to be delivered in the process of delivery to the display. For example, for advertisements, advertisements can be delivered to different display objects such as browsers and software, and the advertisements are displayed to users through these display objects. Users can watch the video of the advertisements by clicking and other operations. Therefore, delivery data can be generated during the process of delivery of advertisements to display objects and user viewing, so as to indicate the viewing situation of the object to be delivered in the process of delivery to the display through the delivery data.

[0068] The delivery data can describe the delivery object from different angles. For example, the delivery data includes but is not limited to: cover exposure, cover clicks, video exposure, video clicks and conversions. Cover exposure and cover clicks describe the use of the cover of the delivery object. The cover is an overview of the delivery object, similar to the abstract of an article. Cover exposure indicates the number of times the cover of the delivery object is displayed, and cover clicks indicate the number of times the cover of the delivery object is clicked by users. Video exposure and video clicks describe the use of the video of the delivery object. Compared with the cover, the video is a detailed description of the delivery object. The amount of light indicates the number of times the video of the object to be delivered is displayed, and the number of video clicks indicates the number of times the video of the object to be delivered is clicked by users; the number of conversions indicates the number of times the object to be delivered is displayed to the number of times the APP (application) of the object to be delivered is installed. The conversion rate can be obtained through the conversion number and the video clicks. The conversion rate is used to indicate the probability of the object to be delivered being displayed to the installation of the APP (application) of the object to be delivered. For example, the object to be delivered is a promotional advertisement for an application, and the conversion rate indicates the probability of the promotional advertisement being delivered to the application it points to being installed in the device by the user. These data describing the object to be delivered from different aspects can be recorded in one delivery data.

[0069] In this embodiment, one way of obtaining the delivery data of the object to be delivered in step S11 is: obtaining the real-time delivery data of the object to be delivered in the current time period. The current time period can be determined as the end time point of the target object, and the preset time is moved forward to obtain the real-time delivery data. The value of the preset time is not limited in this embodiment, for example, the real-time delivery data of the day is obtained.

[0070] Real-time delivery data can be collected through but not limited to Kafka messages. By obtaining the real-time delivery data of the object to be delivered, adjustments can be made according to the real-time delivery situation of the object to be delivered, thereby realizing real-time adjustment of the target object. Compared with adjusting the target object according to historical delivery data, the adjustment of the target object is more timely. Compared with historical delivery data, real-time delivery data can be the delivery data of the object to be delivered generated in the current time period, so that the latest delivery data of the object to be delivered can be obtained in time, so as to timely discover and make these types of objects to be delivered have the possibility of being target objects.

[0071] In this embodiment, another way of obtaining the delivery data of the object to be delivered in step S11 is: obtaining the historical delivery data of the object to be delivered, where the historical delivery data is delivery data at a time interval from the current time, and the time interval is not limited in this embodiment, such as the historical delivery data can be delivery data within 7 days from the current time.

[0072] Historical delivery data can be obtained through but not limited to scheduled tasks. Scheduled tasks correspond to operating cycles, so that historical delivery data is obtained once every operating cycle. The stability of historical delivery data is greater than that of real-time delivery data. Therefore, the stability of the target object obtained through historical delivery data can be improved.

[0073] In this embodiment, another way of obtaining the delivery data of the objects to be delivered in step S11 is: obtaining the real-time delivery data of each object to be delivered in the current time period; obtaining the historical delivery data of each object to be delivered; aggregating the historical delivery data of each object to be delivered and the real-time delivery data of each object to be delivered to obtain the delivery data of each object to be delivered.

[0074] Aggregation is used to merge the real-time delivery data and historical delivery data of the same delivery object, so that the delivery data of the delivery object can take into account both real-time and stability. Aggregation is used to make one delivery object have one data set, thus preventing the problem of contradictory results from calculations using real-time delivery data and historical delivery data respectively.

[0075] For example, this embodiment can obtain the real-time delivery data of each object to be delivered through but not limited to Kafka messages and obtain the historical delivery data of each object to be delivered through but not limited to scheduled tasks. The acquisition of real-time delivery data and historical delivery data can be simultaneous or sequential, and this embodiment does not limit the sequence of the two. After obtaining the real-time delivery data and historical delivery data, the two are aggregated to merge the real-time delivery data and historical delivery data of the same object to be delivered.

[0076] In step S12, if the delivery data of the object to be delivered does not meet the preset conditions, the delivery data of the object to be delivered is numerically corrected to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty.

[0077] If the delivery data of the object to be delivered does not meet the preset conditions, it means that the number of times the object to be delivered has been exposed is small or even has not been exposed, so it is necessary to make a numerical correction so that the value of the target delivery data is not empty. The delivery data of the object to be delivered includes: cover clicks, cover exposure, video clicks, video exposure and conversions. The corresponding preset conditions include: cover clicks, cover exposure, video clicks, video exposure and conversions are respectively greater than their respective thresholds, that is, a threshold is set for each field in the delivery data. If the value of at least one field in the delivery data is less than or equal to the corresponding threshold, it is determined that the delivery data of the object to be delivered does not meet the preset conditions. If the value of each field in the delivery data is greater than the corresponding threshold, it is determined that the delivery data of the object to be delivered meets the preset conditions. The threshold corresponding to each field is not limited in this embodiment.

[0078] Taking the cover exposure as an example, the threshold corresponding to the cover exposure is 10,000 times. If the cover exposure of the object to be delivered is less than or equal to 10,000 times, it is determined that the object to be delivered does not meet the preset conditions. If the cover exposure of the object to be delivered is greater than 10,000 times, it is necessary to further determine whether the values ​​of other fields are greater than the corresponding thresholds. If the values ​​of each field are greater than the corresponding thresholds, it is determined that the delivery data of the object to be delivered meets the preset conditions. Of course, the preset conditions in this embodiment can also be adopted in other ways, such as the value of at least one field among the above-mentioned cover clicks, cover exposures, video clicks, video exposures and conversions is greater than the corresponding threshold, which is not limited in this embodiment.

[0079] For a currently generated object to be delivered, its delivery data may have a null value. For some objects to be delivered that have been generated but have not been used as target objects or even have not been viewed, their delivery data may also have a null value. For example, if the value of at least one of the fields mentioned above: cover exposure, cover clicks, video exposure, video clicks and conversions is null, it may cause an error in the process of obtaining the viewing probability. For this reason, this embodiment can perform numerical correction on the delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not null.

[0080] The numerical correction method includes but is not limited to: adding the fields with empty values ​​in the delivery data of the object to be delivered to the preset values. For different fields, the added preset values ​​may be the same or different. For example, an addition operation may be performed for the fields with empty values. Alternatively, each field in the delivery data of the object to be delivered is added to the preset values. Similarly, for different fields, the added preset values ​​may be the same or different. For example, an addition operation may be performed for each field. Compared with only adding the fields with empty values ​​to the preset values, the step of filtering out the delivery data with empty values ​​is omitted, thereby speeding up the numerical correction process of the delivery data.

[0081] In step S13, the viewing probability of the object to be delivered is determined according to the target delivery data of the object to be delivered, wherein the viewing probability is used to indicate the probability of the object to be delivered being viewed by the user.

[0082] It can be understood that: the greater the viewing probability, the more times the object to be delivered is viewed, and each viewing can bring a profit. Therefore, the greater the viewing probability, the higher the profit, and the correspondingly greater the possibility of the object to be delivered as the target object. Conversely, the smaller the possibility of the object to be delivered as the target object. Therefore, the viewing probability is used as a reference condition for determining the target object. The higher the accuracy of the viewing probability, the higher the accuracy of the target object.

[0083] For example, one way to calculate viewing probability is but not limited to: viewing probability = target cover click-through rate * target video click-through rate * target conversion rate. The target cover click-through rate, target video click-through rate and target conversion rate can be obtained through target delivery data. Therefore, the accuracy of the target cover click-through rate, target video click-through rate and target conversion rate affects the accuracy of the viewing probability.

[0084] The target delivery data includes target cover exposure, target cover clicks, target video exposure, target video clicks and target conversions. In the related art, the target cover click-through rate, target video click-through rate and target conversion rate are calculated as follows: target cover click-through rate = target cover clicks / target cover exposure; target video click-through rate = target video clicks / target video exposure, target conversion rate = target conversions / target video clicks. However, this method is obtained by calculating the ratio between the two parameters, and the higher the ratio does not mean more delivery. For example, the target cover clicks and the target cover exposure are both 1, and the target cover click-through rate is 100%; the target cover clicks are 100 and the target cover exposure is 200, and the target cover click-through rate is 50%. However, the corresponding number of delivery of the object to be delivered is more than the number of delivery of the object to be delivered with a target cover click-through rate of 100%.

[0085] Therefore, for the target delivery data, this embodiment uses the method of estimating the current delivery trend of the delivery object to obtain the viewing probability. The current delivery trend of the delivery object reflects the development trend of the delivery object over time and can more objectively reflect the delivery object.

[0086] For example, if the cover click count of an object to be delivered 1 is 1 and the cover exposure is 1, judging from the specific values ​​alone, the cover of the object to be delivered 1 is exposed once and clicked once, and the cover click rate is 100%; if the cover click count of an object to be delivered 2 is 10 and the cover exposure is 100, judging from the specific values, the cover click rate of the object to be delivered 2 is 10%. Although the cover click rate of the object to be delivered 2 is less than that of the object to be delivered 1, the cover click rate of the object to be delivered 1 is only the click rate of a time period. As time goes by, the cover click rate of the object to be delivered 1 may be less than the cover click rate of the object to be delivered 2. Therefore, simply comparing the ratio between the two parameters will reduce objectivity and accuracy. For this reason, this embodiment can obtain the viewing probability by estimating the current delivery trend of the object to be delivered. Correspondingly, a feasible way to determine the viewing probability of the object to be delivered based on the target delivery data of the object to be delivered is as follows:

[0087] According to the distribution of the target delivery data of the object to be delivered, the current delivery trend of the object to be delivered is estimated to obtain the estimated result of the object to be delivered; according to the estimated result of the object to be delivered, the viewing probability of the object to be delivered is determined.

[0088] The estimated result of the object to be delivered includes at least one of a target cover click-through rate, a target video click-through rate, and a target conversion rate. The estimation process of the target cover click-through rate, the target video click-through rate, and the target conversion rate is as follows:

[0089] The estimation process of the target cover click rate includes: constructing a cover distribution function of the target cover clicks and the target cover exposures according to the target cover clicks in the target delivery data and the target cover exposures in the target delivery data; and determining the target cover click rate of the object to be delivered according to the cover distribution function. For example, a distribution sampling function corresponding to the target cover clicks and the target cover exposures is pre-set, such as a Beta distribution sampling function as the distribution sampling function corresponding to the target cover clicks and the target cover exposures, and the Beta distribution sampling function corresponding to the target cover clicks and the target cover exposures is constructed with one of the target cover clicks and the target cover exposures as the horizontal coordinate and the other as the vertical coordinate, and then the target cover click rate of the object to be delivered is obtained by the Beta distribution sampling function, such as target cover click rate = f (target cover clicks, target cover exposures), and f() can be but is not limited to the Beta distribution sampling function.

[0090] The process of estimating the target video click rate includes: constructing a video distribution function of the target video clicks and the target video exposures according to the target video clicks in the target delivery data and the target video exposures in the target delivery data; and determining the target video click rate of the object to be delivered according to the video distribution function. The process of estimating the target video click rate is similar to the process of estimating the target cover click rate, such as target video click rate = f (target video clicks, target video exposures), and the detailed process will not be repeated here.

[0091] The target conversion rate estimation process includes: constructing the conversion distribution function of the target conversion number and the target video click volume according to the target conversion number and the target video click volume in the target delivery data; and determining the target conversion rate of the object to be delivered according to the conversion distribution function. The target conversion rate estimation process is similar to the target cover click rate estimation process, target conversion rate = f (target conversion number, target video click volume), f () can be but not limited to the Beta distribution sampling function, and the detailed process is not repeated here.

[0092] In addition to considering the above-mentioned target cover click-through rate, target video click-through rate and target conversion rate in the process of determining the viewing probability, the influence of the unit bid of the display object where the object to be placed is located can also be considered, because the unit bid will also affect the revenue. For example, if the number of views is higher but the unit bid is lower, its revenue may be less than the revenue of the object to be placed with a high unit bid but a small number of views. Therefore, in the process of determining the viewing probability, the unit bid can be comprehensively considered, and the realization efficiency of the object to be placed is used as a form of expression of the viewing probability. In this embodiment, a calculation method for the realization efficiency of the object to be placed is: realization efficiency = target cover click-through rate * target video click-through rate * target conversion rate * unit bid. The unit bid can be obtained in the process of obtaining the delivery data. If the real-time delivery data of the object to be placed is obtained, the unit bid is the unit bid corresponding to the real-time delivery data. In this way, because the unit bid will change with the delivery of the target object, the unit bid corresponding to the real-time delivery data reflects the unit price of the object to be placed in the current time period, so the realization efficiency is determined by the unit bid corresponding to the real-time delivery data to be consistent with the current bid of the object to be placed.

[0093] In actual application, after obtaining the delivery data of each object to be delivered, this embodiment can correct the delivery data of each object to be delivered, and use the corrected target delivery data to determine the viewing probability, so that each object to be delivered can be estimated based on its current delivery trend to obtain the estimated result of the object to be delivered; according to the estimated result of the object to be delivered, the viewing probability of the object to be delivered is determined. For the specific process, please refer to the above process.

[0094] In step S14, the target object is determined from all the objects to be delivered according to the viewing probability of each object to be delivered.

[0095] The target object is an object to be placed that can be outputted and selected from all the objects to be placed, such as an object to be placed that is selected for display in the display object. Since the number of objects to be placed that can be outputted is limited, this embodiment can sort the objects to be placed according to the viewing probability of each object to be placed, such as sorting them from high to low according to the viewing probability, starting from the object to be placed ranked in the first position to the object to be placed ranked in the preset position, and the selected object to be placed is determined as the target object, and the target object is outputted.

[0096] It can be seen from the above technical solution that after obtaining the delivery data of each object to be delivered, if the delivery data of the object to be delivered does not meet the preset conditions, the delivery data of the object to be delivered is numerically corrected to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty; according to the target delivery data of the object to be delivered, the viewing probability of the object to be delivered is determined, and according to the viewing probability of each object to be delivered, the target object is determined from all the objects to be delivered. For the object to be delivered, the delivery data of the object to be delivered can be numerically corrected so that the value of the target delivery data of the object to be delivered is not empty, so as to determine the viewing probability of the object to be delivered whose delivery data value is empty. In this way, in the process of determining the target object based on the viewing probability, the object to be delivered whose delivery data value is empty will be used as a candidate object, so that the object to be delivered whose delivery data value is empty has the possibility of being used as a target object. For example, for an object to be delivered that is currently generated and has not been used as a target object or even has not been viewed, its delivery data may have an empty value. By numerically correcting the delivery data, this type of object to be delivered may be used as a target object, so that its viewing probability can be determined based on the corrected target delivery data, thereby improving the accuracy of target object determination.

[0097] Figure 2 is a flow chart of another data processing method according to an exemplary embodiment, which may include the following steps:

[0098] In step S21, the delivery data of each object to be delivered is obtained.

[0099] In step S22, if the delivery data of the object to be delivered does not meet the preset conditions, the delivery data of the object to be delivered is numerically corrected to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty.

[0100] In step S23, the viewing probability of the object to be delivered is determined according to the target delivery data of the object to be delivered, wherein the viewing probability is used to indicate the probability of the object to be delivered being viewed by the user.

[0101] In step S24, if the delivery data of the object to be delivered meets the preset conditions, the viewing probability of the object to be delivered is determined according to the delivery data of the object to be delivered.

[0102] If the delivery data of the object to be delivered meets the preset conditions, it means that the object to be delivered has been exposed many times, and the delivery data of the object to be delivered can effectively describe the object to be delivered. In this case, the viewing probability can be obtained based on the proportional relationship between the actual values ​​of each parameter in the delivery data. For example, the delivery data includes: cover clicks, cover exposure, video clicks, video exposure and conversion rate. Cover click rate = cover clicks / cover exposure; video click rate = video clicks / video exposure, conversion rate = conversion number / video clicks, and viewing probability = cover click rate*video click rate*conversion rate.

[0103] The calculation process of the viewing probability in this embodiment is described by a calculation formula:

[0104]

[0105] Among them, ctr1, ctr2 and cvr are cover click-through rate, video click-through rate and conversion rate respectively. Cover click-through rate = cover clicks / cover exposure; video click-through rate = video clicks / video exposure, conversion rate = conversions / video clicks; ectr1, ectr2 and ecvr are target cover click-through rate, target video click-through rate and target conversion rate respectively. Different from the calculation method of ctr1, ctr2 and cvr, ectr1, ectr2 and ecvr are obtained by combining the preset distribution sampling function.

[0106] In step S25, the target object is determined from all the objects to be delivered according to the viewing probability of each object to be delivered.

[0107] It can be seen from the above technical solution that after obtaining the delivery data of the object to be delivered, different methods are used to obtain the corresponding viewing probabilities for the objects to be delivered that meet the preset conditions and those that do not meet the preset conditions, so that each object to be delivered can choose a determination method that matches it to determine the viewing probability, thereby improving the accuracy of the viewing probability.

[0108] After determining the target object, the data processing method provided in this embodiment may further include the following steps: obtaining a first number of target objects determined based on the viewing probability obtained from the target delivery data and a second number of target objects determined based on the viewing probability obtained from the delivery data; if the quantitative relationship between the first number and the second number does not satisfy the preset quantitative relationship, adjusting the preset conditions.

[0109] In the process of determining the viewing probability, the objects to be delivered whose delivery data do not meet the preset conditions obtain the viewing probability according to the target delivery data, and the objects to be delivered whose delivery data meet the preset conditions obtain the viewing probability according to the delivery data. Then, the target objects determined according to the viewing probability include: the objects to be delivered whose delivery data do not meet the preset conditions and the objects to be delivered whose delivery data meet the preset conditions. In this embodiment, the quantities of these two types of objects to be delivered determined as target objects, i.e., the first quantity and the second quantity mentioned above, can be obtained. If the quantitative relationship between the first quantity and the second quantity does not meet the preset quantitative relationship, the preset conditions are adjusted so that when the target objects are determined again according to the adjusted preset conditions, the quantitative relationship between the first quantity and the second quantity meets the preset quantitative relationship.

[0110] For example, the preset quantity relationship indicates a ratio range between the first quantity and the second quantity. If the ratio between the first quantity and the second quantity is not within the ratio range, it indicates to adjust the preset conditions. For example, the threshold corresponding to at least one field among the cover clicks, cover exposures, video clicks, video exposures and conversion numbers in the preset conditions can be adjusted, thereby realizing dynamic adjustment of the threshold according to the preset quantity relationship. According to the adjusted preset conditions, the number of objects to be delivered whose delivery data does not meet the preset conditions and the number of objects to be delivered whose delivery data meet the preset conditions among all target objects can be limited.

[0111] In this embodiment, in addition to limiting the number of objects to be delivered whose delivery data do not meet the preset conditions and the number of objects to be delivered whose delivery data meet the preset conditions among all target objects by adjusting the preset conditions, other methods can also be used for limitation. For example, another method is: in the process of determining the target objects according to the viewing probability obtained from the target delivery data and the target objects according to the viewing probability obtained from the delivery data, according to the viewing probability obtained from the target delivery data and the first preset number, a first preset number of target objects are determined from the objects to be delivered whose viewing probability is determined according to the target delivery data; according to the viewing probability obtained from the delivery data and the second preset number, a second preset number of target objects are determined from the objects to be delivered whose viewing probability is determined according to the delivery data.

[0112] The first preset number and the second preset number are not limited in this embodiment. After obtaining the viewing probability of each object to be delivered, all objects to be delivered whose viewing probabilities are determined according to the target delivery data can be taken as the first object set, and all objects to be delivered whose viewing probabilities are determined according to the delivery data can be taken as the second object set. For the first object set, according to the viewing probability of each object to be delivered in the first object set, a first preset number of objects to be delivered are selected as target objects. Similarly, for the second object set, according to the viewing probability of each object to be delivered in the second object set, a second preset number of objects to be delivered are selected as target objects. In this way, the preset conditions for distinguishing the objects to be delivered may be different from the above preset conditions. For example, the delivery number field is added to the preset conditions to divide the objects to be delivered in combination with the delivery number.

[0113] Figure 3 is a flow chart of another data processing method according to an exemplary embodiment, which may include the following steps:

[0114] In step S31, the delivery data of each object to be delivered is obtained.

[0115] In step S32, the delivery data of each object to be delivered is divided according to the delivery scenario to obtain a delivery data set corresponding to each delivery scenario. In addition to the above: cover exposure, cover clicks, video exposure, video clicks and conversions, the delivery data of each object to be delivered may also include a delivery scenario field, so as to divide the delivery data according to the delivery scenario, and divide the delivery data of each object to be delivered belonging to the same delivery scenario into a delivery data set. If the delivery scenario is not a field in the delivery data, the delivery data can be marked with the delivery scenario, such as if the delivery scenario and the delivery data are in one data record, thereby marking the delivery scenario to which the delivery data belongs.

[0116] For the delivery data in the delivery data set of each delivery scenario, the following steps S33 to S36 are respectively executed, and the target object under each delivery scenario is determined with each delivery scenario as a unit. For each delivery scenario, the delivery data under a delivery scenario can also be divided twice according to the above preset conditions, and different calculation methods are performed for the delivery data that meets the preset conditions and the delivery data that does not meet the preset conditions. For details, please refer to the relevant description in the above embodiment, which will not be elaborated in this embodiment.

[0117] In step S33, if the delivery data of the object to be delivered does not meet the preset conditions, the delivery data of the object to be delivered is numerically corrected to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty.

[0118] In step S34, the viewing probability of the object to be delivered is determined according to the target delivery data of the object to be delivered, wherein the viewing probability is used to indicate the probability of the object to be delivered being viewed by the user.

[0119] In step S35, if the delivery data of the object to be delivered meets the preset conditions, the viewing probability of the object to be delivered is determined according to the delivery data of the object to be delivered.

[0120] In step S36, the target object belonging to each delivery scene is determined according to the viewing probability of each object to be delivered in each delivery data set, and independent optimization is achieved under different delivery scenes, ensuring the preferred delivery of the target object in each delivery scene.

[0121] According to the above technical solution, after obtaining the delivery data of each object to be delivered, the delivery data of each object to be delivered is divided according to the delivery scenario to obtain a delivery data set corresponding to each delivery scenario. In this way, the target object in any delivery scenario can be determined based on the delivery scenario, and independent optimization can be achieved in different delivery scenarios, ensuring the preferred delivery of the target object in each delivery scenario.

[0122] Figure 4 It is a block diagram of a data processing device according to an exemplary embodiment, which may include: an acquisition unit 10, a correction unit 20, a first determination unit 30 and a second determination unit 40.

[0123] The acquisition unit 10 is configured to execute the acquisition of the delivery data of each object to be delivered. One way in which the acquisition unit 10 acquires the delivery data of each object to be delivered is: the acquisition unit is configured to execute the acquisition of the real-time delivery data of each object to be delivered in the current time period, and acquire the historical delivery data of each object to be delivered; the historical delivery data of each object to be delivered and the real-time delivery data of each object to be delivered are aggregated to obtain the delivery data of each object to be delivered. For other ways of acquiring the delivery data of each object to be delivered, please refer to the above description, which will not be repeated here.

[0124] The correction unit 20 is configured to perform numerical correction on the delivery data of the object to be delivered if the delivery data of the object to be delivered does not meet the preset conditions, so as to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty.

[0125] If the delivery data of the object to be delivered does not meet the preset conditions, it means that the number of times the object to be delivered has been exposed is small or even has not been exposed, so it is necessary to make a numerical correction so that the value of the target delivery data is not empty. The delivery data of the object to be delivered includes: cover clicks, cover exposure, video clicks, video exposure and conversion number, and the corresponding preset conditions include: cover clicks, cover exposure, video clicks, video exposure and conversion number are respectively greater than their respective thresholds, that is, a threshold is set for each field in the delivery data. If the value of at least one field in the delivery data is less than or equal to the corresponding threshold, it is determined that the delivery data of the object to be delivered does not meet the preset conditions. If the value of each field in the delivery data is greater than the corresponding threshold, it is determined that the delivery data of the object to be delivered meets the preset conditions. The threshold corresponding to each field is not limited in this embodiment. The numerical correction method includes but is not limited to: adding the field whose value is empty in the delivery data of the object to be delivered to the preset value. For different fields, the added preset values ​​can be the same or different.

[0126] The first determination unit 30 is configured to determine the viewing probability of the object to be delivered according to the target delivery data of the object to be delivered, and the viewing probability is used to indicate the probability of the object to be delivered from delivery to being viewed by the user. One way is: the first determination unit 30 is configured to estimate the current delivery trend of the object to be delivered according to the distribution of the target delivery data of the object to be delivered, and obtain the estimation result of the object to be delivered. According to the estimation result of the object to be delivered, the viewing probability of the object to be delivered is determined.

[0127] The estimated result of the object to be delivered includes at least one of a target cover click rate, a target video click rate, and a target conversion rate. The estimation process of the target cover click rate, the target video click rate, and the target conversion rate is as follows:

[0128] The first determination unit 30 is configured to construct a cover distribution function of the target cover click volume and the target cover exposure volume according to the target cover click volume and the target cover exposure volume in the target delivery data, and determine the target cover click rate of the object to be delivered according to the cover distribution function.

[0129] The first determination unit 30 is configured to construct a video distribution function of the target video click volume and the target video exposure volume according to the target video click volume and the target video exposure volume in the target delivery data, and determine the target video click rate of the object to be delivered according to the video distribution function.

[0130] The first determination unit 30 is configured to construct a conversion distribution function of the target conversion number and the target video click volume according to the target conversion number and the target video click volume in the target delivery data, and determine the target conversion rate of the delivery object according to the conversion distribution function.

[0131] The second determination unit 40 is configured to determine the target object from all the objects to be placed according to the viewing probability of each object to be placed. The target object is an object to be placed that can be outputted and selected from all the objects to be placed, such as an object to be placed selected for display in a display object. Because the number of objects to be placed that can be outputted is limited, this embodiment can sort the objects to be placed according to the viewing probability of each object to be placed, such as sorting them from high to low according to the viewing probability, starting from the object to be placed ranked in the first place to the object to be placed ranked in a preset place, and the selected object to be placed is determined as the target object, and the target object is output.

[0132] It can be seen from the above technical solution that after obtaining the delivery data of each object to be delivered, if the delivery data of the object to be delivered does not meet the preset conditions, the delivery data of the object to be delivered is numerically corrected to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty; according to the target delivery data of the object to be delivered, the viewing probability of the object to be delivered is determined, and according to the viewing probability of each object to be delivered, the target object is determined from all the objects to be delivered. For the object to be delivered, the delivery data of the object to be delivered can be numerically corrected so that the value of the target delivery data of the object to be delivered is not empty, so as to determine the viewing probability of the object to be delivered whose delivery data value is empty. In this way, in the process of determining the target object based on the viewing probability, the object to be delivered whose delivery data value is empty will be used as a candidate object, so that the object to be delivered whose delivery data value is empty has the possibility of being used as a target object. For example, for an object to be delivered that is currently generated and has not been used as a target object or even has not been viewed, its delivery data may have an empty value. By numerically correcting the delivery data, this type of object to be delivered may be used as a target object, so that its viewing probability can be determined based on the corrected target delivery data, thereby improving the accuracy of target object determination.

[0133] The data processing device provided in this embodiment also includes: a third determination unit, which is configured to execute, if the delivery data of the object to be delivered meets the preset conditions, determine the viewing probability of the object to be delivered according to the delivery data of the object to be delivered, so that the data processing device adopts different calculation methods to obtain the corresponding viewing probabilities for the objects to be delivered that meet the preset conditions and those that do not meet the preset conditions, so that each object to be delivered can select the calculation method that matches it for calculation, thereby improving the accuracy of the viewing probability.

[0134] If both the first determination unit and the third determination unit can obtain the viewing probability of the object to be delivered, then in the process of determining the target object by the second determination unit, one method is that the second determination unit 40 is configured to execute, based on the viewing probability obtained through the target delivery data and the first preset number, to determine a first preset number of target objects from the objects to be delivered whose viewing probability is determined according to the target delivery data; and to determine a second preset number of target objects from the objects to be delivered whose viewing probability is determined according to the delivery data, based on the viewing probability obtained through the delivery data and the second preset number.

[0135] After the target object is determined by the second determination unit, the acquisition unit 10 is configured to further acquire the first number of target objects determined according to the viewing probability obtained from the target delivery data and the second number of target objects determined according to the viewing probability obtained from the delivery data. The corresponding data processing device also includes: an adjustment unit configured to adjust the preset conditions if the quantitative relationship between the first number and the second number does not satisfy the preset quantitative relationship, so that when the target object is determined again according to the adjusted preset conditions, the quantitative relationship between the first number and the second number satisfies the preset quantitative relationship.

[0136] For example, the preset quantity relationship indicates a ratio range between the first quantity and the second quantity. If the ratio between the first quantity and the second quantity is not within the ratio range, it indicates to adjust the preset conditions. For example, the threshold corresponding to at least one field among the cover clicks, cover exposures, video clicks, video exposures and conversion numbers in the preset conditions can be adjusted, thereby realizing dynamic adjustment of the threshold according to the preset quantity relationship. According to the adjusted preset conditions, the number of objects to be delivered whose delivery data does not meet the preset conditions and the number of objects to be delivered whose delivery data meet the preset conditions among all target objects can be limited.

[0137] Figure 5 It is a block diagram of another data processing device according to an exemplary embodiment, and can also include: a division unit 50, configured to divide the delivery data of each object to be delivered according to the delivery scene, and obtain a delivery data set corresponding to each delivery scene. A second determination unit 40 is configured to determine the target object belonging to each delivery scene according to the viewing probability of each object to be delivered in each delivery data set, thereby determining the target object in any delivery scene in units of delivery scenes, realizing independent optimization in different delivery scenes, and ensuring the preferred delivery of the target object in each delivery scene.

[0138] Regarding the device in the above embodiment, the specific manner in which each unit performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0139] This embodiment also provides an electronic device, including: a processor; and a memory for storing instructions executable by the processor.

[0140] The processor is configured to execute instructions to implement the above data processing method.

[0141] This embodiment also provides a storage medium, in which instructions are stored, and when the instructions in the storage medium are executed by the processor, the processor can execute the above data processing method. Optionally, the storage medium can be a non-temporary computer-readable storage medium, for example, the non-temporary computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0142] This embodiment also provides a computer program product, which, when executed on an electronic device, is suitable for executing a program that initializes the steps of the above-mentioned image processing method.

[0143] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0144] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A data processing method, characterized in that: include: Acquire delivery data of each object to be delivered, wherein the delivery data indicates how the object to be delivered is viewed during the process of delivering the object to be delivered to the display; If the delivery data of the object to be delivered does not meet the preset conditions, the delivery data of the object to be delivered is numerically corrected to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty; Determining the viewing probability of the object to be delivered according to the target delivery data of the object to be delivered, wherein the viewing probability is used to indicate the probability of the object to be delivered being viewed by the user from delivery; According to the viewing probability of each object to be delivered, the target object is determined from all the objects to be delivered; The step of performing numerical correction on the delivery data of the object to be delivered comprises: Adding a field whose value is empty in the delivery data of the object to be delivered to a preset value, or adding each field in the delivery data of the object to be delivered to a preset value; The determining, according to the target delivery data of the object to be delivered, the viewing probability of the object to be delivered comprises: According to the distribution of the target delivery data of the object to be delivered, the current delivery trend of the object to be delivered is estimated to obtain the estimated result of the object to be delivered, and the current delivery trend of the object to be delivered reflects the development trend of the object to be delivered over time; The viewing probability of the object to be delivered is determined according to the estimation result of the object to be delivered.

2. The method according to claim 1, characterized in that The estimated result of the object to be delivered includes at least one of a target cover click-through rate, a target video click-through rate, and a target conversion rate; The target cover click rate estimation process includes: constructing a cover distribution function of the target cover click amount and the target cover exposure amount according to the target cover click amount in the target delivery data and the target cover exposure amount in the target delivery data; determining the target cover click rate of the object to be delivered according to the cover distribution function; The target video click rate estimation process includes: constructing a video distribution function of the target video click amount and the target video exposure amount according to the target video click amount in the target delivery data and the target video exposure amount in the target delivery data; determining the target video click rate of the object to be delivered according to the video distribution function; The target conversion rate estimation process includes: constructing a conversion distribution function of the target conversion number and the target video click volume according to the target conversion number and the target video click volume in the target delivery data; and determining the target conversion rate of the object to be delivered according to the conversion distribution function.

3. The method according to claim 1, characterized in that The method further includes: if the delivery data of the object to be delivered meets a preset condition, determining the viewing probability of the object to be delivered according to the delivery data of the object to be delivered.

4. The method according to claim 3, characterized in that The method further includes: obtaining a first number of target objects determined according to the viewing probability obtained from the target delivery data and a second number of target objects determined according to the viewing probability obtained from the delivery data; If the quantitative relationship between the first quantity and the second quantity does not satisfy a preset quantitative relationship, the preset condition is adjusted.

5. The method according to claim 3, characterized in that: Determining the target object from all the objects to be delivered according to the viewing probability of each object to be delivered includes: According to the viewing probability obtained through the target delivery data and the first preset number, determining a first preset number of target objects from the objects to be delivered whose viewing probability is determined according to the target delivery data; According to the viewing probability obtained through the delivery data and the second preset number, a second preset number of target objects are determined from the objects to be delivered whose viewing probability is determined according to the delivery data.

6. The method according to any one of claims 1 to 5, characterized in that The delivery data includes: cover clicks, cover exposure, video clicks, video exposure and conversions; The preset conditions include: the cover click volume, cover exposure volume, video click volume, video exposure volume and conversion number are respectively greater than their respective thresholds.

7. The method according to any one of claims 1 to 5, characterized in that The obtaining of delivery data of each object to be delivered includes: Obtain the real-time delivery data of each object to be delivered in the current time period; Obtain the historical delivery data of each object to be delivered; The historical delivery data of each object to be delivered and the real-time delivery data of each object to be delivered are aggregated to obtain the delivery data of each object to be delivered.

8. The method according to any one of claims 1 to 5, characterized in that The method further includes: dividing the delivery data of each to-be-delivered object according to the delivery scenario to obtain a delivery data set corresponding to each delivery scenario; Determining the target object from all the objects to be delivered according to the viewing probability of each object to be delivered includes: determining the target object belonging to each delivery scene according to the viewing probability of each object to be delivered in each delivery data set.

9. A data processing device, characterized in that: include: An acquisition unit is configured to acquire delivery data of each object to be delivered, wherein the delivery data indicates a viewing status of the object to be delivered during delivery of the object to be delivered to the display; A correction unit is configured to perform a numerical correction on the delivery data of the object to be delivered if the delivery data of the object to be delivered does not meet a preset condition, so as to obtain the target delivery data of the object to be delivered, so that the value of the target delivery data of the object to be delivered is not empty; A first determining unit is configured to determine the viewing probability of the object to be delivered according to the target delivery data of the object to be delivered, wherein the viewing probability is used to indicate the probability of the object to be delivered being viewed by the user from delivery; A second determining unit is configured to determine a target object from all objects to be delivered according to the viewing probability of each object to be delivered; The correction unit is specifically used for: Adding a field whose value is empty in the delivery data of the object to be delivered to a preset value, or adding each field in the delivery data of the object to be delivered to a preset value; The first determining unit is configured to estimate the current delivery trend of the object to be delivered according to the distribution of the target delivery data of the object to be delivered, and obtain the estimation result of the object to be delivered, wherein the current delivery trend of the object to be delivered reflects the development trend of the object to be delivered over time; The viewing probability of the object to be delivered is determined according to the estimation result of the object to be delivered.

10. The device according to claim 9, characterized in that The estimated result of the object to be delivered includes at least one of a target cover click-through rate, a target video click-through rate, and a target conversion rate; The first determining unit is configured to construct a cover distribution function of the target cover click volume and the target cover exposure volume according to the target cover click volume in the target delivery data and the target cover exposure volume in the target delivery data; and determine the target cover click rate of the to-be-delivered object according to the cover distribution function; The first determining unit is configured to construct a video distribution function of the target video click volume and the target video exposure volume according to the target video click volume in the target delivery data and the target video exposure volume in the target delivery data; and determine the target video click rate of the object to be delivered according to the video distribution function; The first determination unit is configured to execute a conversion distribution function of the target conversion number and the target video click volume according to the target conversion number and the target video click volume in the target delivery data; and determine the target conversion rate of the object to be delivered according to the conversion distribution function.

11. The device according to claim 9, characterized in that The device further includes: a third determining unit configured to determine the viewing probability of the object to be delivered according to the delivery data of the object to be delivered if the delivery data of the object to be delivered meets a preset condition.

12. The device according to claim 11, characterized in that The acquisition unit is configured to further acquire a first number of target objects determined according to the viewing probability obtained from the target delivery data and a second number of target objects determined according to the viewing probability obtained from the delivery data; The device further includes: an adjustment unit configured to adjust the preset condition if the quantity relationship between the first quantity and the second quantity does not satisfy a preset quantity relationship.

13. The device according to claim 11, characterized in that The second determination unit is configured to determine a first preset number of target objects from the objects to be delivered that have obtained the target delivery data based on the viewing probability and the first preset number obtained through the target delivery data; and to determine a second preset number of target objects from the objects to be delivered that have obtained the delivery data based on the viewing probability and the second preset number obtained through the delivery data.

14. The device according to any one of claims 9 to 13, characterized in that The delivery data includes: cover clicks, cover exposure, video clicks, video exposure and conversions; The preset conditions include: the cover click volume, cover exposure volume, video click volume, video exposure volume and conversion number are respectively greater than their respective thresholds.

15. The device according to any one of claims 9 to 13, characterized in that The acquisition unit is configured to execute acquisition of real-time delivery data of each object to be delivered in the current time period, and to acquire historical delivery data of each object to be delivered; and to aggregate the historical delivery data of each object to be delivered and the real-time delivery data of each object to be delivered to obtain the delivery data of each object to be delivered.

16. The device according to any one of claims 9 to 13, characterized in that The device further includes: a division unit configured to divide the delivery data of each to-be-delivered object according to the delivery scenario to obtain a delivery data set corresponding to each delivery scenario; The second determination unit is configured to determine the target object belonging to each delivery scene according to the viewing probability of each to-be-delivered object in each delivery data set.

17. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the data processing method as described in any one of claims 1 to 8.

18. A storage medium, characterized in that: The storage medium stores instructions, and when the instructions in the storage medium are executed by a processor, the processor is enabled to execute the data processing method as described in any one of claims 1 to 8.

Images