A conversion rate analysis method and related apparatus

Abstract

Embodiments of the present application disclose a conversion rate analysis method and related device, on one hand, the conversion data corresponding to the historical delivery information can be expanded as prior data based on the relevance between the information recommendation objects, the conversion rate corresponding to the to-be-analyzed information is analyzed, on the other hand, the conversion data corresponding to the to-be-analyzed information before the target time after actual delivery can be obtained as posterior data, so as to reduce the influence of the error based on the analysis of the conversion rate of the associated objects, so that the conversion rate predicted by combining the two aspects can be more consistent with the real conversion rate of the to-be-analyzed information after the target time, and high-precision conversion rate analysis in the sparse scene of conversion data is realized.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a conversion rate analysis method and related apparatus. Background Technology

[0002] Advertising is a common promotional and marketing method in various fields. For example, when selling products or promoting various game programs, advertising is required to attract people to buy products or download game programs.

[0003] Conversion rate is one of the key metrics for measuring advertising effectiveness. It refers to the percentage of people who click on an ad but ultimately convert their clicks. Therefore, to predict advertising performance, an analytical method capable of accurately analyzing conversion rates is needed. Related technologies primarily utilize conversion rate prediction models, which learn how to accurately analyze the conversion rates of ads by collecting conversion data from ads placed by advertisers over historical periods.

[0004] However, many advertisers have placed relatively few ads in a historical period, resulting in a lack of ad conversion data available for analysis and learning. This makes it easy for related technologies to obtain low-accuracy conversion rates through model analysis. Summary of the Invention

[0005] To address the aforementioned technical issues, this application provides a conversion rate analysis method that can expand conversion data while ensuring data validity, and correct the conversion rate analysis by combining posterior data, thereby improving the accuracy of conversion rate analysis under low conversion data conditions.

[0006] The embodiments of this application disclose the following technical solutions:

[0007] In a first aspect, embodiments of this application disclose a conversion rate analysis method, the method comprising:

[0008] Obtain information to be analyzed, which is used to recommend target objects;

[0009] Obtain the first conversion data corresponding to the information to be analyzed. The first conversion data is used to characterize the conversion rate corresponding to the information to be analyzed before the target time. The conversion rate is used to characterize the proportion of objects that perform conversion behavior for the objects recommended by the corresponding information among the objects that click on the corresponding information.

[0010] Determine the historical delivery information corresponding to the information to be analyzed. The historical delivery information is used to recommend associated objects. The associated objects are objects that meet the object association conditions with the target object.

[0011] Obtain the second conversion data corresponding to the historical delivery information, wherein the second conversion data is used to characterize the conversion rate corresponding to the historical delivery information;

[0012] By combining the first conversion data and the second conversion data, the conversion rate of the information to be analyzed after the target time is analyzed.

[0013] Secondly, embodiments of this application disclose a conversion rate analysis device, the device comprising a first acquisition unit, a second acquisition unit, a determination unit, a third acquisition unit, and an analysis unit:

[0014] The first acquisition unit is used to acquire information to be analyzed, and the information to be analyzed is used to recommend target objects;

[0015] The second acquisition unit is used to acquire the first conversion data corresponding to the information to be analyzed. The first conversion data is used to characterize the conversion rate corresponding to the information to be analyzed before the target time. The conversion rate is used to characterize the proportion of objects that perform conversion behavior for the objects recommended by the corresponding information among the objects that click on the corresponding information.

[0016] The determining unit is used to determine the historical delivery information corresponding to the information to be analyzed. The historical delivery information is used to recommend associated objects. The associated objects are objects that meet the object association conditions with the target object.

[0017] The third acquisition unit is used to acquire the second conversion data corresponding to the historical delivery information, and the second conversion data is used to characterize the conversion rate corresponding to the historical delivery information.

[0018] The analysis unit is used to combine the first conversion data and the second conversion data to analyze the conversion rate of the information to be analyzed after the target time.

[0019] In one possible implementation, the conversion count corresponding to the historical delivery information is greater than a preset number, where the conversion count is the number of objects that perform conversion actions on the objects recommended by the corresponding information among the objects that are clicked.

[0020] In one possible implementation, the determining unit is specifically used for:

[0021] The first object association condition is used as the object association condition to obtain the first historical delivery information. The first historical delivery information is used to recommend the first object. The first object and the target object satisfy the first object association condition.

[0022] Based on the fact that the conversion number corresponding to the first historical delivery information has not reached the preset number, the second object association condition is used as the object association condition, and the second historical delivery information is obtained. The second historical delivery information is used to recommend the second object. The second object and the target object meet the second object association condition. The degree of association between the second object and the target object is less than the degree of association between the first object and the target object.

[0023] If the conversion count corresponding to the second historical delivery information reaches the preset number, the second historical delivery information is determined as the historical delivery information.

[0024] In one possible implementation, the third acquisition unit is specifically used for:

[0025] Obtain the conversion data corresponding to the historical delivery information in the first historical time period, and the conversion data corresponding to the first historical time period is used to characterize the conversion rate of the historical delivery information in the first historical time period.

[0026] Based on the fact that the number of conversions corresponding to the historical delivery information in the first historical period did not reach the preset number, the conversion data corresponding to the historical delivery information in the second historical period is obtained. The conversion data corresponding to the second historical period is used to characterize the conversion rate corresponding to the historical delivery information in the second historical period. The second historical period includes the first historical period, and the start time of the second historical period is earlier than the start time of the first historical period.

[0027] Based on the fact that the conversion number corresponding to the historical delivery information in the second historical period reaches the preset number, the conversion data corresponding to the historical delivery information in the second historical period is determined as the second conversion data.

[0028] In one possible implementation, the analysis unit is specifically used for:

[0029] Based on the first conversion data, the posterior conversion rate corresponding to the information to be analyzed is determined, wherein the posterior conversion rate is the conversion rate represented by the first conversion data;

[0030] Based on the second conversion data, predict the prior conversion rate corresponding to the information to be analyzed;

[0031] By combining the prior conversion rate and the posterior conversion rate, the conversion rate of the information to be analyzed after the target time is analyzed.

[0032] In one possible implementation, the device further includes a fourth acquisition unit:

[0033] The fourth acquisition unit is used to acquire a first weight corresponding to the prior conversion rate and a second weight corresponding to the posterior conversion rate. The first weight is used to control the degree of reference to the prior conversion rate when determining the conversion rate of the information to be analyzed after the target time. The second weight is used to control the degree of reference to the posterior conversion rate when determining the conversion rate of the information to be analyzed after the target time.

[0034] The analysis unit is specifically used for:

[0035] By combining the first weight, the prior conversion rate, the second weight, and the posterior conversion rate, the conversion rate of the information to be analyzed after the target time is analyzed.

[0036] In one possible implementation, the first conversion data includes the first number of objects that clicked on the information to be analyzed before the target time, and the fourth acquisition unit is specifically used for:

[0037] Based on the second conversion data and the number of the first objects, a first parameter and a second parameter corresponding to the beta distribution are determined. The ratio of the first parameter to the second parameter corresponds to the ratio between the conversion rate and the non-conversion rate represented by the second conversion data. The ratio of the first parameter to the second parameter and the first object number corresponds to the ratio of the first weight to the second weight. The first parameter and the second parameter are used to adjust the distribution mode represented by the beta distribution. The beta distribution is used to represent the distribution mode of the prior conversion rate.

[0038] The ratio of the sum of the first parameter to the sum of the second parameter is determined as the first weight, and the ratio of the number of the first object to the sum of the second parameter is determined as the second weight. The sum of the first parameter is the sum of the first parameter and the second parameter, and the sum of the second parameter is the sum of the first parameter and the number of the first object.

[0039] The analysis unit is specifically used for:

[0040] The expected value corresponding to the beta distribution is determined as the prior conversion rate.

[0041] In one possible implementation, the sum of the first parameters equals the number of the first objects.

[0042] In one possible implementation, the information to be analyzed is used to recommend the target object to the object to be recommended, and the device further includes a fifth acquisition unit and a first prediction unit:

[0043] The fifth acquisition unit is used to acquire first feature information corresponding to the target object and second feature information corresponding to the object to be recommended. The first feature information is used to characterize the object feature corresponding to the target object, and the second feature information is used to characterize the object feature corresponding to the object to be recommended.

[0044] The first prediction unit is configured to predict the first conversion rate corresponding to the information to be analyzed based on the first feature information and the second feature information;

[0045] The analysis unit is specifically used for:

[0046] By combining the first conversion data and the second conversion data, the second conversion rate corresponding to the information to be analyzed after the target time is analyzed;

[0047] The second conversion rate is adjusted based on the difference between the first conversion rate and the second conversion rate to obtain the conversion rate of the information to be analyzed after the target time. The adjustment is used to reduce the difference between the first conversion rate and the second conversion rate.

[0048] In one possible implementation, the device further includes a sixth acquisition unit, a seventh acquisition unit, a second prediction unit, and a first adjustment unit:

[0049] The sixth acquisition unit is used to acquire historical sample delivery information and the sample conversion rate corresponding to the historical sample delivery information. The historical sample delivery information is used to recommend sample objects to sample objects to be recommended. The sample conversion rate is the conversion rate obtained by delivering the historical sample delivery information in a historical period.

[0050] The seventh acquisition unit is used to acquire first sample feature information and second sample feature information. The first sample feature information is used to characterize the object features corresponding to the sample object, and the second sample feature information is used to characterize the object features corresponding to the sample object to be recommended.

[0051] The second prediction unit is used to predict the undetermined conversion rate corresponding to the historical delivery information of the sample based on the first sample feature information and the second sample feature information through an initial conversion rate prediction model;

[0052] The first adjustment unit is used to adjust the model parameters corresponding to the initial conversion rate prediction model according to the difference between the undetermined conversion rate and the sample conversion rate, so as to obtain the conversion rate prediction model;

[0053] The first prediction unit is specifically used for:

[0054] Based on the first feature information and the second feature information, the conversion rate prediction model is used to predict the first conversion rate corresponding to the information to be analyzed.

[0055] In one possible implementation, the analysis unit is specifically used for:

[0056] Based on the difference between the first conversion rate and the second conversion rate and the adjustment parameter, the second conversion rate is adjusted to obtain the conversion rate of the information to be analyzed after the target time. The adjustment parameter is used to control the adjustment intensity of the second conversion rate based on the difference between the first conversion rate and the second conversion rate.

[0057] In one possible implementation, the second acquisition unit is specifically used for:

[0058] Based on a preset time interval of n times between the target time and the initial click time corresponding to the information to be analyzed, first conversion data corresponding to the information to be analyzed is obtained. The first conversion data is used to characterize the conversion rate of the information to be analyzed between the initial click time and the target time. The initial click time is the time when the information to be analyzed is first clicked, and n is a positive integer.

[0059] In one possible implementation, the apparatus further includes a generation unit:

[0060] The generation unit is configured to generate a prompt message based on the fact that the difference between the conversion rate of the information to be analyzed after the target time and the conversion rate of the information to be analyzed after a historical time is greater than a difference threshold. The prompt message is used to indicate that the conversion rate of the information to be analyzed has changed abnormally. The historical time is the previous time that is separated from the target time by the preset time interval. The conversion rate of the information to be analyzed after the historical time is obtained based on the conversion rate of the information to be analyzed before the historical time.

[0061] In one possible implementation, the device further includes a second adjustment unit:

[0062] The second adjustment unit is used to adjust the delivery method of the information to be analyzed after the target time according to the conversion rate of the information to be analyzed after the target time. The delivery method is used to determine the prominence of displaying the information to be analyzed.

[0063] Thirdly, embodiments of this application disclose a computer device, which includes a processor and a memory:

[0064] The memory is used to store computer programs and to transfer the computer programs to the processor;

[0065] The processor is configured to execute the conversion rate analysis method described in any one of the first aspects according to the instructions in the computer program;

[0066] Fourthly, embodiments of this application disclose a computer-readable storage medium for storing a computer program for executing the conversion rate analysis method described in any one of the first aspects;

[0067] Fifthly, embodiments of this application disclose a computer program product including a computer program, which, when run on a computer device, causes the computer device to execute the conversion rate analysis method described in any one of the first aspects.

[0068] As can be seen from the above technical solutions, when performing conversion rate analysis on the information to be analyzed, this application can obtain historical delivery information that is highly correlated with the target objects recommended by the information to be analyzed to avoid insufficient conversion data. The conversion data corresponding to the historical delivery information can, to a certain extent, characterize the conversion status of the recommended target objects after delivery. Therefore, on the one hand, the conversion data corresponding to this part of the historical delivery information can be used as prior data to analyze the conversion rate of the information to be analyzed after the target time. On the other hand, in order to reduce the error caused by analyzing the conversion rate based on related objects, the conversion data of the information to be analyzed before the target time can also be obtained as posterior data to analyze the actual conversion status of the information to be analyzed, and the conversion rate of the information to be analyzed after the target time can be analyzed based on this. As can be seen from the above, this application can, on the one hand, use the conversion data of related objects as prior data to increase the amount of data referenced for conversion rate analysis, and at the same time, analyze the conversion rate from a higher-dimensional object dimension. On the other hand, it can reduce the analysis error of combining related objects by collecting the actual conversion data of the information to be analyzed after the launch as posterior data. Thus, it can conduct a more comprehensive and accurate analysis of the conversion rate of the information to be analyzed after the target time, providing a reliable data foundation for the subsequent processing of the information to be analyzed. Attached Figure Description

[0069] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0070] Figure 1This is a schematic diagram of a device interaction scenario provided in an embodiment of this application;

[0071] Figure 2 A schematic diagram illustrating a conversion rate analysis method in a practical application scenario provided by an embodiment of this application;

[0072] Figure 3 A flowchart of a conversion rate analysis method provided in an embodiment of this application;

[0073] Figure 4 A schematic diagram illustrating a conversion rate analysis method provided in an embodiment of this application;

[0074] Figure 5 A schematic diagram illustrating a conversion rate analysis method provided in an embodiment of this application;

[0075] Figure 6 A flowchart illustrating a conversion rate analysis method in a practical application scenario, provided as an embodiment of this application;

[0076] Figure 7 A structural block diagram of a conversion rate analysis device provided in an embodiment of this application;

[0077] Figure 8 A structural diagram of a terminal provided in an embodiment of this application;

[0078] Figure 9 This is a structural diagram of a server provided in an embodiment of this application. Detailed Implementation

[0079] The embodiments of this application will now be described with reference to the accompanying drawings.

[0080] Information delivery is one of the main methods of promotion and advertising. To measure the effectiveness of information delivery, it is usually necessary to analyze the conversion rate of the delivered information. Conversion rate refers to the percentage of people who click on the delivered information and then convert their clicks into actual purchases of the recommended products or services. A click refers to the act of browsing the delivered information. The conversion behavior may vary depending on the recommended products or services. For example, when the delivered information is an advertisement for a product, the conversion behavior may be the act of purchasing the product; when the delivered information is an advertisement for a game, the conversion behavior may be the act of registering for the game. A higher conversion rate indicates a better information delivery effect.

[0081] In related technologies, conversion rate prediction models are mainly used to predict the conversion rate of the information to be analyzed. Training a conversion rate prediction model requires a large amount of conversion data. For example, when it is necessary to predict the conversion rate of information delivered by a specific advertiser, it is necessary to obtain the conversion data of the information delivered by that advertiser in historical periods to train the model, so that the model can learn the conversion characteristics of the information delivered by that advertiser.

[0082] Therefore, when a particular advertiser has limited conversion data for its campaigns over a historical period, the conversion rate prediction model struggles to accurately predict the effectiveness of its campaigns due to the lack of corresponding conversion data. This model becomes highly susceptible to fluctuations in the conversion data. Fluctuations refer to the situation where insufficient conversion data leads to a high degree of randomness in the conversion rate representation, making it difficult to accurately reflect the true conversion characteristics of the information. Consequently, related technologies struggle to provide accurate and effective conversion rate analysis for scenarios with limited conversion data.

[0083] To address the aforementioned technical issues, this application provides a conversion rate analysis method. On one hand, it can expand the conversion data corresponding to historical delivery information as prior data based on the correlation between information recommendation objects to analyze the conversion rate of the information to be analyzed. On the other hand, it can obtain the conversion data corresponding to the information to be analyzed after actual delivery but before the target time as posterior data to reduce the impact of errors caused by analyzing conversion rates based on related objects. Thus, the conversion rate predicted by combining the two aspects can better match the actual conversion rate of the information to be analyzed after the target time, achieving high-precision conversion rate analysis in scenarios with sparse conversion data.

[0084] Understandably, this method can be applied to computer devices capable of conversion rate analysis, such as terminal devices or servers. This method can be executed independently by a terminal device or server, or it can be applied to network scenarios where the terminal device and server communicate, executing in cooperation. The terminal device can be a mobile phone, tablet, laptop, desktop computer, etc. The terminal device can also include various virtual reality devices, such as augmented reality (AR) devices like AR glasses and AR screens, and virtual reality (VR) devices like VR headsets. The server can be understood as an application server or a web server. In actual deployment, the server can be a standalone server, a cluster server, or a cloud server, etc.

[0085] See Figure 1 , Figure 1This is a schematic diagram of a device interaction scenario provided in an embodiment of this application, which introduces two possible device interaction scenarios. In scenario 1, the computer device can be a terminal device 101. The terminal device 101 can send a data acquisition request to the server 102 used to store conversion data to obtain the conversion data stored by the server 102, and then use the conversion data to perform conversion rate analysis and output the obtained conversion rate. In scenario 2, the computer device can be a server 103. When conversion rate analysis is required, the terminal device 104 will send a conversion rate analysis request to the server 103. After receiving the request, the server 103 can obtain the conversion data on its own, perform conversion rate analysis, and return the obtained conversion rate to the terminal device 104.

[0086] To facilitate understanding of the technical solution provided in this application, the conversion rate analysis method provided in this application will be introduced next in conjunction with a practical application scenario.

[0087] See Figure 2 , Figure 2 This is a schematic diagram illustrating a conversion rate analysis method in a practical application scenario provided by an embodiment of this application. In this practical application scenario, device interaction can be as follows: Figure 1 As shown in Scenario 1.

[0088] Terminal device 101 can first acquire the information to be analyzed and the corresponding first conversion data. The first conversion data is used to characterize the conversion rate of the information to be analyzed after delivery and before the target time. Since the first conversion data is real data, it can be regarded as posterior data in conversion rate analysis, used to characterize the actual conversion situation of the information to be analyzed. Terminal device 101 can also acquire historical delivery information and the corresponding second conversion data based on object association conditions. The historical delivery information is used to recommend related objects. The related objects and the target objects meet the object association conditions, that is, the two objects are highly related. Therefore, the information used to recommend these two objects usually has certain commonalities in conversion.

[0089] Based on this, the computer device can use the second conversion data, which represents the conversion rate of historical delivery information, as prior data to analyze the conversion rate of the information to be analyzed after the target time, ensuring a certain level of analytical accuracy. However, since analyzing delivery information corresponding to associated objects still introduces some analytical error, the terminal device 101 can combine the prior and posterior data, using the actual conversion situation represented by the posterior data to correct the errors introduced by the prior data. This allows for a more accurate analysis of the conversion rate of the information to be analyzed after the target time by combining these two types of data.

[0090] Therefore, on the one hand, obtaining conversion data corresponding to historical campaign information based on object association conditions can compensate for the problem of sparse conversion data caused by the limited number of campaign information and short campaign time corresponding to the information to be analyzed. This provides more sufficient analyzable data for conversion rate analysis while ensuring data validity. On the other hand, by using conversion data that represents the actual conversion rate of the information to be analyzed before the target time, a more accurate analysis of the actual conversion situation corresponding to the information to be analyzed can be performed, reducing the analysis error based on prior data analysis. Ultimately, combining prior and posterior data can improve the accuracy of conversion rate analysis.

[0091] Next, the technical solution provided in this application will be described in detail with reference to the accompanying drawings.

[0092] See Figure 3 , Figure 3 A flowchart of a conversion rate analysis method provided in this application embodiment, the method including:

[0093] S301: Obtain information to be analyzed.

[0094] The information to be analyzed can be any type of information that can be delivered, such as advertising information or push notifications. This information is used to recommend to target audiences, who can be any object that can be recommended through the information, such as various products or programs. Delivery refers to the process of displaying the information to the target audience, which can include various delivery methods, such as delivery through various programs, or delivery via SMS, web pages, etc., without limitation here.

[0095] S302: Obtain the first transformation data corresponding to the information to be analyzed.

[0096] In this application, conversion data is used to characterize the conversion rate corresponding to the information. Conversion data can include various data forms, such as data that directly represents the conversion rate, or data that can be used to calculate the conversion rate, such as the number of clicked objects, the number of converted objects, and the number of unconverted objects.

[0097] The first conversion data is used to characterize the conversion rate of the information to be analyzed before the target time. The target time can be any time after the information to be analyzed is delivered. The conversion rate is used to characterize the percentage of objects that perform conversion behavior for the objects recommended by the corresponding information among the objects that click on the corresponding information. That is, the first conversion data is the actual conversion data after the information to be analyzed is delivered, and it can characterize the actual conversion situation of the information to be analyzed. Based on this, the computer device can use the first conversion data as posterior data for analyzing the conversion rate of the information to be analyzed. Posterior data refers to data used to characterize the verification results.

[0098] S303: Determine the historical delivery information corresponding to the information to be analyzed.

[0099] Because the advertiser of the information being analyzed had a limited number of ads placed in the historical period and the ad placement time was relatively short, there may be limited conversion data available for analyzing the conversion rate of the information being analyzed. Therefore, to improve the accuracy of conversion rate analysis, computer equipment can acquire more sufficient conversion data for analysis.

[0100] Understandably, when multiple pieces of information recommend highly relevant objects, their conversion rates often share certain commonalities. For example, two pieces of information recommending the same product typically have similar conversion rates, as do two pieces of information recommending the same type of game. Based on this, computer devices can analyze the conversion rates of the information to be analyzed by using the conversion data corresponding to information recommending objects highly relevant to the target object.

[0101] Computer devices can preset an object association condition, which is used to measure whether two objects have a high degree of association. If two objects meet the object association condition, then it can be determined that the two objects have a high degree of association. The object association condition can include various conditions, such as two objects recommended by different information being the same object, two objects being objects under the same brand, two objects being objects of the same category, etc., which are not limited here. When the object association condition is different, the degree of association between the determined associated objects may also be different, which will be described in detail below, and will not be elaborated here.

[0102] Based on object association conditions, computer devices can obtain historical delivery information corresponding to the information to be analyzed. This historical delivery information refers to information that has been delivered within a historical period, thus enabling the acquisition of conversion data corresponding to the historical delivery information. Specifically, the historical delivery information is used to recommend related objects, which are objects that meet the object association conditions with the target object, i.e., objects with a high degree of association with the target object.

[0103] S304: Obtain the second conversion data corresponding to the historical delivery information.

[0104] The second conversion data is used to characterize the conversion rate corresponding to historical delivery information. Because the target audience and related objects have a high degree of correlation, the conversion rate represented by the second conversion data shares certain commonalities with the conversion rate corresponding to the information to be analyzed. Based on this, the computer device can use the second conversion data as prior data for analyzing the conversion rate corresponding to the information to be analyzed. Prior data refers to data collected before actual verification and used to predict the verification results.

[0105] S305: Combining the first conversion data and the second conversion data, analyze the conversion rate of the information to be analyzed after the target time.

[0106] There are various ways to combine two sets of data for conversion rate analysis, which will be described in detail below and will not be elaborated here. By combining the second conversion data, computer devices can analyze the conversion characteristics of information used to recommend related objects associated with the target object, thereby enabling a more accurate prediction of the conversion rate of the information to be analyzed after the target time.

[0107] Since the conversion rate analysis is based on historical delivery information, there may be some errors, leading to a discrepancy between the analyzed conversion rate and the actual conversion after the delivery of the information. Computer equipment can supplement and correct the conversion rate analysis based on the second conversion data using first conversion data that characterizes the actual conversion of the information to be analyzed, thus obtaining the final conversion rate of the information after the target time. Therefore, the first conversion data can reduce the errors caused by conversion rate analysis based on historical delivery information, while the second conversion data can compensate for the sparsity of conversion data. Thus, combining the first and second conversion data allows for a more accurate analysis of the conversion rate of the information after the target time, considering both prior and posterior data, even in scenarios with sparse conversion data. Furthermore, the second conversion data can characterize the commonalities in information conversion among multiple pieces of information related to the object categories associated with the recommended target object, making this conversion rate analysis method more targeted to specific object categories and further improving the accuracy of the conversion rate analysis.

[0108] The following section will provide a detailed description of the solutions involved in this application.

[0109] Understandably, if the number of conversions corresponding to an information is too low, the conversion rate represented by the conversion data may be highly unreliable. For example, if an information is only converted twice, the total number of clicks on the information may be low, such as only 2 clicks. If analysis is performed based on this conversion data, a 100% conversion rate may be calculated, which is clearly distorted.

[0110] Therefore, to improve the quality of the acquired second conversion data, computer devices can limit the number of conversions corresponding to historical campaign information when acquiring historical campaign information, thereby increasing the reliability of the conversion data. For example, the computer device can require the number of conversions corresponding to historical campaign information to be greater than a preset number. This preset number is used to measure whether the information has a large number of conversions, for example, it can be 10 times. The number of conversions is the number of objects that performed conversion behavior for the recommended objects in the corresponding information among the clicked objects. Since the number of conversions is greater than the preset number, it can be said that the historical campaign information has undergone a relatively large number of conversions. Therefore, the second conversion data corresponding to the historical campaign information can represent a more realistic conversion rate corresponding to the historical campaign information with a large number of clicks, thereby reducing the randomness of the second conversion data corresponding to the historical campaign information, and thus reducing the error brought about by conversion rate analysis based on the second conversion data, further improving the accuracy of conversion rate analysis.

[0111] When retrieving historical campaign information that meets the above requirements based on object association conditions, relying on a single object association condition may result in the inability to obtain the necessary historical campaign information. Therefore, to ensure sufficient conversion data for conversion rate analysis, computer devices can be configured with multiple object association conditions to retrieve historical campaign information.

[0112] When executing step S303, the computer device may execute steps S3031-S3033 (not shown in the figure). Steps S3031-S3033 are one possible implementation of step S303, including:

[0113] S3031: Use the first object association condition as the object association condition to obtain the first historical delivery information.

[0114] When acquiring historical delivery information, computer devices can employ multi-level object association conditions to ensure the acquisition of reliable and effective conversion data for analysis. Taking the first and second object association conditions in a multi-level framework as an example, the computer device can use the first object association condition as the object association condition to acquire the first historical delivery information. This first historical delivery information is used to recommend the first object. The first object and the target object satisfy the first object association condition; that is, under the determination of the first object association condition, the first object and the target object have a high degree of association.

[0115] S3032: Based on the fact that the conversion quantity corresponding to the first historical delivery information has not reached the preset quantity, the second object association condition is used as the object association condition to obtain the second historical delivery information.

[0116] After acquiring the first historical campaign information, the computer device first determines whether the first historical campaign information has a sufficient number of conversions, and then determines whether the conversion data corresponding to the first historical campaign information is reliable and valid. The computer device will determine whether the number of conversions corresponding to the first historical campaign information has reached a preset number. If it has, it means that the first historical campaign information has a sufficient number of conversions. Therefore, the conversion data corresponding to the first historical campaign information is relatively reliable and has a low degree of randomness, and can therefore be used as the second conversion data for conversion rate analysis.

[0117] If the preset number is not reached, it indicates that the conversion rate corresponding to the first historical delivery information is low. In this case, the conversion data corresponding to the first historical delivery information is highly random and has low reliability. Therefore, directly performing conversion rate analysis based on the conversion data corresponding to the first historical delivery information may reduce the accuracy of the conversion rate analysis. When the required historical delivery information cannot be obtained based on the first object association condition, the computer device can use a second object association condition at another level as the object association condition to obtain the second historical delivery information. The second historical delivery information is used to recommend the second object. The second object and the target object meet the second object association condition, that is, under the determination of the second object association condition, the degree of association between the second object and the target object is high.

[0118] Among them, the degree of association between the second object and the target object is less than that between the first object and the target object. That is, the association condition of the second object has a lower requirement for the degree of association between objects, so it can obtain more historical delivery information that meets the conditions, thereby increasing the probability of obtaining historical delivery information with sufficient conversion data.

[0119] S3033: Based on the conversion quantity corresponding to the second historical delivery information reaching the preset quantity, the second historical delivery information is determined as historical delivery information.

[0120] Similarly, after obtaining the second historical campaign information, the computer device can determine whether the conversion count of the second historical campaign information is sufficient. The determination process is similar to that for the first historical campaign information and will not be elaborated here. If the conversion count corresponding to the second historical campaign information reaches a preset number, it indicates that the second historical campaign information has a relatively sufficient number of conversions, and its corresponding conversion data can be used for more accurate conversion rate analysis. Therefore, the computer device can identify the second historical campaign information as the aforementioned historical campaign information for conversion rate analysis.

[0121] In this way, computer equipment can improve the reliability of the acquired second conversion data while ensuring the relevance between the related objects recommended by historical delivery information and the target object, and reduce the error impact caused by the randomness in the second conversion data, thereby ensuring the accuracy of conversion rate analysis.

[0122] The multi-level object association conditions can be set in various ways. For example, the first-level object association condition can be that multiple pieces of information recommend the same type of object, such as mobile phones or computers; the second-level object association condition can be that multiple pieces of information recommend objects belonging to the same brand, such as products under a certain brand.

[0123] In addition to obtaining highly reliable conversion data through multiple object association conditions, the computer device can also ensure the acquisition of sufficient and reliable conversion data by adjusting the time window for acquiring the conversion data. In one possible implementation, when executing step S303, the computer device can execute steps S3034-S3036 (not shown in the figure). Steps S3034-S3036 are a possible implementation of step S303, including:

[0124] S3034: Obtain the conversion data corresponding to the first historical time period for historical campaign information.

[0125] The conversion data corresponding to the first historical period is used to characterize the conversion rate of historical campaign information in the first historical period.

[0126] S3035: Based on the fact that the number of conversions in the first historical period did not reach the preset number, obtain the conversion data corresponding to the historical delivery information in the second historical period.

[0127] Understandably, conversion data from historical periods closer to the current time is more timely. Therefore, this conversion data is more accurate in representing the conversion status of information used to recommend related objects, and the conversion rate analyzed based on this data is also more accurate. Based on this, computer devices can prioritize acquiring conversion data from historical periods closer to the current time.

[0128] Similar to the above, during the conversion data acquisition process, the computer device can determine whether the acquired conversion data is sufficient based on a preset quantity. If the number of conversions corresponding to the historical campaign information in the first historical time period reaches the preset quantity, it indicates that the conversion data corresponding to the historical campaign information in the first historical time period is relatively sufficient and can be used as prior data for conversion rate analysis. Conversely, if the preset quantity is not reached, it indicates that the conversion data corresponding to the historical campaign information in the first historical time period is insufficient, and conversion rate analysis based on this data will introduce significant errors due to randomness. Therefore, the computer device can broaden the time window for acquiring conversion data, acquiring conversion data corresponding to the historical campaign information in a second historical time period to obtain more conversion data.

[0129] The conversion data corresponding to the second historical time period is used to characterize the conversion rate of historical campaign information within that period. The second historical time period includes the first historical time period, and the start point of the second historical time period is earlier than that of the first historical time period. In other words, the second historical time period is a historical period whose start point is earlier than the current time, thus allowing for the acquisition of more conversion data. For example, the first historical time period could be within 4 hours before the current time, and the second historical time period could be within 8 hours of the current time.

[0130] S3036: Based on the historical delivery information, the conversion number corresponding to the historical delivery information in the second historical period reaches a preset number, and the conversion data corresponding to the historical delivery information in the second historical period is determined as the second conversion data.

[0131] Similarly, during the process of adjusting the time window to acquire conversion data, the computer equipment can continuously determine whether sufficient conversion data has been obtained. If the number of conversions corresponding to historical campaign information in the second historical time period reaches a preset number, it indicates that the conversion frequency of historical campaign information in the second historical time period is relatively sufficient, and the corresponding conversion data in the second historical time period is highly reliable. Therefore, the conversion data corresponding to historical campaign information in the second historical time period can be identified as the second conversion data. Conversely, if the preset number is not reached, the time window can be further widened to acquire more conversion data, such as extending it to 12 hours prior to the current moment.

[0132] This approach achieves two main benefits. First, by continuously expanding the historical time window, it ensures a sufficient number of conversions in the acquired conversion data, thereby increasing the reliability of the conversion data and reducing the impact of the randomness of the conversion data on conversion rate analysis. Second, while continuously expanding the time window, it still includes the historical period closest to the current moment, thus ensuring the timeliness of the conversion data to a certain extent and improving the accuracy of conversion rate analysis based on the conversion data.

[0133] The above-described scheme for acquiring transformed data based on multi-layered object association conditions and multiple time windows can be executed individually or in combination. For example... Figure 4 As shown, the object association conditions can include three levels. The first level condition is that the information of the two recommended objects is information placed by the same advertiser. The second level condition is that the two objects correspond to the same object category. The third level condition is that the two objects correspond to the same object brand. The degree of association between the associated objects determined by the three levels of conditions and the target object gradually decreases. The time window can be selected from 4 hours, 24 hours, and 48 hours before the current time, until the historical advertising information that has reached the preset number of conversions is determined.

[0134] Next, we will provide a detailed explanation of the conversion rate analysis process.

[0135] First, in one possible implementation, when executing step S305, the computer device can execute steps S3051-S3053 (not shown in the figure). Steps S3051-S3053 are a possible implementation of step S305, including:

[0136] S3051: Based on the first conversion data, determine the posterior conversion rate corresponding to the information to be analyzed.

[0137] As mentioned above, the first conversion data can characterize the actual conversion situation of the information to be analyzed after deployment. Therefore, computer equipment can determine the posterior conversion rate corresponding to the information to be analyzed based on the first conversion data. The posterior conversion rate is the conversion rate represented by the first conversion data. The posterior conversion rate is a conversion rate determined based on the actual conversion situation and can characterize the actual conversion characteristics of the information to be analyzed. The advantage of the posterior conversion rate is that it is closer to the actual conversion situation, while the disadvantage is that the amount of first conversion data used to determine the posterior conversion rate may be small, and there is a certain degree of random error.

[0138] S3052: Based on the second conversion data, predict the prior conversion rate corresponding to the information to be analyzed.

[0139] Since the second conversion data can characterize the conversion characteristics of recommendation information corresponding to related objects with a high degree of correlation with the target object, and the recommendation information corresponding to objects with a high degree of correlation usually has relatively consistent conversion characteristics, the computer device can predict the prior conversion rate corresponding to the information to be analyzed based on the second conversion data. This prior conversion rate is close to the conversion rate characterized by the second conversion data. The prior conversion rate can be regarded as the conversion rate obtained by the computer device based on the conversion characteristics characterized by historical conversion data. Its advantage is that the amount of second conversion data is relatively sufficient, so it can reliably predict the conversion rate of the information to be analyzed after the target time. Its disadvantage is that the second conversion data is not the conversion data corresponding to the information to be analyzed itself, so the analyzed posterior conversion rate may have a certain deviation from the actual conversion situation of the information to be analyzed.

[0140] S3053: Combine prior and posterior conversion rates to analyze the conversion rate of the information to be analyzed after the target time.

[0141] As can be seen from the above, the disadvantage of prior conversion rate can be compensated by the advantage of posterior conversion rate, and the advantage of prior conversion rate can compensate for the disadvantage of posterior conversion rate. Therefore, computer equipment can combine prior conversion rate and posterior conversion rate to analyze the conversion rate of the information to be analyzed after the target time. This ensures that the analyzed conversion rate can not only match the conversion characteristics of the information used to recommend the target object, but also match the actual conversion situation of the information to be analyzed, thus guaranteeing the accuracy of conversion rate analysis.

[0142] The combination of prior and posterior conversion rates can be varied, for example, as shown in formula (1), by directly taking the average of the two:

[0143]

[0144] Where E(θ) is the analyzed conversion rate, E(prior) is the prior conversion rate, and E(sample) is the posterior conversion rate.

[0145] To more accurately combine prior and posterior conversion rates, one possible implementation is that the computer device can assign corresponding weights to the prior and posterior conversion rates respectively, so as to reasonably control the degree of reference given to the prior and posterior conversion rates when analyzing conversion rates.

[0146] The computer device can first obtain the first weight corresponding to the prior conversion rate and the second weight corresponding to the posterior conversion rate. The first weight is used to control the degree of reference to the prior conversion rate when determining the conversion rate of the information to be analyzed after the target time. The second weight is used to control the degree of reference to the posterior conversion rate when determining the conversion rate of the information to be analyzed after the target time. The degree of reference is usually proportional to the weight.

[0147] When executing step S3053, the computer device may execute step S30531 (not shown in the figure). Step S30531 is a possible implementation of step S3053, including:

[0148] S30531: Combining the first weight, prior conversion rate, second weight, and posterior conversion rate, analyze the conversion rate of the information to be analyzed after the target time.

[0149] When analyzing conversion rates, computer equipment can use a first weight to reference the prior conversion rate and a second weight to reference the posterior conversion rate, thereby reasonably controlling the role of the prior and posterior conversion rates in the conversion rate analysis and ultimately obtaining a more accurate conversion rate.

[0150] The setting of the first and second weights can take many forms, primarily depending on the relative importance of the prior and posterior data in conversion rate analysis. For example, if prior and posterior data are to play an equal role in conversion rate analysis, the first and second weights can be made close to or equal to each other. If the credibility of the prior data is considered greater than that of the posterior data (e.g., when the amount of data in the second conversion data is much larger than that in the first conversion data), the computer device can set the first weight greater than the second weight. If the credibility of the prior data is considered less than that of the posterior data (e.g., when the correlation between the related objects recommended in historical campaign information and the target object is not high enough, or when the amount of data in the first conversion data is large), the computer device can set the second weight greater than the first weight; no limitation is imposed here.

[0151] Specifically, in one possible implementation, the computer device can utilize the beta distribution to combine prior and posterior conversion rates for conversion rate analysis. The conversion events of the information to be analyzed can be approximated as a binomial distribution (0 for clicks without conversion, 1 for clicks with conversion). Therefore, the conversion rate can be represented by the beta distribution Beta(α,β), where α is the first parameter and β is the second parameter. The first and second parameters are used to adjust the distribution pattern represented by the beta distribution, i.e., the distribution pattern of the conversion rate.

[0152] In this implementation, the first conversion data includes the number of objects that clicked on the information to be analyzed before the target time. For example, the first conversion data can be characterized by the number of objects and the number of objects that were converted. When obtaining the first weight corresponding to the prior conversion rate and the second weight corresponding to the posterior conversion rate, the computer device can first determine the first and second parameters corresponding to the beta distribution based on the second conversion data and the number of objects.

[0153] The process of analyzing conversion rate based on beta distribution can be represented by the following formula:

[0154] First, computer equipment can simulate the conversion rate represented by the second conversion data using the beta distribution. The beta distribution can be used to characterize the distribution of the prior conversion rate. When using the beta distribution to characterize the distribution of the prior conversion rate, the corresponding expected conversion rate can be shown in formula (2):

[0155]

[0156] Therefore, if we want E(θ) to be close to the prior conversion rate represented by the second conversion data, the computer device can make the ratio of the first parameter to the second parameter correspond to the ratio between the conversion rate and the non-conversion rate represented by the second conversion data.

[0157] After adding posterior data, the distribution of conversion rate can be approximated by the beta distribution analysis as shown in formula (3):

[0158]

[0159] Where p(θ|X) is the conversion rate distribution after adding posterior data, X is the posterior data (i.e., the first conversion data), and x i For the object that undergoes transformation, denoted as the number of conversions in the posterior data, and 'n' as the number of clicks in the posterior data, i.e., the number of first objects.

[0160] At this point, the expected value of the beta distribution (i.e., the conversion rate of the analyzed information after the target time) E(θ) can be expressed as shown in formula (4):

[0161]

[0162] By simply splitting the fraction in formula (4), we can obtain formula (5):

[0163]

[0164] As can be seen from the above, This can be considered as the prior conversion rate. This can be considered as the posterior conversion rate, and the conversion rate represented by the first conversion data. It can be considered as the first weight. It can be regarded as the second weight, that is, the following formula (6) is obtained:

[0165]

[0166] The ratio between the first weight and the second weight is That is, the ratio of the sum of the first parameter and the second parameter to the number of the first object corresponds to the ratio of the first weight to the second weight.

[0167] Therefore, the computer device can obtain its desired first weight and second weight by adjusting the ratio of the sum of the first parameter and the second parameter to the number of first objects. Furthermore, the computer device can determine the first weight as the ratio of the sum of the first parameter to the sum of the second parameter, and the second weight as the ratio of the number of first objects to the sum of the second parameter, where the first parameter sum is the sum of the first parameter and the second parameter, and the second parameter sum is the sum of the first parameter sum and the number of first objects.

[0168] As can be seen from the above, after setting the required weight ratio, the first weight and the second weight need to satisfy the following formulas (7) and (8):

[0169]

[0170] Where k is the proportion of the first weight in the sum of the first weight and the second weight. For example, when k = 0.5, it can be said that the first weight and the second weight are the same. The conversion rate represented by the second conversion data, prior click For the number of clicked objects in the second conversion data, prior conv This represents the number of converted objects in the second conversion data. Therefore, by obtaining the second conversion data and the required weight ratio, the first and second parameters can be derived in reverse.

[0171] When executing step S3052, the computer device may execute step S30521 (not shown in the figure). Step S30521 is a possible implementation of step S3052, including:

[0172] S30521: Determine the expected value corresponding to the beta distribution as the prior conversion rate.

[0173] This process has already been explained in the above formula explanation, and will not be repeated here.

[0174] In this way, computer equipment can fully utilize the characteristic of beta distribution to analyze parameter distribution by combining prior and posterior data. It can accurately analyze conversion rates by combining second conversion data (prior data) and first conversion data (posterior data), providing a reliable mathematical basis for the aforementioned conversion rate analysis method and ensuring its rationality. Furthermore, since the prior conversion rate and weight ratio can be adjusted by regulating the first and second parameters corresponding to the beta distribution, the parameter analysis method based on beta distribution can accommodate conversion data corresponding to various historical delivery information and meet the needs of various weight ratios, thus exhibiting greater versatility.

[0175] As mentioned above, the first and second weights in this application can adopt various weight ratios. In one possible implementation, since the advantages of the prior conversion rate and the posterior conversion rate can compensate for their respective disadvantages, in order to balance the impact of the prior and posterior conversion rates on conversion rate analysis and allow both to fully leverage their respective advantages, the computer device can make the first and second weights the same. This way, when analyzing the conversion rate of the information to be analyzed after the target time, the prior and posterior conversion rates are referenced with the same level of intensity. As can be seen from the above formula, the computer device can make the sum of the first parameters equal to the number of the first objects to achieve a balance between the first and second weights.

[0176] In addition to predicting conversion rates based on historical conversion data of delivered information, computer devices can also combine information from other dimensions to predict conversion rates. For example, since the delivered information is used to recommend specific objects to a target audience, different target audiences may have different levels of intent to receive information recommending different objects. Therefore, by analyzing the object characteristics of different target audiences and the object characteristics of different objects recommended through the information, it is possible to analyze the patterns of information conversion at the object characteristic dimension.

[0177] Based on this, in one possible implementation, if the information to be analyzed is used to recommend target objects to the object to be recommended, the computer device can also obtain first feature information corresponding to the target object and second feature information corresponding to the object to be recommended. The first feature information is used to characterize the object characteristics corresponding to the target object, such as multi-dimensional features like the target object's color, size, object type, and price; the second feature information is used to characterize the object characteristics corresponding to the object to be recommended, such as behavioral characteristics of the conversion behavior performed by the object to be recommended in historical periods, and the object's own attribute characteristics. The object to be recommended can be any object that can be delivered information, such as a user browsing information.

[0178] Based on the first and second feature information, the computer device can analyze the probability of a target object converting when recommending it to another target object, thus predicting the first conversion rate corresponding to the analyzed information. The first conversion rate is analyzed from two dimensions: the object being recommended and the object targeted by the information delivery, respectively, and their corresponding object characteristics. The limitation of the first conversion rate is that if there is a lack of information on recommending similar target objects, or conversion data for recommending objects similar to the target object, to summarize the information conversion characteristics for such object features, the first conversion rate based on the object characteristics of the target object and the other target object may contain some errors.

[0179] When executing step S305, the computer device may execute steps S3054-S3055 (not shown in the figure). Steps S3054-S3055 are one possible implementation of step S305, including:

[0180] S3054: Combining the first conversion data and the second conversion data, analyze the second conversion rate of the information to be analyzed after the target time.

[0181] The analysis process for the second conversion rate has been detailed above and will not be repeated here. The second conversion rate is the conversion rate analyzed from two data dimensions: historical conversion data and actual conversion data. It allows for the analysis of conversion rates from the perspective of conversion rates corresponding to similar information and actual conversion rates corresponding to the information to be analyzed.

[0182] S3055: Adjust the second conversion rate based on the difference between the first conversion rate and the second conversion rate to obtain the conversion rate of the information to be analyzed after the target time, and adjust it to reduce the difference between the first conversion rate and the second conversion rate.

[0183] Since the second conversion rate has higher accuracy, the computer device can adjust the first conversion rate based on the difference between the first and second conversion rates to reduce the difference between the first and second conversion rates, thereby correcting the error of the first conversion rate and using the corrected conversion rate as the conversion rate corresponding to the information to be analyzed after the target time.

[0184] The advantage of this analytical approach lies in its ability to analyze conversion rates by combining the first conversion rate with the characteristics of both the target audience and the intended recipient. Furthermore, combining the second conversion rate allows for analysis based on the conversion patterns of similar information over historical periods and the actual conversion performance of the information being analyzed. Therefore, combining both rates expands the analytical dimensions of conversion rate analysis. Simultaneously, the conversion data from historical campaigns used in the second conversion rate analysis compensates for errors caused by the sparsity of conversion data in the first conversion rate analysis, thereby improving the comprehensiveness and accuracy of the analysis.

[0185] The adjustment method can be represented by the following formula:

[0186] First, the computer device can calculate the difference parameter pcvr between the first conversion rate and the second conversion rate. bias As shown in the following formula (9):

[0187]

[0188] Where CVR is the second conversion rate, PCVR is the first conversion rate, and ΔT is the time window for conversion rate analysis, which is the historical period before the target time.

[0189] Then, the computer device can calculate the adjustment factor f for adjusting the first conversion rate based on the difference parameter, as shown in formula (10):

[0190]

[0191] The adjustment factor is calculated using a fractional method, which can effectively control the adjustment intensity of the first conversion rate based on the difference parameter, and avoid over-adjustment of the first conversion rate.

[0192] Finally, the computer device can adjust the first conversion rate based on this adjustment factor, as shown in formula (11):

[0193] PCVR * =pcvr×f (11)

[0194] Among them, PCVR * The adjusted conversion rate is the conversion rate of the analyzed information after the target time.

[0195] The methods for analyzing conversion rates by incorporating object characteristics can be varied. For example, conversion rates can be predicted using a conversion rate prediction model. In one possible implementation, a computer device can train the desired conversion rate prediction model in the following way.

[0196] Computer equipment can first obtain historical sample delivery information and the corresponding sample conversion rate. The historical sample delivery information can be of any type. This historical sample delivery information is used to recommend sample objects to the sample to be recommended objects. The sample conversion rate is the conversion rate obtained by delivering historical sample delivery information in the historical period, that is, the actual conversion rate corresponding to the historical sample delivery information.

[0197] Then, the computer device can acquire first sample feature information and second sample feature information. The first sample feature information is used to characterize the object features corresponding to the sample object, and the second sample feature information is used to characterize the object features corresponding to the sample object to be recommended. This feature information can be extracted by the computer device itself or extracted by other devices and sent to the computer device; this is not limited here. Based on the first and second sample feature information, the computer device can predict the undetermined conversion rate corresponding to the historical delivery information of the sample using an initial conversion rate prediction model. This undetermined conversion rate is the conversion rate corresponding to the historical delivery information of the sample analyzed by the initial conversion rate prediction model based on the object features.

[0198] Since the sample conversion rate represents the actual conversion rate corresponding to the historical delivery information of the sample, the difference between the sample conversion rate and the undetermined conversion rate can characterize the prediction accuracy of the initial conversion rate prediction model when predicting conversion rates based on object characteristics. This prediction accuracy is inversely correlated with the difference. Therefore, the computer device can adjust the model parameters corresponding to the initial conversion rate prediction model based on the difference between the undetermined conversion rate and the sample conversion rate to obtain the conversion rate prediction model. During the adjustment process, the computer device can continuously reduce the difference between the determined undetermined conversion rate and the sample conversion rate, thereby enabling the initial conversion rate prediction model to learn how to accurately predict the conversion rate corresponding to the delivery information based on object characteristics. Consequently, the conversion rate prediction model can be used for conversion rate analysis based on object characteristics.

[0199] When predicting the first conversion rate corresponding to the information to be analyzed based on the first feature information and the second feature information, the processing device can predict the first conversion rate corresponding to the information to be analyzed based on the first feature information and the second feature information through the above conversion rate prediction model. In this way, the powerful learning and data processing capabilities of the model can be utilized to fully and effectively analyze the first conversion rate in the dimension of object features, thereby ensuring the accuracy of the first conversion rate.

[0200] like Figure 5 As shown, after training the conversion rate prediction model using training samples, the object feature information corresponding to the target object recommended by the information to be analyzed and the object feature information corresponding to the recommended object when the information to be analyzed is delivered can be input into the conversion rate prediction model, and the first conversion rate predicted by the model is output.

[0201] Meanwhile, the computer equipment can use the conversion data corresponding to the historical delivery information as prior data and the conversion data corresponding to the information to be analyzed before the target time as posterior data. By combining the prior data and the posterior data, the second conversion rate corresponding to the information to be analyzed after the target time can be analyzed. Based on the first conversion rate and the second conversion rate, the adjustment factor is determined. Finally, the first conversion rate is adjusted based on the adjustment factor to obtain the final predicted conversion rate corresponding to the information to be analyzed after the target time.

[0202] In order to more reasonably adjust the first conversion rate, when executing step S3055, the computer device can execute step S30551 (not shown in the figure). Step S30551 is a possible implementation of step S3055, including:

[0203] S30551: Based on the difference between the first conversion rate and the second conversion rate and the adjustment parameter, adjust the second conversion rate to obtain the conversion rate of the information to be analyzed after the target time.

[0204] The computer device can set adjustment parameters for the adjustment process of the first conversion rate. These parameters control the adjustment strength of the second conversion rate based on the difference between the first and second conversion rates. For example, if the computer device believes that the accuracy of the conversion rate determined based on prior and posterior data is greater than the accuracy of the conversion rate analyzed based on object features, it can increase the adjustment strength by adjusting the adjustment parameters, making the adjusted conversion rate closer to the second conversion rate. This allows the prior and posterior data to have a greater influence in the conversion rate analysis. Conversely, it can decrease the adjustment strength by adjusting the adjustment parameters, making the adjusted conversion rate closer to the first conversion rate. This allows the object features to have a greater influence in the conversion rate analysis. In this way, the computer device can more rationally control the adjustment method of the first conversion rate based on the second conversion rate, enabling the conversion rate analysis process to refer more strongly to data dimensions with higher credibility, thereby further improving the accuracy of conversion rate analysis when combining multi-dimensional data.

[0205] Understandably, the conversion rate of the delivered information may change over time. For example, when the conversion rate is low, the promotion is ineffective, and the conversion rate may continue to decline. Conversely, when the conversion rate is high, it indicates good promotion effectiveness, and the conversion rate may continue to rise. Since the conversion rate is closely related to the information value of the delivered information, computer equipment can continuously analyze the conversion rate of the information to be analyzed based on changes in the delivery time to more accurately analyze the information value of the delivered information.

[0206] For example, a computer device can preset a time interval, and analyze the conversion rate of the information to be analyzed in subsequent periods every preset time interval, so as to update the information value of the information to be analyzed in a timely manner.

[0207] When performing step S302, the computer device may perform step S3021 (not shown in the figure). Step S3021 is a possible implementation of step S302, including:

[0208] S3021: Based on the time interval between the target time and the initial click time corresponding to the information to be analyzed reaching n times the preset time interval, obtain the first conversion data corresponding to the information to be analyzed.

[0209] The initial click time is the moment when the information to be analyzed is first clicked. Since conversion rate is the percentage of clicked items that convert, valid conversion data can only be obtained after an item clicks on the information to be analyzed. The computer device can analyze the conversion rate of the information to be analyzed in subsequent time periods, starting from the initial click time, at preset time intervals. For example, the first conversion data can be used to characterize the conversion rate of the information to be analyzed between the initial click time and the target time, reflecting the actual conversion situation of the information to be analyzed from the initial click time to the target time, where n is a positive integer. When the (n+1)th preset time interval is reached, the computer device will also trigger an analysis of the conversion rate.

[0210] In this way, computer equipment can continuously monitor the conversion rate of the information to be analyzed in subsequent time periods after it has been delivered. This allows for ongoing measurement of the information value generated after delivery, enabling timely detection of significant changes in conversion rates or information value. Consequently, the deliverer can adjust their delivery strategy accordingly to achieve better results. For example, if conversion rate analysis predicts a significant drop in the conversion rate of the information after a certain point in time, the deliverer can adjust the delivery method to increase the prominence of the information and improve the corresponding conversion rate; alternatively, the deliverer can cancel the delivery to avoid wasting excessive resources on information with low conversion value.

[0211] Based on this, this application also provides various processing methods for subsequent processing based on the analyzed conversion rates. For example, in one possible implementation, the computer device can preset a difference threshold, which is used to measure whether there is a large difference between two conversion rates.

[0212] After analyzing the conversion rate, the computer device can compare the analyzed conversion rate with the conversion rate of the information to be analyzed after a historical time. The historical time is the previous time that is separated from the target time by a preset time interval. The conversion rate of the information to be analyzed after the historical time is obtained based on the conversion rate of the information to be analyzed before the historical time. The conversion rate after the historical time is the conversion rate obtained from the previous conversion rate analysis.

[0213] If the difference between the conversion rate of the information to be analyzed after the target time and the conversion rate of the information to be analyzed after a historical time exceeds a difference threshold, it indicates that the conversion rate of the information to be analyzed will change significantly after the target time, the conversion value of the information to be analyzed will fluctuate significantly, and the delivery of the information to be analyzed is likely to be abnormal. Based on this, the computer device can generate a prompt message to indicate that the conversion rate of the information to be analyzed has changed abnormally. In this application, abnormal change refers to a change in conversion rate that is too large.

[0214] This notification can promptly alert advertisers that the conversion rate of the information being analyzed is about to fluctuate significantly. This allows advertisers to adjust their information delivery strategies in a timely manner to obtain more information conversion value or reduce the waste of ineffective information delivery resources, thereby bringing better information delivery results to advertisers.

[0215] Furthermore, in one possible implementation, the computer device can automatically adjust the delivery method of the information to be analyzed based on the analyzed conversion rate. For example, the computer device can adjust the delivery method of the information to be analyzed after the target time based on the conversion rate corresponding to the information after the target time. This delivery method determines the prominence of displaying the information to be analyzed. The higher the prominence, the more times the information to be analyzed will be exposed, which usually improves the click-through rate and conversion rate of the information to a certain extent. The adjustment of the delivery method can include various methods, such as changing the delivery frequency of the information to be analyzed, the duration of each delivery, and adding delivery channels, etc., which are not limited here.

[0216] The logic by which computer devices adjust ad delivery based on conversion rates can include various approaches. For example, under one logic, the computer device can make the salience of the adjusted ad delivery method positively correlated with the conversion rate; that is, the higher the analyzed conversion rate, the higher the salience of the information being analyzed can be displayed. The advantage of this adjustment logic is that it allows information with high conversion rates to receive more full exposure, thus maximizing its high conversion value; while reducing the ad delivery intensity for information with low conversion rates, thereby minimizing unnecessary resource consumption.

[0217] Under the second logic, computer devices can provide information with higher exposure through information with low conversion rates. The advantage is that they can improve the conversion effect of information with poor conversion results in a timely manner, so that the target audience recommended by the information can receive effective publicity and promotion.

[0218] Therefore, computer equipment can automatically adjust the delivery method based on conversion rate analysis results, so that the information delivery effect can meet the diverse information delivery needs. At the same time, the computer equipment can also improve the efficiency of the delivery method adjustment, so that the delivery method can change with the future conversion rate, ultimately achieving a better information conversion effect.

[0219] To facilitate understanding of the technical solution provided in this application, the conversion rate analysis method provided in this application will be introduced next in conjunction with a practical application scenario.

[0220] See Figure 6 , Figure 6 A flowchart illustrating a conversion rate analysis method in a practical application scenario provided in this application embodiment, the method comprising:

[0221] S601: Extract the first feature information corresponding to the target object to be recommended and the second feature information corresponding to the object to be recommended.

[0222] Among them, the information to be analyzed is the information of recommending target objects to the object to be recommended. The first feature information is used to characterize the object features of the target object, and the second feature information is used to characterize the object features corresponding to the object to be recommended.

[0223] S602: Based on the first feature information and the second feature information, predict the first conversion rate corresponding to the information to be analyzed using the conversion rate prediction model.

[0224] The first conversion rate is the conversion rate of the information to be analyzed from the dimension of object features, and it is easily affected by the error caused by the sparsity of the converted data.

[0225] S603: Based on multi-level object association conditions and multiple historical time periods, obtain the historical delivery information corresponding to the information to be analyzed.

[0226] Computer devices can obtain relatively abundant historical campaign information by using multi-level object association conditions with varying requirements for object association, as well as multiple historical time periods with different intervals from the current time. For example, the computer device can continuously adjust object association and historical time periods until it obtains 10 pieces of historical campaign information, each corresponding to a conversion count of 10 times.

[0227] S604: Obtain the second conversion data corresponding to the historical delivery information.

[0228] The second conversion data is used to characterize the conversion rate of historical delivery information in historical time periods, and is used as prior data for conversion rate analysis.

[0229] S605: Obtain the first transformation data corresponding to the information to be analyzed before the target time.

[0230] The first conversion data is used to characterize the actual conversion rate of the information to be analyzed after deployment and before the target time, and is used as posterior data for conversion rate analysis.

[0231] S606: Determine the first parameter and the second parameter corresponding to the beta distribution based on the first transformation data and the second transformation data.

[0232] Among them, the first parameter and the second parameter need to satisfy the conversion rate represented by the second conversion data, and at the same time, the first weight and the second weight determined based on the beta distribution need to be close to each other, for example, they can be the same. Based on this, the first parameter and the second parameter that meet the requirements can be determined.

[0233] S607: Determine the posterior conversion rate and the prior conversion rate based on the first conversion data using the beta distribution.

[0234] The calculation method for prior and posterior conversion rates using beta distribution has been detailed above and will not be repeated here. Prior conversion rate is used to analyze conversion rates based on the conversion characteristics of historical delivery information similar to the recommended targets in the information being analyzed, while posterior conversion rate is used to analyze conversion rates based on the actual conversion characteristics of the information being analyzed. Besides beta distribution, computer devices can also perform conversion rate distribution analysis using other mathematical methods, such as least squares, which are not limited here.

[0235] S608: Determine the second conversion rate by combining the prior conversion rate and the posterior conversion rate.

[0236] S609: Determine the adjustment factor for adjusting the first conversion rate based on the difference between the first conversion rate and the second conversion rate.

[0237] This adjustment factor is used to characterize the adjustment method of the first conversion rate. Its determination process has been introduced in the above formula and will not be repeated here.

[0238] S610: Adjust the first conversion rate based on the adjustment factor to obtain the conversion rate of the information to be analyzed after the target time.

[0239] By adjusting the first conversion rate, it can be made close to the second conversion rate, thus enabling the analysis of the conversion rate by combining the above multiple dimensions.

[0240] S611: Based on the conversion rate of the information to be analyzed after the target time, adjust the delivery method of the information to be analyzed after the target time.

[0241] As can be seen from the above, this application has the following technical effects:

[0242] 1. By combining the first and second conversion data, a more accurate analysis of the conversion rate of the information to be analyzed after the target time can be achieved from both prior and posterior data dimensions, even in scenarios where conversion data is sparse. Simultaneously, the second conversion data can also characterize the commonalities in information conversion among multiple pieces of information related to the object categories associated with the recommended target object. This makes the conversion rate analysis method more targeted to specific object categories, further improving the accuracy of the conversion rate analysis.

[0243] 2. By combining the first conversion rate predicted by the conversion rate prediction model, the conversion rate can be further analyzed by combining the dimensions of object characteristics, thereby further improving the data dimensions referenced in the conversion rate analysis and improving the accuracy of the conversion rate analysis.

[0244] 3. Since the second conversion rate uses rich prior data for analysis, adjusting the first conversion rate based on the difference between the second and first conversion rates can compensate for the conversion rate prediction error caused by the conversion rate prediction model in scenarios with sparse conversion data, thereby improving the accuracy of conversion rate analysis.

[0245] 4. This application uses multi-level object association conditions and multiple historical time periods to obtain historical campaign information, which can ensure the sufficiency and reliability of prior data. While ensuring the accuracy of conversion rate analysis, it solves the problem of sparse conversion data to a certain extent.

[0246] 5. This application can automatically adjust the subsequent information delivery method based on the analyzed conversion rate to ensure that the final information to be analyzed can achieve a more effective conversion effect and reduce unnecessary information delivery resource consumption.

[0247] Based on the conversion rate analysis method provided in the above embodiments, this application also provides a conversion rate analysis device, see [link to relevant documentation]. Figure 7 , Figure 7 This application provides a structural block diagram of a conversion rate analysis device 700, which includes a first acquisition unit 701, a second acquisition unit 702, a determination unit 703, a third acquisition unit 704, and an analysis unit 705.

[0248] The first acquisition unit 701 is used to acquire information to be analyzed, the information to be analyzed being used to recommend target objects;

[0249] The second acquisition unit 702 is used to acquire the first conversion data corresponding to the information to be analyzed. The first conversion data is used to characterize the conversion rate corresponding to the information to be analyzed before the target time. The conversion rate is used to characterize the proportion of objects that perform conversion behavior for the objects recommended by the corresponding information among the objects that click on the corresponding information.

[0250] The determining unit 703 is used to determine the historical delivery information corresponding to the information to be analyzed. The historical delivery information is used to recommend associated objects. The associated objects are objects that meet the object association conditions with the target object.

[0251] The third acquisition unit 704 is used to acquire the second conversion data corresponding to the historical delivery information, and the second conversion data is used to characterize the conversion rate corresponding to the historical delivery information.

[0252] The analysis unit 705 is used to combine the first conversion data and the second conversion data to analyze the conversion rate of the information to be analyzed after the target time.

[0253] In one possible implementation, the conversion count corresponding to the historical delivery information is greater than a preset number, where the conversion count is the number of objects that perform conversion actions on the objects recommended by the corresponding information among the objects that are clicked.

[0254] In one possible implementation, the determining unit 703 is specifically used for:

[0255] The first object association condition is used as the object association condition to obtain the first historical delivery information. The first historical delivery information is used to recommend the first object. The first object and the target object satisfy the first object association condition.

[0256] Based on the fact that the conversion number corresponding to the first historical delivery information has not reached the preset number, the second object association condition is used as the object association condition, and the second historical delivery information is obtained. The second historical delivery information is used to recommend the second object. The second object and the target object meet the second object association condition. The degree of association between the second object and the target object is less than the degree of association between the first object and the target object.

[0257] If the conversion count corresponding to the second historical delivery information reaches the preset number, the second historical delivery information is determined as the historical delivery information.

[0258] In one possible implementation, the third acquisition unit 704 is specifically used for:

[0259] Obtain the conversion data corresponding to the historical delivery information in the first historical time period, and the conversion data corresponding to the first historical time period is used to characterize the conversion rate of the historical delivery information in the first historical time period.

[0260] Based on the fact that the number of conversions corresponding to the historical delivery information in the first historical period did not reach the preset number, the conversion data corresponding to the historical delivery information in the second historical period is obtained. The conversion data corresponding to the second historical period is used to characterize the conversion rate corresponding to the historical delivery information in the second historical period. The second historical period includes the first historical period, and the start time of the second historical period is earlier than the start time of the first historical period.

[0261] Based on the fact that the conversion number corresponding to the historical delivery information in the second historical period reaches the preset number, the conversion data corresponding to the historical delivery information in the second historical period is determined as the second conversion data.

[0262] In one possible implementation, the analysis unit 705 is specifically used for:

[0263] Based on the first conversion data, the posterior conversion rate corresponding to the information to be analyzed is determined, wherein the posterior conversion rate is the conversion rate represented by the first conversion data;

[0264] Based on the second conversion data, predict the prior conversion rate corresponding to the information to be analyzed;

[0265] By combining the prior conversion rate and the posterior conversion rate, the conversion rate of the information to be analyzed after the target time is analyzed.

[0266] In one possible implementation, the device further includes a fourth acquisition unit:

[0267] The fourth acquisition unit is used to acquire a first weight corresponding to the prior conversion rate and a second weight corresponding to the posterior conversion rate. The first weight is used to control the degree of reference to the prior conversion rate when determining the conversion rate of the information to be analyzed after the target time. The second weight is used to control the degree of reference to the posterior conversion rate when determining the conversion rate of the information to be analyzed after the target time.

[0268] The analysis unit 705 is specifically used for:

[0269] By combining the first weight, the prior conversion rate, the second weight, and the posterior conversion rate, the conversion rate of the information to be analyzed after the target time is analyzed.

[0270] In one possible implementation, the first conversion data includes the first number of objects that clicked on the information to be analyzed before the target time, and the fourth acquisition unit is specifically used for:

[0271] Based on the second conversion data and the number of the first objects, a first parameter and a second parameter corresponding to the beta distribution are determined. The ratio of the first parameter to the second parameter corresponds to the ratio between the conversion rate and the non-conversion rate represented by the second conversion data. The ratio of the first parameter to the second parameter and the first object number corresponds to the ratio of the first weight to the second weight. The first parameter and the second parameter are used to adjust the distribution mode represented by the beta distribution. The beta distribution is used to represent the distribution mode of the prior conversion rate.

[0272] The ratio of the sum of the first parameter to the sum of the second parameter is determined as the first weight, and the ratio of the number of the first object to the sum of the second parameter is determined as the second weight. The sum of the first parameter is the sum of the first parameter and the second parameter, and the sum of the second parameter is the sum of the first parameter and the number of the first object.

[0273] The analysis unit 705 is specifically used for:

[0274] The expected value corresponding to the beta distribution is determined as the prior conversion rate.

[0275] In one possible implementation, the sum of the first parameters equals the number of the first objects.

[0276] In one possible implementation, the information to be analyzed is used to recommend the target object to the object to be recommended, and the device further includes a fifth acquisition unit and a first prediction unit:

[0277] The fifth acquisition unit is used to acquire first feature information corresponding to the target object and second feature information corresponding to the object to be recommended. The first feature information is used to characterize the object feature corresponding to the target object, and the second feature information is used to characterize the object feature corresponding to the object to be recommended.

[0278] The first prediction unit is configured to predict the first conversion rate corresponding to the information to be analyzed based on the first feature information and the second feature information;

[0279] The analysis unit 705 is specifically used for:

[0280] By combining the first conversion data and the second conversion data, the second conversion rate corresponding to the information to be analyzed after the target time is analyzed;

[0281] The second conversion rate is adjusted based on the difference between the first conversion rate and the second conversion rate to obtain the conversion rate of the information to be analyzed after the target time. The adjustment is used to reduce the difference between the first conversion rate and the second conversion rate.

[0282] In one possible implementation, the device further includes a sixth acquisition unit, a seventh acquisition unit, a second prediction unit, and a first adjustment unit:

[0283] The sixth acquisition unit is used to acquire historical sample delivery information and the sample conversion rate corresponding to the historical sample delivery information. The historical sample delivery information is used to recommend sample objects to sample objects to be recommended. The sample conversion rate is the conversion rate obtained by delivering the historical sample delivery information in a historical period.

[0284] The seventh acquisition unit is used to acquire first sample feature information and second sample feature information. The first sample feature information is used to characterize the object features corresponding to the sample object, and the second sample feature information is used to characterize the object features corresponding to the sample object to be recommended.

[0285] The second prediction unit is used to predict the undetermined conversion rate corresponding to the historical delivery information of the sample based on the first sample feature information and the second sample feature information through an initial conversion rate prediction model;

[0286] The first adjustment unit is used to adjust the model parameters corresponding to the initial conversion rate prediction model according to the difference between the undetermined conversion rate and the sample conversion rate, so as to obtain the conversion rate prediction model;

[0287] The first prediction unit is specifically used for:

[0288] Based on the first feature information and the second feature information, the conversion rate prediction model is used to predict the first conversion rate corresponding to the information to be analyzed.

[0289] In one possible implementation, the analysis unit 705 is specifically used for:

[0290] Based on the difference between the first conversion rate and the second conversion rate and the adjustment parameter, the second conversion rate is adjusted to obtain the conversion rate of the information to be analyzed after the target time. The adjustment parameter is used to control the adjustment intensity of the second conversion rate based on the difference between the first conversion rate and the second conversion rate.

[0291] In one possible implementation, the second acquisition unit 702 is specifically used for:

[0292] Based on a preset time interval of n times between the target time and the initial click time corresponding to the information to be analyzed, first conversion data corresponding to the information to be analyzed is obtained. The first conversion data is used to characterize the conversion rate of the information to be analyzed between the initial click time and the target time. The initial click time is the time when the information to be analyzed is first clicked, and n is a positive integer.

[0293] In one possible implementation, the apparatus further includes a generation unit:

[0294] The generation unit is configured to generate a prompt message based on the fact that the difference between the conversion rate of the information to be analyzed after the target time and the conversion rate of the information to be analyzed after a historical time is greater than a difference threshold. The prompt message is used to indicate that the conversion rate of the information to be analyzed has changed abnormally. The historical time is the previous time that is separated from the target time by the preset time interval. The conversion rate of the information to be analyzed after the historical time is obtained based on the conversion rate of the information to be analyzed before the historical time.

[0295] In one possible implementation, the device further includes a second adjustment unit:

[0296] The second adjustment unit is used to adjust the delivery method of the information to be analyzed after the target time according to the conversion rate of the information to be analyzed after the target time. The delivery method is used to determine the prominence of displaying the information to be analyzed.

[0297] This application also provides a computer device; please refer to [link to relevant documentation]. Figure 8 As shown, the computer device can be a terminal device; for example, a mobile phone can be used as a terminal device.

[0298] Figure 8 This diagram illustrates a partial structural representation of a mobile phone related to the terminal device provided in this embodiment. (Reference) Figure 8 The mobile phone includes components such as a radio frequency (RF) circuit 710, a memory 720, an input unit 730, a display unit 740, a sensor 750, an audio circuit 760, a wireless Fidelity (WiFi) module 770, a processor 780, and a power supply 790. Those skilled in the art will understand that... Figure 8 The mobile phone structure shown does not constitute a limitation on the mobile phone and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0299] The following is combined with Figure 8 A detailed introduction to each component of a mobile phone:

[0300] RF circuit 710 can be used for receiving and transmitting signals during information transmission or calls. Specifically, it receives downlink information from the base station and processes it with processor 780; additionally, it transmits uplink data to the base station. Typically, RF circuit 710 includes, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a low-noise amplifier (LNA), and a duplexer. Furthermore, RF circuit 710 can also communicate wirelessly with networks and other devices. The aforementioned wireless communication can use any communication standard or protocol, including but not limited to Global System for Mobile Communications (GSM), General Packet Radio Service (GPRS), Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA), Long Term Evolution (LTE), email, and Short Messaging Service (SMS).

[0301] The memory 720 can be used to store software programs and modules. The processor 780 executes various mobile phone functions and data processing by running the software programs and modules stored in the memory 720. The memory 720 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, applications required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 720 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0302] The input unit 730 can be used to receive input numerical or character information, and to generate key signal inputs related to user settings and function control of the mobile phone. Specifically, the input unit 730 may include a touch panel 731 and other input devices 732. The touch panel 731, also known as a touch screen, can collect touch operations performed by the user on or near it (such as operations performed by the user using a finger, stylus, or any suitable object or accessory on or near the touch panel 731), and drive the corresponding connected devices according to a pre-set program. Optionally, the touch panel 731 may include two parts: a touch detection device and a touch controller. The touch detection device detects the user's touch position and the signal generated by the touch operation, and transmits the signal to the touch controller; the touch controller receives touch information from the touch detection device, converts it into touch point coordinates, and sends it to the processor 780, and can also receive and execute commands sent by the processor 780. In addition, the touch panel 731 can be implemented using various types such as resistive, capacitive, infrared, and surface acoustic wave. In addition to the touch panel 731, the input unit 730 may also include other input devices 732. Specifically, other input devices 732 may include, but are not limited to, one or more of the following: physical keyboard, function keys (such as volume control buttons, power buttons, etc.), trackball, mouse, joystick, etc.

[0303] The display unit 740 can be used to display information input by the user or information provided to the user, as well as various menus of the mobile phone. The display unit 740 may include a display panel 741, which may optionally be configured as a Liquid Crystal Display (LCD), Organic Light-Emitting Diode (OLED), or similar display panel. Further, a touch panel 731 may cover the display panel 741. When the touch panel 731 detects a touch operation on or near it, it transmits the information to the processor 780 to determine the type of touch event. Subsequently, the processor 780 provides corresponding visual output on the display panel 741 based on the type of touch event. Although in Figure 8 In this embodiment, the touch panel 731 and the display panel 741 are two separate components to realize the input and output functions of the mobile phone. However, in some embodiments, the touch panel 731 and the display panel 741 can be integrated to realize the input and output functions of the mobile phone.

[0304] The mobile phone may also include at least one sensor 750, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor may include an ambient light sensor and a proximity sensor. The ambient light sensor can adjust the brightness of the display panel 741 according to the ambient light level, and the proximity sensor can turn off the display panel 741 and / or backlight when the phone is moved to the ear. As a type of motion sensor, an accelerometer sensor can detect the magnitude of acceleration in various directions (generally three axes). When stationary, it can detect the magnitude and direction of gravity and can be used for applications that recognize the phone's posture (such as landscape / portrait switching, related games, magnetometer posture calibration), vibration recognition-related functions (such as pedometer, taps), etc. Other sensors that may be configured in the mobile phone, such as gyroscopes, barometers, hygrometers, thermometers, and infrared sensors, will not be described in detail here.

[0305] Audio circuit 760, speaker 761, and microphone 762 provide an audio interface between the user and the mobile phone. Audio circuit 760 converts received audio data into electrical signals and transmits them to speaker 761, where speaker 761 converts them into sound signals for output. On the other hand, microphone 762 converts collected sound signals into electrical signals, which are received by audio circuit 760, converted into audio data, and then processed by processor 780 before being transmitted via RF circuit 710 to, for example, another mobile phone, or the audio data can be output to memory 720 for further processing.

[0306] WiFi is a short-range wireless transmission technology. Through the WiFi module 770, mobile phones can help users send and receive emails, browse web pages, and access streaming media, providing users with wireless broadband internet access. Although Figure 8 The WiFi module 770 is shown, but it is understood that it is not an essential component of a mobile phone and can be omitted as needed without changing the essence of the invention.

[0307] The processor 780 is the control center of the mobile phone, connecting various parts of the phone through various interfaces and lines. It executes software programs and / or modules stored in the memory 720, and calls data stored in the memory 720 to perform various functions and process data, thereby performing overall detection of the phone. Optionally, the processor 780 may include one or more processing units; preferably, the processor 780 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into the processor 780.

[0308] The mobile phone also includes a power supply 790 (such as a battery) that supplies power to various components. Preferably, the power supply can be logically connected to the processor 780 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system.

[0309] Although not shown, mobile phones may also include a camera, Bluetooth module, etc., which will not be described in detail here.

[0310] In this embodiment, the processor 780 included in the terminal device also has the following functions:

[0311] Obtain information to be analyzed, which is used to recommend target objects;

[0312] Obtain the first conversion data corresponding to the information to be analyzed. The first conversion data is used to characterize the conversion rate corresponding to the information to be analyzed before the target time. The conversion rate is used to characterize the proportion of objects that perform conversion behavior for the objects recommended by the corresponding information among the objects that click on the corresponding information.

[0313] Determine the historical delivery information corresponding to the information to be analyzed. The historical delivery information is used to recommend associated objects. The associated objects are objects that meet the object association conditions with the target object.

[0314] Obtain the second conversion data corresponding to the historical delivery information, wherein the second conversion data is used to characterize the conversion rate corresponding to the historical delivery information;

[0315] By combining the first conversion data and the second conversion data, the conversion rate of the information to be analyzed after the target time is analyzed.

[0316] This application also provides a server; please refer to [link / reference]. Figure 9 As shown, Figure 9 This is a structural diagram of a server 800 provided in an embodiment of this application. The server 800 can vary significantly due to different configurations or performance. It may include one or more Central Processing Units (CPUs) 822 (e.g., one or more processors) and a memory 832, and one or more storage media 830 (e.g., one or more mass storage devices) for storing application programs 842 or data 844. The memory 832 and storage media 830 can be temporary or persistent storage. The program stored in the storage media 830 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the server. Furthermore, the CPU 822 may be configured to communicate with the storage media 830 and execute the series of instruction operations in the storage media 830 on the server 800.

[0317] Server 800 may also include one or more power supplies 826, one or more wired or wireless network interfaces 850, one or more input / output interfaces 858, and / or one or more operating systems 841, such as Windows Server. TM Mac OS X TM Unix TM Linux TM FreeBSD TM etc.

[0318] The steps performed by the server in the above embodiments can be based on Figure 9 The server structure shown.

[0319] This application also provides a computer-readable storage medium for storing a computer program that executes any one of the conversion rate analysis methods described in the foregoing embodiments.

[0320] This application also provides a computer program product including a computer program, which, when run on a computer device, causes the computer device to execute the conversion rate analysis method described in any of the above embodiments.

[0321] It is understood that in the specific implementation of this application, user information (such as conversion data) and other related data are involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0322] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium can be at least one of the following media: read-only memory (ROM), RAM, magnetic disk, or optical disk, etc., and other media capable of storing program code.

[0323] It should be noted that the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for the device and system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments. The device and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of the solution in this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0324] The above description is merely one specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A conversion rate analysis method, characterized in that, The method includes: Obtain information to be analyzed, which is used to recommend target objects; Obtain the first conversion data corresponding to the information to be analyzed. The first conversion data is used to characterize the conversion rate corresponding to the information to be analyzed before the target time. The conversion rate is used to characterize the proportion of objects that perform conversion behavior for the objects recommended by the corresponding information among the objects that click on the corresponding information. Determine the historical delivery information corresponding to the information to be analyzed. The historical delivery information is used to recommend associated objects. The associated objects are objects that meet the object association conditions with the target object. Obtain the second conversion data corresponding to the historical delivery information, wherein the second conversion data is used to characterize the conversion rate corresponding to the historical delivery information; By combining the first conversion data and the second conversion data, the conversion rate of the information to be analyzed after the target time is analyzed.

2. The method according to claim 1, characterized in that, The conversion count corresponding to the historical delivery information is greater than the preset number. The conversion count is the number of objects that perform conversion behavior for the objects recommended by the corresponding information among the objects clicked.

3. The method according to claim 2, characterized in that, The step of determining the historical delivery information corresponding to the information to be analyzed includes: The first object association condition is used as the object association condition to obtain the first historical delivery information. The first historical delivery information is used to recommend the first object. The first object and the target object satisfy the first object association condition. Based on the fact that the conversion number corresponding to the first historical delivery information has not reached the preset number, the second object association condition is used as the object association condition, and the second historical delivery information is obtained. The second historical delivery information is used to recommend the second object. The second object and the target object meet the second object association condition. The degree of association between the second object and the target object is less than the degree of association between the first object and the target object. If the conversion count corresponding to the second historical delivery information reaches the preset number, the second historical delivery information is determined as the historical delivery information.

4. The method according to claim 2, characterized in that, The step of obtaining the second conversion data corresponding to the historical delivery information includes: Obtain the conversion data corresponding to the historical delivery information in the first historical time period, and the conversion data corresponding to the first historical time period is used to characterize the conversion rate of the historical delivery information in the first historical time period. Based on the fact that the number of conversions corresponding to the historical delivery information in the first historical period did not reach the preset number, the conversion data corresponding to the historical delivery information in the second historical period is obtained. The conversion data corresponding to the second historical period is used to characterize the conversion rate corresponding to the historical delivery information in the second historical period. The second historical period includes the first historical period, and the start time of the second historical period is earlier than the start time of the first historical period. Based on the fact that the conversion number corresponding to the historical delivery information in the second historical period reaches the preset number, the conversion data corresponding to the historical delivery information in the second historical period is determined as the second conversion data.

5. The method according to claim 1, characterized in that, The step of combining the first conversion data and the second conversion data to analyze the conversion rate of the information to be analyzed after the target time includes: Based on the first conversion data, the posterior conversion rate corresponding to the information to be analyzed is determined, wherein the posterior conversion rate is the conversion rate represented by the first conversion data; Based on the second conversion data, predict the prior conversion rate corresponding to the information to be analyzed; By combining the prior conversion rate and the posterior conversion rate, the conversion rate of the information to be analyzed after the target time is analyzed.

6. The method according to claim 5, characterized in that, The method further includes: Obtain a first weight corresponding to the prior conversion rate and a second weight corresponding to the posterior conversion rate. The first weight is used to control the degree of reference to the prior conversion rate when determining the conversion rate of the information to be analyzed after the target time. The second weight is used to control the degree of reference to the posterior conversion rate when determining the conversion rate of the information to be analyzed after the target time. The step of combining the prior conversion rate and the posterior conversion rate to analyze the conversion rate of the information to be analyzed after the target time includes: By combining the first weight, the prior conversion rate, the second weight, and the posterior conversion rate, the conversion rate of the information to be analyzed after the target time is analyzed.

7. The method according to claim 6, characterized in that, The first conversion data includes the number of first objects that clicked on the information to be analyzed before the target time. Obtaining the first weight corresponding to the prior conversion rate and the second weight corresponding to the posterior conversion rate includes: Based on the second conversion data and the number of the first objects, a first parameter and a second parameter corresponding to the beta distribution are determined. The ratio of the first parameter to the second parameter corresponds to the ratio between the conversion rate and the non-conversion rate represented by the second conversion data. The ratio of the first parameter to the second parameter and the first object number corresponds to the ratio of the first weight to the second weight. The first parameter and the second parameter are used to adjust the distribution mode represented by the beta distribution. The beta distribution is used to represent the distribution mode of the prior conversion rate. The ratio of the sum of the first parameter to the sum of the second parameter is determined as the first weight, and the ratio of the number of the first object to the sum of the second parameter is determined as the second weight. The sum of the first parameter is the sum of the first parameter and the second parameter, and the sum of the second parameter is the sum of the first parameter and the number of the first object. The step of predicting the prior conversion rate corresponding to the information to be analyzed based on the second conversion data includes: The expected value corresponding to the beta distribution is determined as the prior conversion rate.

8. The method according to claim 7, characterized in that, The sum of the first parameter is equal to the number of the first objects.

9. The method according to claim 1, characterized in that, The information to be analyzed is used to recommend the target object to the object to be recommended, and the method further includes: Obtain first feature information corresponding to the target object and second feature information corresponding to the object to be recommended, wherein the first feature information is used to characterize the object features corresponding to the target object and the second feature information is used to characterize the object features corresponding to the object to be recommended. Based on the first feature information and the second feature information, predict the first conversion rate corresponding to the information to be analyzed; The step of combining the first conversion data and the second conversion data to analyze the conversion rate of the information to be analyzed after the target time includes: By combining the first conversion data and the second conversion data, the second conversion rate corresponding to the information to be analyzed after the target time is analyzed; The second conversion rate is adjusted based on the difference between the first conversion rate and the second conversion rate to obtain the conversion rate of the information to be analyzed after the target time. The adjustment is used to reduce the difference between the first conversion rate and the second conversion rate.

10. The method according to claim 9, characterized in that, The method further includes: Obtain historical sample delivery information and the corresponding sample conversion rate. The historical sample delivery information is used to recommend sample objects to sample objects to be recommended. The sample conversion rate is the conversion rate obtained by delivering the historical sample delivery information in a historical period. Obtain first sample feature information and second sample feature information, wherein the first sample feature information is used to characterize the object features corresponding to the sample object, and the second sample feature information is used to characterize the object features corresponding to the sample object to be recommended. Based on the first sample feature information and the second sample feature information, the undetermined conversion rate corresponding to the historical delivery information of the sample is predicted by the initial conversion rate prediction model; Based on the difference between the undetermined conversion rate and the sample conversion rate, the model parameters corresponding to the initial conversion rate prediction model are adjusted to obtain the conversion rate prediction model. The step of predicting the first conversion rate corresponding to the information to be analyzed based on the first feature information and the second feature information includes: Based on the first feature information and the second feature information, the conversion rate prediction model is used to predict the first conversion rate corresponding to the information to be analyzed.

11. The method according to claim 9, characterized in that, The step of adjusting the second conversion rate based on the difference between the first conversion rate and the second conversion rate to obtain the conversion rate of the information to be analyzed after the target time includes: Based on the difference between the first conversion rate and the second conversion rate and the adjustment parameter, the second conversion rate is adjusted to obtain the conversion rate of the information to be analyzed after the target time. The adjustment parameter is used to control the adjustment intensity of the second conversion rate based on the difference between the first conversion rate and the second conversion rate.

12. The method according to claim 1, characterized in that, The step of obtaining the first conversion data corresponding to the information to be analyzed includes: Based on a preset time interval of n times between the target time and the initial click time corresponding to the information to be analyzed, first conversion data corresponding to the information to be analyzed is obtained. The first conversion data is used to characterize the conversion rate of the information to be analyzed between the initial click time and the target time. The initial click time is the time when the information to be analyzed is first clicked, and n is a positive integer.

13. The method according to claim 12, characterized in that, The method further includes: If the difference between the conversion rate of the information to be analyzed after the target time and the conversion rate of the information to be analyzed after a historical time is greater than a difference threshold, a prompt message is generated. The prompt message is used to indicate that the conversion rate of the information to be analyzed has changed abnormally. The historical time is the previous time that is separated from the target time by the preset time interval. The conversion rate of the information to be analyzed after the historical time is obtained based on the conversion rate of the information to be analyzed before the historical time.

14. The method according to claim 1, characterized in that, The method further includes: Based on the conversion rate of the information to be analyzed after the target time, the delivery method of the information to be analyzed after the target time is adjusted, and the delivery method is used to determine the prominence of displaying the information to be analyzed.

15. A conversion rate analysis device, characterized in that, The device includes a first acquisition unit, a second acquisition unit, a determination unit, a third acquisition unit, and an analysis unit: The first acquisition unit is used to acquire information to be analyzed, and the information to be analyzed is used to recommend target objects; The second acquisition unit is used to acquire the first conversion data corresponding to the information to be analyzed. The first conversion data is used to characterize the conversion rate corresponding to the information to be analyzed before the target time. The conversion rate is used to characterize the proportion of objects that perform conversion behavior for the objects recommended by the corresponding information among the objects that click on the corresponding information. The determining unit is used to determine the historical delivery information corresponding to the information to be analyzed. The historical delivery information is used to recommend associated objects. The associated objects are objects that meet the object association conditions with the target object. The third acquisition unit is used to acquire the second conversion data corresponding to the historical delivery information, and the second conversion data is used to characterize the conversion rate corresponding to the historical delivery information. The analysis unit is used to combine the first conversion data and the second conversion data to analyze the conversion rate of the information to be analyzed after the target time.

16. A computer device, characterized in that, The computer device includes a processor and memory: The memory is used to store computer programs and to transfer the computer programs to the processor; The processor is configured to execute the conversion rate analysis method according to any one of claims 1-14 according to instructions in the computer program.

17. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program for performing the conversion rate analysis method according to any one of claims 1-14.

18. A computer program product comprising a computer program, which, when run on a computer device, causes the computer device to perform the conversion rate analysis method according to any one of claims 1-14.

Images