Page processing method and device, training method and device, electronic equipment and storage medium

Abstract

The present disclosure provides a page processing method and device, a training method and device, an electronic device, and a storage medium, relates to the technical field of artificial intelligence, and in particular to big data and deep learning technology. The specific implementation scheme is as follows: performing feature extraction on object data related to a guide page to obtain object feature data; using the object feature data to obtain a plurality of page conversion rate prediction values respectively corresponding to a plurality of conversion paths, wherein the plurality of conversion paths correspond to the guide page; determining a target page conversion rate prediction value of the guide page according to the plurality of page conversion rate prediction values; and adjusting the content of the guide page according to the target page conversion rate prediction value.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of artificial intelligence, in particular to big data and deep learning technology. Specifically, it relates to a page processing method and device, an electronic device, and a storage medium. BACKGROUND

[0002] With the continuous development of Internet technology, the Internet has become the main means of information recommendation. Information recommendation can be made according to the content displayed on the pages of a website or an application. In the process of information recommendation, it is expected that users can complete conversion operations based on the content displayed on the pages. SUMMARY

[0003] The present disclosure provides a page processing method, a training method, a device, an electronic device, and a storage medium.

[0004] According to an aspect of the present disclosure, a page processing method is provided, including: performing feature extraction on object data related to a guide page to obtain object feature data; using the object feature data to obtain a plurality of page conversion rate prediction values corresponding to a plurality of conversion paths, respectively, wherein the plurality of conversion paths correspond to the guide page; determining a target page conversion rate prediction value of the guide page according to the plurality of page conversion rate prediction values; and adjusting the content of the guide page according to the target page conversion rate prediction value.

[0005] According to another aspect of the present disclosure, a training method of a page conversion rate prediction model is provided, including: performing feature extraction on sample object data related to a sample guide page to obtain sample object feature data; using the sample object feature data to obtain a plurality of sample page conversion rate prediction values corresponding to a plurality of sample conversion paths, respectively, wherein the plurality of sample conversion paths correspond to the sample guide page; determining a target sample page conversion rate prediction value of the sample guide page according to the plurality of sample page conversion rate prediction values; and training a predetermined model using the target sample page conversion rate prediction value and a target sample page conversion rate true value of the sample guide page to obtain the page conversion rate prediction model.

[0006] According to another aspect of the present disclosure, there is provided a page processing apparatus, comprising: a first obtaining module configured to perform feature extraction on object data associated with a guide page to obtain object feature data; a second obtaining module configured to obtain a plurality of page conversion rate prediction values corresponding to a plurality of conversion paths respectively using the object feature data, wherein the plurality of conversion paths correspond to the guide page; a first determining module configured to determine a target page conversion rate prediction value of the guide page according to the plurality of page conversion rate prediction values; and an adjusting module configured to adjust content of the guide page according to the target page conversion rate prediction value.

[0007] According to another aspect of the present disclosure, there is provided a page conversion rate prediction model training apparatus, comprising: a third obtaining module configured to perform feature extraction on sample object data associated with a sample guide page to obtain sample object feature data; a fourth obtaining module configured to obtain a plurality of sample page conversion rate prediction values corresponding to a plurality of sample conversion paths respectively using the sample object feature data, wherein the plurality of sample conversion paths correspond to the sample guide page; a second determining module configured to determine a target sample page conversion rate prediction value of the sample guide page according to the plurality of sample page conversion rate prediction values; and a fifth obtaining module configured to train a predetermined model using the target sample page conversion rate prediction value and a target sample page conversion rate real value of the sample guide page to obtain the page conversion rate prediction model.

[0008] According to another aspect of the present disclosure, there is provided an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method as described above.

[0009] According to another aspect of the present disclosure, there is provided a non-transitory computer readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to perform the method as described above.

[0010] According to another aspect of the present disclosure, there is provided a computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the method as described above.

[0011] It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0012] The accompanying drawings are used to better understand the present scheme, and do not constitute a limitation on the present disclosure. Among them:

[0013] Figure 1 An exemplary system architecture to which the page processing method, the training method and the device of the page conversion rate prediction model according to embodiments of the present disclosure can be applied is schematically shown;

[0014] Figure 2 A flowchart of the page processing method according to embodiments of the present disclosure is schematically shown;

[0015] Figure 3 An exemplary schematic diagram of a content processing process according to embodiments of the present disclosure is schematically shown;

[0016] Figure 4 A flowchart of the training method of the page conversion rate prediction model according to embodiments of the present disclosure is schematically shown;

[0017] Figure 5 An exemplary schematic diagram of a training process of the page conversion rate prediction model according to embodiments of the present disclosure is schematically shown;

[0018] Figure 6 A block diagram of the page processing device according to embodiments of the present disclosure is schematically shown;

[0019] Figure 7 A block diagram of the training device of the page conversion rate prediction model according to embodiments of the present disclosure is schematically shown; and

[0020] Figure 8 A block diagram of an electronic device suitable for implementing the page processing method and the training method of the page conversion rate prediction model according to embodiments of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0021] Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, which include various details of the embodiments of the present disclosure to help understanding, and should be considered as merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Also, in order to be clear and concise, the description below omits the description of well-known functions and structures.

[0022] Information is recommended according to the content displayed by the page of a website or an application. In the process of information being recommended, it is expected that the user can complete a conversion operation based on the content displayed by the page. The page can include a landing page. The landing page (i.e., a promotion page or a target page) can refer to a carrier page in which it is expected that the user completes a task.

[0023] The guide page can be set in the landing page, and the guide page can refer to a page for guiding a user to perform a related operation. The related operation can refer to an operation related to the content of the guide page. The page conversion rate of the guide page can be predicted, so as to adjust the content of the guide page according to the page conversion rate prediction value, and improve the page conversion rate.

[0024] There are multiple conversion paths from the user entering the landing page to the conversion occurring. Therefore, the present embodiment proposes a page processing scheme. The object data related to the guide page is subjected to feature extraction to obtain object feature data. The object feature data is used to obtain multiple page conversion rate prediction values respectively corresponding to multiple conversion paths. The multiple conversion paths correspond to the guide page. According to the multiple page conversion rate prediction values, a target page conversion rate prediction value of the guide page is determined. According to the target page conversion rate prediction value, the content of the guide page is adjusted.

[0025] The object feature data is used to obtain a page conversion rate prediction value corresponding to each conversion path in the multiple conversion paths, and the target page conversion rate prediction value is obtained based on the page conversion rate prediction values of different conversion paths, which realizes effective utilization of the contribution of different conversion paths to the page conversion rate prediction value to determine the target page conversion rate prediction value, and further improves the accuracy of the page conversion rate prediction value. On this basis, the content of the guide page is adjusted according to the target page conversion rate prediction value, and the page conversion rate is improved.

[0026] Figure 1 An exemplary system architecture to which the page processing method, the training method and the device of the page conversion rate prediction model according to the present embodiment can be applied is schematically shown.

[0027] It should be noted that Figure 1 The shown is only an example of a system architecture to which the present embodiment can be applied, to help those skilled in the art understand the technical content of the present disclosure, but does not mean that the present embodiment cannot be used in other devices, systems, environments or scenarios. For example, in another embodiment, the exemplary system architecture to which the page processing method, the training method and the device of the page conversion rate prediction model can be applied can include a terminal device, but the terminal device can not need to interact with the server to implement the page processing method, the training method and the device of the page conversion rate prediction model provided by the present embodiment.

[0028] As Figure 1 shown, the system architecture 100 according to the embodiment can include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used as a medium to provide a communication link between the terminal devices 101, 102, 103 and the server 105. The network 104 can include various connection types, such as wired and / or wireless communication links, etc.

[0029] The user can use the terminal devices 101, 102, and 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications can be installed on the terminal devices 101, 102, and 103, such as knowledge reading applications, web browser applications, search applications, instant messaging tools, email clients, and / or social platform software, etc. (only as examples).

[0030] The terminal devices 101, 102, and 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smartphones, tablet computers, laptop computers, desktop computers, etc.

[0031] The server 105 can be various types of servers providing various services, such as a background management server providing support for content browsed by the user using the terminal devices 101, 102, and 103 (only as an example). The background management server can analyze and process received user requests and other data, and feed back the processing results (such as web pages, information, or data, etc. obtained or generated according to user requests) to the terminal devices.

[0032] The server 105 can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, solving the defects of large management difficulty and weak business scalability in traditional physical hosts and VPS services (Virtual Private Server, VPS). The server 105 can also be a server of a distributed system, or a server combined with a blockchain.

[0033] It should be noted that the page processing method and the training method of the page conversion rate prediction model provided in the embodiments of the present disclosure can generally be executed by the server 105. Correspondingly, the page processing apparatus and the training apparatus of the page conversion rate prediction model provided in the embodiments of the present disclosure can generally be arranged in the server 105. The page processing method and the training method of the page conversion rate prediction model provided in the embodiments of the present disclosure can also be executed by a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, and 103 and / or the server 105. Correspondingly, the page processing apparatus and the training apparatus of the page conversion rate prediction model provided in the embodiments of the present disclosure can also be arranged in a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, and 103 and / or the server 105.

[0034] Alternatively, the page processing method and the training method of the page conversion rate prediction model provided in the embodiments of the present disclosure can also be generally executed by the terminal device 101, 102, or 103. Accordingly, the page processing apparatus and the training apparatus of the page conversion rate prediction model provided in the embodiments of the present disclosure can also be arranged in the terminal device 101, 102, or 103.

[0035] It should be understood that Figure 1 The number of terminal devices, networks, and servers in the above-mentioned system is only illustrative. Any number of terminal devices, networks, and servers can be provided according to implementation needs.

[0036] Figure 2 A flowchart of a page processing method according to an embodiment of the present disclosure is schematically shown.

[0037] As Figure 2 shown, the method 200 includes operations S210-S240.

[0038] In operation S210, feature extraction is performed on object data related to a guide page to obtain object feature data.

[0039] In operation S220, a plurality of page conversion rate prediction values respectively corresponding to a plurality of conversion paths are obtained by using the object feature data. The plurality of conversion paths correspond to the guide page.

[0040] In operation S230, a target page conversion rate prediction value of the guide page is determined according to the plurality of page conversion rate prediction values.

[0041] In operation S240, the content of the guide page is adjusted according to the target page conversion rate prediction value.

[0042] According to the embodiments of the present disclosure, the guide page can include at least one of a floating layer and a pop-up window. The object data can include at least one of the following: data related to a promotion, data related to a terminal device, data related to a user, and data related to a page. The data related to the terminal device can include at least one of the following: a model type of the terminal device and a network type of the terminal device. The data related to the user can include at least one of the following: user basic attribute data, user preference data, and user purchase behavior data. The data related to the page can include at least one of the following: a historical page conversion rate real value and image data of an image included in the guide page. The object feature data can be data obtained by performing feature extraction on medium features.

[0043] According to an embodiment of the present disclosure, a conversion path can refer to a path experienced from a user entering a page to a page conversion occurring. The path from the user entering the page to the page conversion occurring can include multiple. For example, the conversion path can be a path of the user entering the page, the guide page being displayed, the user responding to the guide page, and the user performing a conversion operation, i.e., the conversion path can be user enters a page → guide page is displayed → user responds to guide page → user performs a conversion operation. The "responding" in the user responding to the guide page can include a click or a swipe. The click can include a single click or a double click.

[0044] According to an embodiment of the present disclosure, for a certain guide page, the page conversion rate can refer to a ratio based on a number of page conversion operations occurring based on the guide page and a total number of page operations. The total number of page operations includes a number of page conversion operations occurring based on the guide page and a number of page conversion operations not occurring based on the guide page. The page conversion operation occurring based on the guide page can refer to a guide into a related page in response to the guide page, and the user performing a conversion operation in the related page. The conversion operation can refer to an operation corresponding to the purpose of the guide page. For example, if the guide page is used to guide the user to purchase an item, the conversion operation can refer to an item purchase operation.

[0045] According to an embodiment of the present disclosure, the page conversion rate prediction value can refer to a value obtained by predicting the page conversion rate. The target page conversion rate prediction value can be obtained according to the page conversion rate prediction value corresponding to each of the plurality of conversion paths.

[0046] According to an embodiment of the present disclosure, object data related to the guide page can be obtained. The object data can be feature extracted to obtain object feature data. For example, the object data can be feature extracted by a model obtained by training a preset model using sample object data to obtain object feature data. The preset model can include a deep learning model.

[0047] According to an embodiment of the present disclosure, after obtaining the object feature data, the object feature data can be processed to obtain the page conversion rate prediction value corresponding to each of the plurality of conversion paths. For example, the object feature data can be processed by a model obtained by training a preset model using sample object data to obtain the page conversion rate prediction value corresponding to each of the plurality of conversion paths. The model obtained by training the preset model using the sample object data can include a branch network for processing a page conversion rate prediction operation corresponding to each of the plurality of conversion paths.

[0048] According to an embodiment of the present disclosure, after obtaining the plurality of page conversion rate prediction values, a target page conversion rate prediction value of the guide page can be obtained according to the page conversion rate prediction value corresponding to each of the plurality of conversion paths. For example, the plurality of page conversion rate prediction values can be weighted and summed, and the result is determined as the target page conversion rate prediction value.

[0049] According to an embodiment of the present disclosure, the content of the guide page can be adjusted according to the target page conversion rate prediction value of the guide page, so as to improve the target page conversion rate prediction value of the guide page. For example, in a case where it is determined that the target page conversion rate prediction value of the guide page is less than or equal to the page conversion rate prediction threshold, the content of the guide page can be adjusted. In a case where it is determined that the target page conversion rate prediction value of the guide page is greater than the page conversion rate prediction threshold, the content of the guide page can be kept unchanged.

[0050] According to an embodiment of the present disclosure, if there are a plurality of guide pages, operations S210-S240 can be performed on each of the plurality of guide pages. In addition, a target guide page can be determined from the plurality of guide pages according to the target page conversion rate prediction value corresponding to each of the plurality of guide pages. The target guide page can be the guide page with the maximum target page conversion rate prediction value. Adjusting the content of the guide page according to the target page conversion rate prediction value of the guide page can include: in a case where it is determined that the target page conversion rate prediction value of the guide page is less than or equal to the page conversion rate prediction threshold, the content of the guide page can be adjusted according to the content of the target guide page. For example, the content of the guide page is set to the content of the target guide page.

[0051] According to an embodiment of the present disclosure, the object feature data is used to obtain the page conversion rate prediction value corresponding to each of the plurality of conversion paths, and the target page conversion rate prediction value is obtained based on the page conversion rate prediction values of different conversion paths, which realizes effective use of the contribution of different conversion paths to the page conversion rate prediction value to determine the target page conversion rate prediction value, and further improves the accuracy of the page conversion rate prediction value. On this basis, the content of the guide page is adjusted according to the target page conversion rate prediction value, which improves the page conversion rate.

[0052] According to an embodiment of the present disclosure, the plurality of conversion paths include a page response path and a page non-response path.

[0053] According to an embodiment of the present disclosure, operation S220 can include the following operations.

[0054] The object feature data is subjected to page response conversion rate prediction to obtain a page response conversion rate prediction value corresponding to the page response path. The object feature data is subjected to page non-response conversion rate prediction to obtain a page non-response conversion rate prediction value corresponding to the page non-response path.

[0055] According to an embodiment of the present disclosure, the page response path can refer to a path in which the object responds to the guide page and conversion occurs. The page non-response path can refer to a path in which the object does not respond to the guide page and conversion occurs. The page response conversion rate prediction value can refer to a conversion rate prediction value in which page conversion occurs in the case of the page response path. The page non-response conversion rate prediction value can refer to a conversion rate prediction value in which page conversion occurs in the case of the page non-response path.

[0056] According to an embodiment of the present disclosure, a model obtained by training a preset model using sample object data can be used to predict the page conversion rate of the object feature data, to obtain a page response conversion rate prediction value corresponding to the page response path. A model obtained by training a preset model using sample object data can be used to predict the page non-response conversion rate of the object feature data, to obtain a page non-response conversion rate prediction value corresponding to the page non-response path.

[0057] According to an embodiment of the present disclosure, operation S230 can include the following operations.

[0058] Based on the conditional probability method, the target page conversion rate prediction value of the guide page is determined according to the page response conversion rate prediction value and the page non-response conversion rate prediction value.

[0059] According to an embodiment of the present disclosure, the conditional probability method can refer to a method for determining a page conversion rate prediction value of a guide page in the case of page conversion. The case of page conversion can include a page response rate in the case of the page response path and a page non-response conversion rate in the case of the page non-response path.

[0060] According to an embodiment of the present disclosure, a probability value corresponding to the page response path and a probability value corresponding to the page non-response path can be determined. The target conversion rate prediction value is determined according to the probability value corresponding to the page response path and the page response rate prediction value, and the probability value corresponding to the page non-response path and the page non-response rate prediction value. The probability value corresponding to the page response path can refer to a probability value of page response. The probability value corresponding to the page non-response path can refer to a probability value of non-page response.

[0061] According to an embodiment of the present disclosure, based on the conditional probability method, the target page conversion rate prediction value of the guide page is determined according to the page response conversion rate prediction value and the page non-response conversion rate prediction value, which can include the following operations.

[0062] The page response rate prediction is performed on the object feature data to obtain a page response rate prediction value. The page response conversion rate prediction probability value is obtained according to the page response rate prediction value and the page response conversion rate prediction value. The page non-response conversion rate prediction probability value is obtained according to the page response rate prediction value and the page non-response conversion rate prediction value. The target page conversion rate prediction value of the guide page is determined according to the page response conversion rate prediction probability value and the page non-response conversion rate prediction probability value.

[0063] According to an embodiment of the present disclosure, the page response rate prediction value can refer to a probability value of occurrence of page response. The page response rate prediction value can be obtained by using a preset model trained by sample object data to perform page response rate prediction on object feature data.

[0064] According to an embodiment of the present disclosure, after the page response rate prediction value is determined, the first contribution proportion value and the second contribution proportion value can be determined according to the page response rate prediction value. The first contribution proportion value can refer to a contribution proportion value of the page response conversion rate prediction value in the target page conversion rate prediction value. The second contribution proportion value can refer to a contribution proportion value of the page non-response conversion rate prediction value in the target page conversion rate prediction value.

[0065] According to an embodiment of the present disclosure, the page response conversion rate prediction probability value can be obtained according to the first contribution proportion value and the page response conversion rate prediction value. The page non-response conversion rate prediction probability value can be obtained according to the second contribution proportion value and the page non-response conversion rate prediction value. The target page conversion rate prediction value can be determined according to the page response conversion rate prediction probability value and the page non-response conversion rate prediction probability value. For example, the sum of the page response conversion rate prediction probability value and the page non-response conversion rate prediction probability value can be determined as the target page conversion rate prediction value.

[0066] According to an embodiment of the present disclosure, obtaining the page response conversion rate prediction probability value according to the page response rate prediction value and the page response conversion rate prediction value can include the following operations.

[0067] A first product between the page response rate prediction value and the page response conversion rate prediction value is determined. The first product is determined as the page response conversion rate prediction probability value.

[0068] According to an embodiment of the present disclosure, obtaining the page non-response conversion rate prediction probability value according to the page response rate prediction value and the page non-response conversion rate prediction value can include the following operations.

[0069] A second product between the page response rate prediction value and the page non-response conversion rate prediction value is determined. A difference value between the page non-response conversion rate prediction value and the second product is determined as the page non-response conversion rate prediction probability value.

[0070] According to an embodiment of the present disclosure, the page response conversion rate prediction probability value can be determined according to the following formula (1).

[0071] CTCVR = CTR cta *CTR cta-clk (1)

[0072] According to an embodiment of the present disclosure, the CTCVR represents a page response conversion rate prediction probability value. The CTR cta represents a page response rate prediction value. The CTR cta-clk represents a page response conversion rate prediction value.

[0073] According to an embodiment of the present disclosure, the page response conversion rate prediction probability value can be determined according to the following formula (2).

[0074] NCTCVR = CTR cta-noclk -CTR cta *CTR cta-noclk (2)

[0075] According to an embodiment of the present disclosure, the NCTCVR represents a page non-response conversion rate prediction probability value. The CTR cta-noclk represents a page non-response conversion rate prediction value.

[0076] According to an embodiment of the present disclosure, the target page conversion rate prediction value can be determined according to the following formula (3).

[0077] CVR = CTCVR + NCTCVR (3)

[0078] According to an embodiment of the present disclosure, the CVR represents a target page conversion rate prediction value.

[0079] According to an embodiment of the present disclosure, the object feature data is obtained by processing and guiding page-related object data by using a backbone network included in the page conversion rate prediction model. The page response rate prediction value is obtained by processing the object feature data by using a page response rate branch network included in the page conversion rate prediction model. The page response conversion rate prediction value is obtained by processing the object feature data by using a page response conversion rate branch network included in the page conversion rate prediction model. The page non-response conversion rate prediction value is obtained by processing the object feature data by using a page non-response conversion rate branch network included in the page conversion rate prediction model.

[0080] According to an embodiment of the present disclosure, the page conversion rate prediction model can include a trunk network and three parallel branch networks, i.e., a page response rate branch network, a page response conversion rate branch network, and a page non-response conversion rate branch network. The page response rate branch network can be used to process operations of predicting a page response rate. The page response conversion rate branch network can be used to process page conversion rate prediction operations corresponding to a page response path. The page non-response conversion rate branch network can be used to process page conversion rate prediction operations corresponding to a page non-response path.

[0081] According to an embodiment of the present disclosure, the network structures of the trunk network, the page response rate branch network, the page response conversion rate branch network, and the page non-response conversion rate branch network can be configured according to actual business needs, which are not limited herein.

[0082] According to an embodiment of the present disclosure, the page conversion rate prediction model can be obtained by training a preset model using sample object data. For example, feature extraction is performed on the sample object data to obtain sample object feature data. Using the sample object feature data, a plurality of sample page conversion rate prediction values corresponding to a plurality of sample conversion paths are obtained. The plurality of sample conversion paths correspond to a sample guide page. According to the plurality of sample page conversion rate prediction values, a sample page conversion rate prediction value of the sample guide page is determined. Using the sample page conversion rate prediction value and a sample page conversion rate real value of the sample guide page, the predetermined model is trained to obtain the page conversion rate prediction model.

[0083] The page processing method according to the embodiments of the present disclosure will be further described below with reference to Figure 3 , in combination with specific embodiments.

[0084] Figure 3 An example schematic diagram of a content processing process according to an embodiment of the present disclosure is schematically shown.

[0085] As shown in Figure 3 , in the content processing process 300, object data 301 is input into a trunk network 302 to obtain object feature data 303.

[0086] The object feature data 303 is input into a page response rate branch network 304 to obtain a page response rate prediction value 305. The object feature data 303 is input into a page response conversion rate branch network 306 to obtain a page response conversion rate prediction value 307. The object feature data 303 is input into a page non-response conversion rate branch network 308 to obtain a page non-response conversion rate prediction value 309.

[0087] According to the page response rate prediction value 305 and the page response conversion rate prediction value 307, a page response rate prediction probability value 310 is obtained. According to the page response rate prediction value 305 and the page non-response conversion rate prediction value 309, a page non-response conversion rate prediction probability value 311 is obtained. According to the page response rate prediction probability value 310 and the page non-response conversion rate prediction probability value 311, a target page conversion rate prediction value 312 of the guide page is obtained.

[0088] The page final conversion rate prediction model can be directly created without splitting the conversion path. However, there are multiple conversion paths from the user entering the landing page to the conversion occurring. In actual application, the contributions of the user response page and the user non-response page to the page conversion rate are different.

[0089] Therefore, the difference in the contribution of different conversion paths to the page conversion rate can be obtained by using the modeling method of splitting the conversion path, so as to effectively utilize the contribution of different conversion paths to the page conversion rate prediction value to determine the target page conversion rate prediction value and improve the accuracy of the page conversion rate prediction value. Thus, the present embodiment of the present disclosure proposes a training scheme of a page conversion rate prediction model based on a conversion path.

[0090] The training scheme of the page conversion rate prediction model will be described below. Figures 4-5 The training scheme of the page conversion rate prediction model will be described below.

[0091] Figure 4 A flowchart of a training method of a page conversion rate prediction model according to an embodiment of the present disclosure is schematically shown.

[0092] As shown in the flowchart, the method 400 can include operations S410-S440. Figure 4

[0093] Operation S410, feature extraction is performed on sample object data related to a sample guide page to obtain sample object feature data.

[0094] Operation S420, using the sample object feature data, a plurality of sample page conversion rate prediction values respectively corresponding to a plurality of sample conversion paths are obtained. The plurality of sample conversion paths correspond to the sample guide page.

[0095] Operation S430, according to the plurality of sample page conversion rate prediction values, a target sample page conversion rate prediction value of the sample guide page is determined.

[0096] Operation S440, using the target sample page conversion rate prediction value and the target sample page conversion rate real value of the sample guide page, a predetermined model is trained to obtain a page conversion rate prediction model.

[0097] ​According to an embodiment of the present disclosure, the sample conversion path can be the same conversion path as the conversion path, which is not described herein again.

[0098] According to an embodiment of the present disclosure, the output value can be obtained based on the loss function, by using the target sample page conversion rate prediction value of the sample guide page and the target sample page conversion rate real value. The model parameter of the predetermined model is adjusted according to the output value until a predetermined condition is met. The predetermined model trained under the condition that the predetermined condition is met is determined as the page conversion rate prediction model.

[0099] According to an embodiment of the present disclosure,

[0100] According to an embodiment of the present disclosure, the plurality of sample conversion paths include a sample page response path and a sample page non-response path corresponding to the sample guide page.

[0101] According to an embodiment of the present disclosure, operation S420 can include the following operations.

[0102] The sample object feature data is subjected to page response conversion rate prediction to obtain a sample page response conversion rate prediction value corresponding to the sample page response path. The sample object feature data is subjected to page non-response conversion rate prediction to obtain a sample page non-response conversion rate prediction value corresponding to the sample page non-response path.

[0103] According to an embodiment of the present disclosure, operation S430 can include the following operations.

[0104] Based on the conditional probability method, the target sample page conversion rate prediction value of the sample guide page is determined according to the sample page response conversion rate prediction value and the sample page non-response conversion rate prediction value.

[0105] According to an embodiment of the present disclosure, based on the conditional probability method, the target sample page conversion rate prediction value of the sample guide page is determined according to the sample page response conversion rate prediction value and the sample page non-response conversion rate prediction value, which can include the following operations.

[0106] The sample object feature data is subjected to page response rate prediction to obtain a sample page response rate prediction value. The sample page response conversion rate prediction probability value is obtained according to the sample page response rate prediction value and the sample page response conversion rate prediction value. The sample page non-response conversion rate prediction probability value is obtained according to the sample page response rate prediction value and the page non-response conversion rate prediction value. The target sample page conversion rate prediction value of the sample guide page is determined according to the sample page response conversion rate prediction probability value and the sample page non-response conversion rate prediction probability value.

[0107] According to an embodiment of the present disclosure, the page processing method is different from the above-described page processing method in that the page processing method is for object features, the training method of the page conversion rate prediction model is for sample object features, and the processing operation is the same as the page processing method, which will not be described herein again.

[0108] According to an embodiment of the present disclosure, the predetermined model comprises a backbone network.

[0109] According to an embodiment of the present disclosure, the operation S410 can comprise the following operation.

[0110] The sample object data related to the sample guide page is processed by using the backbone network to obtain sample object feature data.

[0111] According to an embodiment of the present disclosure, the predetermined model comprises a page response rate branch network, a page response conversion rate branch network, and a page non-response conversion rate branch network.

[0112] According to an embodiment of the present disclosure, the page response rate prediction on the sample object feature data to obtain a sample page response rate prediction value can comprise the following operation.

[0113] The sample object feature data is processed by using the page response rate branch network to obtain the sample page response rate prediction value.

[0114] According to an embodiment of the present disclosure, the page response conversion rate prediction on the sample object feature data to obtain a sample page response conversion rate prediction value corresponding to a sample page response path can comprise the following operation.

[0115] The sample object feature data is processed by using the page response conversion rate branch network to obtain the sample page response conversion rate prediction value.

[0116] According to an embodiment of the present disclosure, the page non-response conversion rate prediction on the sample object feature data to obtain a sample page non-response conversion rate prediction value corresponding to a sample page non-response path can comprise the following operation.

[0117] The sample object feature data is processed by using the page non-response conversion rate branch network to obtain the sample page non-response conversion rate prediction value.

[0118] According to an embodiment of the present disclosure, the page response rate branch network can be used to predict the page response rate. The page response conversion rate branch network can be used to predict the page response conversion rate. The page non-response conversion rate branch network can be used to predict the page non-response conversion rate. The sample object feature data can be input into the page response rate branch network to obtain a sample page response rate prediction value. The sample object feature data can be input into the page response conversion rate branch network to obtain a sample page response conversion rate prediction value. The sample object feature data can be input into the page non-response conversion rate branch network to obtain a sample page non-response conversion rate prediction value.

[0119] According to an embodiment of the present disclosure, operation S440 can include the following operations.

[0120] Based on the loss function, the output value is obtained by using the sample page response rate true value and the sample page response rate prediction value, the sample page response conversion rate true value and the sample page response conversion rate prediction value, the sample page non-response conversion rate true value and the sample page non-response conversion rate prediction value, and the target sample page conversion rate prediction value and the target sample page conversion rate true value of the sample guide page. The model parameters of the predetermined model are adjusted according to the output value until the predetermined condition is met. The predetermined model obtained when the predetermined condition is met is determined as the page conversion rate prediction model.

[0121] According to an embodiment of the present disclosure, the predetermined condition can include at least one of the convergence of the output value and the training round reaching the maximum round. The output value can include a first output value, a second output value, a third output value, and a fourth output value. The first output value is obtained by inputting the sample page response rate true value and the sample page response rate prediction value into the loss function. The second output value is obtained by inputting the sample page response conversion rate true value and the sample page response conversion rate prediction value into the loss function. The third output value is obtained by inputting the sample page non-response conversion rate true value and the sample page non-response conversion rate prediction value into the loss function. The fourth output value is obtained by inputting the target sample page conversion rate prediction value and the target sample page conversion rate true value. The model parameters of the predetermined model are further adjusted according to the first output value, the second output value, the third output value, and the fourth output value until the predetermined condition is met.

[0122] The training method of the page conversion rate prediction model according to the embodiments of the present disclosure will be further described below with reference to specific embodiments in combination with Figure 5 .

[0123] Figure 5 An example schematic diagram of the training process of the page conversion rate prediction model according to an embodiment of the present disclosure is schematically shown.

[0124] As Figure 5As shown, in the training process 500, the sample object data 501 is input into the backbone network 502 to obtain sample object feature data 503.

[0125] The sample object feature data 503 is input into the page response rate branch network 504 to obtain a sample page response rate prediction value 505. The sample object feature data 503 is input into the page response conversion rate branch network 506 to obtain a sample page response conversion rate prediction value 507. The sample object feature data 503 is input into the page non-response conversion rate branch network 508 to obtain a sample page non-response conversion rate prediction value 509.

[0126] According to the sample page response rate prediction value 505 and the sample page response conversion rate prediction value 507, a sample page response rate prediction probability value 510 is obtained. According to the sample page response rate prediction value 505 and the sample page non-response conversion rate prediction value 509, a sample page non-response conversion rate prediction probability value 511 is obtained. According to the sample page response rate prediction probability value 510 and the sample page non-response conversion rate prediction probability value 511, a target sample page conversion rate prediction value 512 of the sample guide page is obtained.

[0127] The sample page response rate true value 513 and the sample page response rate prediction value 505, the sample page response conversion rate true value 515 and the sample page response conversion rate prediction value 507, the sample page non-response conversion rate true value 516 and the sample page non-response conversion rate prediction value 509, and the target sample page conversion rate prediction value 512 of the sample guide page and the target sample page conversion rate true value 514 are input into the loss function 517 to obtain an output value 518. The model parameters of the predetermined model are adjusted according to the output value 518 until a predetermined condition is met. The predetermined model obtained when the predetermined condition is met is determined as the page conversion rate prediction model.

[0128] The above is only an exemplary embodiment, but is not limited thereto, and other page processing methods and page conversion rate prediction model training methods known in the art can also be included as long as the accuracy of the page conversion rate prediction value can be improved.

[0129] It should be noted that in the technical solutions of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved in the technical solutions comply with relevant laws and regulations and do not violate public order and good customs.

[0130] Figure 6 A block diagram of a page processing apparatus according to an embodiment of the present disclosure is schematically shown.

[0131] As Figure 6As shown, the page processing apparatus 600 can include a first obtaining module 610, a second obtaining module 620, a first determining module 630, and an adjusting module 640.

[0132] The first obtaining module 610 is configured to perform feature extraction on object data related to the guide page to obtain object feature data.

[0133] The second obtaining module 620 is configured to use the object feature data to obtain a plurality of page conversion rate prediction values respectively corresponding to a plurality of conversion paths.

[0134] The first determining module 630 is configured to determine a target page conversion rate prediction value of the guide page according to the plurality of page conversion rate prediction values.

[0135] The adjusting module 640 is configured to adjust the content of the guide page according to the target page conversion rate prediction value.

[0136] According to an embodiment of the present disclosure, the plurality of conversion paths include a page response path and a page non-response path.

[0137] According to an embodiment of the present disclosure, the second obtaining module 620 can include a first obtaining submodule and a second obtaining submodule.

[0138] The first obtaining submodule is configured to perform page response conversion rate prediction on the object feature data to obtain a page response conversion rate prediction value corresponding to the page response path.

[0139] The second obtaining submodule is configured to perform page non-response conversion rate prediction on the object feature data to obtain a page non-response conversion rate prediction value corresponding to the page non-response path.

[0140] According to an embodiment of the present disclosure, the first determining module 630 can include a first determining submodule.

[0141] The first determining submodule is configured to determine the target page conversion rate prediction value of the guide page according to the page response conversion rate prediction value and the page non-response conversion rate prediction value based on a conditional probability method.

[0142] According to an embodiment of the present disclosure, the first determining submodule can include a first obtaining unit, a second obtaining unit, a third obtaining unit, and a first determining unit.

[0143] The first obtaining unit is configured to perform page response rate prediction on the object feature data to obtain a page response rate prediction value.

[0144] The second obtaining unit is configured to obtain a page response conversion rate prediction probability value according to the page response rate prediction value and the page response conversion rate prediction value.

[0145] The third obtaining unit is configured to obtain a page non-response conversion rate prediction probability value according to the page response rate prediction value and the page non-response conversion rate prediction value.

[0146] The first determining unit is configured to determine a target page conversion rate prediction value of the guide page according to the page response conversion rate prediction probability value and the page non-response conversion rate prediction probability value.

[0147] According to an embodiment of the present disclosure, the second obtaining unit can include a first determining sub-unit and a second determining sub-unit.

[0148] The first determining sub-unit is configured to determine a first product between the page response rate prediction value and the page response conversion rate prediction value.

[0149] The second determining sub-unit is configured to determine the first product as the page response conversion rate prediction probability value.

[0150] According to an embodiment of the present disclosure, the third obtaining unit can include a third determining sub-unit and a fourth determining sub-unit.

[0151] The third determining sub-unit is configured to determine a second product between the page response rate prediction value and the page non-response conversion rate prediction value.

[0152] The fourth determining sub-unit is configured to determine a difference between the page non-response conversion rate prediction value and the second product as the page non-response conversion rate prediction probability value.

[0153] According to an embodiment of the present disclosure, the object feature data is obtained by processing object data related to the guide page by using a backbone network included in the page conversion rate prediction model. The page response rate prediction value is obtained by processing the object feature data by using a page response rate branch network included in the page conversion rate prediction model. The page response conversion rate prediction value is obtained by processing the object feature data by using a page response conversion rate branch network included in the page conversion rate prediction model. The page non-response conversion rate prediction value is obtained by processing the object feature data by using a page non-response conversion rate branch network included in the page conversion rate prediction model.

[0154] Figure 7 A block diagram of a training apparatus of a page conversion rate prediction model according to an embodiment of the present disclosure is schematically shown.

[0155] As shown in Figure 7 The page conversion rate prediction model 700 can include a third obtaining module 710, a fourth obtaining module 720, a second determining module 730 and a fifth obtaining module 740.

[0156] The third obtaining module 710 is configured to perform feature extraction on sample object data related to a sample guide page to obtain sample object feature data.

[0157] The fourth obtaining module 720 is configured to obtain a plurality of sample page conversion rate prediction values corresponding to a plurality of sample conversion paths respectively by using the sample object feature data. The plurality of sample conversion paths correspond to the sample guide page.

[0158] The second determining module 730 is configured to determine a target sample page conversion rate prediction value of the sample guide page according to the plurality of sample page conversion rate prediction values.

[0159] The fifth obtaining module 740 is configured to train a predetermined model by using the target sample page conversion rate prediction value of the sample guide page and a target sample page conversion rate real value, to obtain a page conversion rate prediction model.

[0160] According to an embodiment of the present disclosure, the plurality of sample conversion paths include a sample page response path and a sample page non-response path corresponding to the sample guide page.

[0161] According to an embodiment of the present disclosure, the fourth obtaining module 720 can include a third obtaining submodule and a fourth obtaining submodule.

[0162] The third obtaining submodule is configured to perform page response conversion rate prediction on the sample object feature data to obtain a sample page response conversion rate prediction value corresponding to the sample page response path.

[0163] The fourth obtaining submodule is configured to perform page non-response conversion rate prediction on the sample object feature data to obtain a sample page non-response conversion rate prediction value corresponding to the sample page non-response path.

[0164] According to an embodiment of the present disclosure, the second determining module 730 can include a second determining submodule.

[0165] The second determining submodule is configured to determine the target sample page conversion rate prediction value of the sample guide page according to the sample page response conversion rate prediction value and the sample page non-response conversion rate prediction value based on a conditional probability method.

[0166] According to an embodiment of the present disclosure, the second determining submodule can include a fourth obtaining unit, a fifth obtaining unit, a sixth obtaining unit and a second determining unit.

[0167] The fourth obtaining unit is configured to perform page response rate prediction on the sample object feature data to obtain a sample page response rate prediction value.

[0168] The fifth obtaining unit is configured to obtain a sample page response conversion rate prediction probability value according to the sample page response rate prediction value and the sample page response conversion rate prediction value.

[0169] The sixth obtaining unit is configured to obtain a sample page non-response conversion rate prediction probability value according to the sample page response rate prediction value and the page non-response conversion rate prediction value.

[0170] The second determining unit is configured to determine a target sample page conversion rate prediction value of the sample guide page according to the sample page response conversion rate prediction probability value and the sample page non-response conversion rate prediction probability value.

[0171] According to an embodiment of the present disclosure, the predetermined model comprises a backbone network.

[0172] According to an embodiment of the present disclosure, the third obtaining module 710 can comprise a fifth obtaining submodule.

[0173] The fifth obtaining submodule is configured to process sample object data related to the sample guide page by using the backbone network to obtain sample object feature data.

[0174] According to an embodiment of the present disclosure, the predetermined model comprises a page response rate branch network, a page response conversion rate branch network and a page non-response conversion rate branch network.

[0175] According to an embodiment of the present disclosure, the fourth obtaining unit can comprise a first obtaining submodule.

[0176] The first obtaining submodule is configured to process the sample object feature data by using the page response rate branch network to obtain the sample page response rate prediction value.

[0177] According to an embodiment of the present disclosure, the fifth obtaining unit can comprise a second obtaining submodule.

[0178] The second obtaining submodule is configured to process the sample object feature data by using the page response conversion rate branch network to obtain the sample page response conversion rate prediction value.

[0179] According to an embodiment of the present disclosure, the sixth obtaining unit can comprise a third obtaining submodule.

[0180] The third obtaining submodule is configured to process the sample object feature data by using the page non-response conversion rate branch network to obtain the sample page non-response conversion rate prediction value.

[0181] According to an embodiment of the present disclosure, the fifth obtaining module 740 can comprise a sixth obtaining submodule, an adjusting submodule and a third determining submodule.

[0182] The sixth obtaining sub-module is configured to obtain an output value based on a loss function and the sample page response rate real value and the sample page response rate prediction value, the sample page response conversion rate real value and the sample page response conversion rate prediction value, the sample page non-response conversion rate real value and the sample page non-response conversion rate prediction value, and the target sample page conversion rate prediction value and the target sample page conversion rate real value of the sample guide page.

[0183] The adjusting sub-module is configured to adjust the model parameter of the predetermined model according to the output value until a predetermined condition is met.

[0184] The third determining sub-module is configured to determine the predetermined model obtained when the predetermined condition is met as the page conversion rate prediction model.

[0185] According to embodiments of the present disclosure, the present disclosure further provides an electronic device, a readable storage medium, and a computer program product.

[0186] According to embodiments of the present disclosure, an electronic device comprises at least one processor, and a memory connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method as described above.

[0187] According to embodiments of the present disclosure, a non-transitory computer readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to perform the method as described above.

[0188] According to embodiments of the present disclosure, a computer program product comprises a computer program, and the computer program, when executed by a processor, implements the method as described above.

[0189] Figure 8 A block diagram schematically shows an electronic device suitable for implementing the page processing method and the training method of the page conversion rate prediction model according to embodiments of the present disclosure. The electronic device is intended to represent a variety of forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent a variety of forms of mobile devices, such as personal digital assistants, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit the implementations of the present disclosure described and / or claimed in this document.

[0190] As Figure 8As shown, the electronic device 800 includes a computing unit 801 that can perform various appropriate actions and processes in accordance with a computer program stored in a read-only memory (ROM) 802 or a computer program loaded into a random access memory (RAM) 803 from a storage unit 808. In the RAM 803, various programs and data required for the operation of the electronic device 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other through a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0191] A plurality of components in the electronic device 800 are connected to the I / O interface 805, including an input unit 806 such as a keyboard, a mouse, and the like, an output unit 807 such as various types of displays, a speaker, and the like, a storage unit 808 such as a magnetic disk, an optical disk, and the like, and a communication unit 809 such as a network card, a modem, a wireless communication transceiver, and the like. The communication unit 809 allows the electronic device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0192] The computing unit 801 can be various general-purpose and / or special-purpose processing components having processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, and the like. The computing unit 801 performs various methods and processes described above, such as the page processing method or the training method of the page conversion rate prediction model. For example, in some embodiments, the page processing method or the training method of the page conversion rate prediction model can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the page processing method or the training method of the page conversion rate prediction model described above can be performed. Alternatively, in other embodiments, the computing unit 801 can be configured to perform the page processing method or the training method of the page conversion rate prediction model by any other appropriate means, such as by means of firmware.

[0193] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a complex programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0194] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. The program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, produces the functions / operations specified in the flowcharts and / or the block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0195] In the context of the present disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include but is not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0196] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0197] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0198] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, distributed system servers, or servers incorporating blockchain technology.

[0199] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0200] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A page processing method, comprising: performing feature extraction on object data associated with a guide page to obtain object feature data; obtaining, using the object feature data, a plurality of page conversion rate prediction values respectively corresponding to a plurality of conversion paths, wherein the plurality of conversion paths correspond to the guide page; determining a target page conversion rate prediction value of the guide page according to the plurality of page conversion rate prediction values; and adjusting content of the guide page according to the target page conversion rate prediction value; wherein the plurality of conversion paths include a page response path and a page non-response path, the page response path includes a path in which an object responds to the guide page and conversion occurs, and the page non-response path includes a path in which an object does not respond to the guide page and conversion occurs; wherein the obtaining, using the object feature data, a plurality of page conversion rate prediction values respectively corresponding to a plurality of conversion paths comprises: performing page response conversion rate prediction on the object feature data to obtain a page response conversion rate prediction value corresponding to the page response path; and performing page non-response conversion rate prediction on the object feature data to obtain a page non-response conversion rate prediction value corresponding to the page non-response path; wherein the determining a target page conversion rate prediction value of the guide page according to the plurality of page conversion rate prediction values comprises: determining, based on a conditional probability method, the target page conversion rate prediction value of the guide page according to the page response conversion rate prediction value and the page non-response conversion rate prediction value.

2. The method of claim 1, wherein, The determining, based on a conditional probability method, the target page conversion rate prediction value of the guide page according to the page response conversion rate prediction value and the page non-response conversion rate prediction value comprises: performing page response rate prediction on the object feature data to obtain a page response rate prediction value; obtaining a page response conversion rate prediction probability value according to the page response rate prediction value and the page response conversion rate prediction value; obtaining a page non-response conversion rate prediction probability value according to the page response rate prediction value and the page non-response conversion rate prediction value; and determining the target page conversion rate prediction value of the guide page according to the page response conversion rate prediction probability value and the page non-response conversion rate prediction probability value.

3. The method of claim 2, wherein, The obtaining a page response conversion rate prediction probability value according to the page response rate prediction value and the page response conversion rate prediction value comprises: determining a first product between the page response rate prediction value and the page response conversion rate prediction value; and determining the first product as the page response conversion rate prediction probability value.

4. The method of claim 2 or 3, wherein, The obtaining a page non-response conversion rate prediction probability value according to the page response rate prediction value and the page non-response conversion rate prediction value comprises: determining a second product between the page response rate prediction value and the page non-response conversion rate prediction value; and determining a difference between the page non-response conversion rate prediction value and the second product as the page non-response conversion rate prediction probability value.

5. The method of any one of claims 2-4, wherein, The object feature data is obtained by processing object data related to the guide page by using a main network included in the page conversion rate prediction model; The page response rate prediction value is obtained by processing the object feature data by using a page response rate branch network included in the page conversion rate prediction model; The page response conversion rate prediction value is obtained by processing the object feature data by using a page response conversion rate branch network included in the page conversion rate prediction model; The page non-response conversion rate prediction value is obtained by processing the object feature data by using a page non-response conversion rate branch network included in the page conversion rate prediction model.

6. A training method of a page conversion rate prediction model, comprising: performing feature extraction on sample object data related to a sample guide page to obtain sample object feature data; obtaining a plurality of sample page conversion rate prediction values corresponding to a plurality of sample conversion paths respectively by using the sample object feature data, wherein the plurality of sample conversion paths correspond to the sample guide page; determining a target sample page conversion rate prediction value of the sample guide page according to the plurality of sample page conversion rate prediction values; and training a predetermined model by using the target sample page conversion rate prediction value and a target sample page conversion rate real value of the sample guide page to obtain the page conversion rate prediction model; wherein the plurality of sample conversion paths include a sample page response path and a sample page non-response path corresponding to the sample guide page; the sample page response path includes a path in which an object responds to the sample guide page and conversion occurs; and the sample page non-response path includes a path in which an object does not respond to the sample guide page and conversion occurs; wherein the obtaining of the plurality of sample page conversion rate prediction values corresponding to the plurality of sample conversion paths by using the sample object feature data includes: performing page response conversion rate prediction on the sample object feature data to obtain a sample page response conversion rate prediction value corresponding to the sample page response path; and performing page non-response conversion rate prediction on the sample object feature data to obtain a sample page non-response conversion rate prediction value corresponding to the sample page non-response path; wherein the determining of the target sample page conversion rate prediction value of the sample guide page according to the plurality of sample page conversion rate prediction values includes: determining the target sample page conversion rate prediction value of the sample guide page according to the sample page response conversion rate prediction value and the sample page non-response conversion rate prediction value based on a conditional probability method.

7. The method of claim 6, wherein, The determining of the target sample page conversion rate prediction value of the sample guide page according to the sample page response conversion rate prediction value and the sample page non-response conversion rate prediction value based on the conditional probability method includes: performing page response rate prediction on the sample object feature data to obtain a sample page response rate prediction value; obtaining a sample page response conversion rate prediction probability value according to the sample page response rate prediction value and the sample page response conversion rate prediction value; and determining the target sample page conversion rate prediction value of the sample guide page according to the sample page response conversion rate prediction probability value. According to the sample page response rate prediction value and the sample page non-response conversion rate prediction value, a sample page non-response conversion rate prediction probability value is obtained; and According to the sample page response conversion rate prediction probability value and the sample page non-response conversion rate prediction probability value, a target sample page conversion rate prediction value of the sample guide page is determined.

8. The method of claim 7, wherein, The predetermined model comprises a backbone network; The feature extraction of the sample object data related to the sample guide page comprises: The sample object feature data is obtained by processing the sample object data related to the sample guide page by using the backbone network.

9. The method of claim 7 or 8, wherein, The predetermined model comprises a page response rate branch network, a page response conversion rate branch network and a page non-response conversion rate branch network; The sample page response rate prediction value is obtained by performing page response rate prediction on the sample object feature data, comprising: The sample page response rate prediction value is obtained by processing the sample object feature data by using the page response rate branch network; The sample page response conversion rate prediction value corresponding to the sample page response path is obtained by performing page response conversion rate prediction on the sample object feature data, comprising: The sample page response conversion rate prediction value is obtained by processing the sample object feature data by using the page response conversion rate branch network; The sample page non-response conversion rate prediction value corresponding to the sample page non-response path is obtained by performing page non-response conversion rate prediction on the sample object feature data, comprising: The sample page non-response conversion rate prediction value is obtained by processing the sample object feature data by using the page non-response conversion rate branch network.

10. The method of any one of claims 7-9, wherein, The target sample page conversion rate prediction value of the sample guide page is obtained by using the target sample page conversion rate prediction value and the target sample page conversion rate real value of the sample guide page, and the page conversion rate prediction model is obtained by training the predetermined model, comprising: Based on a loss function, an output value is obtained by using a sample page response rate real value and the sample page response rate prediction value, a sample page response conversion rate real value and the sample page response conversion rate prediction value, a sample page non-response conversion rate real value and the sample page non-response conversion rate prediction value, and the target sample page conversion rate prediction value and the target sample page conversion rate real value of the sample guide page; The model parameters of the predetermined model are adjusted according to the output value until a predetermined condition is met; and The predetermined model obtained when the predetermined condition is met is determined as the page conversion rate prediction model.

11. A page processing apparatus, comprising: A first obtaining module is configured to perform feature extraction on object data related to a guide page to obtain object feature data. A second obtaining module is configured to obtain a plurality of page conversion rate prediction values corresponding to a plurality of conversion paths respectively by using the object feature data, wherein the plurality of conversion paths correspond to the guide page. A first determining module is configured to determine a target page conversion rate prediction value of the guide page according to the plurality of page conversion rate prediction values; and A second determining module is configured to determine a target page conversion rate real value of the guide page according to a page conversion rate real value corresponding to each of the plurality of conversion paths. an adjusting module configured to adjust content of the guide page according to the target page conversion rate prediction value; wherein the multiple conversion paths include a page response path and a page non-response path; the page response path includes a path in which an object responds to the guide page and conversion occurs; and the page non-response path includes a path in which an object does not respond to the guide page and conversion occurs; wherein the second obtaining module comprises: a first obtaining sub-module configured to perform page response conversion rate prediction on the object feature data to obtain a page response conversion rate prediction value corresponding to the page response path; and a second obtaining sub-module configured to perform page non-response conversion rate prediction on the object feature data to obtain a page non-response conversion rate prediction value corresponding to the page non-response path. wherein the first determining module comprises: a first determining sub-module configured to determine the target page conversion rate prediction value of the guide page based on a conditional probability method according to the page response conversion rate prediction value and the page non-response conversion rate prediction value.

12. The apparatus of claim 11, wherein, The first determining sub-module comprises: a first obtaining unit configured to perform page response rate prediction on the object feature data to obtain a page response rate prediction value; a second obtaining unit configured to obtain a page response conversion rate prediction probability value according to the page response rate prediction value and the page response conversion rate prediction value; a third obtaining unit configured to obtain a page non-response conversion rate prediction probability value according to the page response rate prediction value and the page non-response conversion rate prediction value; and a first determining unit configured to determine the target page conversion rate prediction value of the guide page according to the page response conversion rate prediction probability value and the page non-response conversion rate prediction probability value.

13. The apparatus of claim 12, wherein, The second obtaining unit comprises: a first determining sub-unit configured to determine a first product between the page response rate prediction value and the page response conversion rate prediction value; and a second determining sub-unit configured to determine the first product as the page response conversion rate prediction probability value.

14. A training device of a page conversion rate prediction model, comprising: a third obtaining module configured to perform feature extraction on sample object data related to a sample guide page to obtain sample object feature data; a fourth obtaining module configured to obtain multiple sample page conversion rate prediction values corresponding to multiple sample conversion paths respectively by using the sample object feature data, wherein the multiple sample conversion paths correspond to the sample guide page; a second determining module configured to determine a target sample page conversion rate prediction value of the sample guide page according to multiple sample page conversion rate prediction values; and a fifth obtaining module configured to train a predetermined model by using the target sample page conversion rate prediction value and a target sample page conversion rate real value of the sample guide page to obtain the page conversion rate prediction model. The plurality of sample conversion paths include a sample page response path and a sample page non-response path corresponding to the sample guide page; the sample page response path includes a path in which the object responds to the sample guide page and conversion occurs; and the sample page non-response path includes a path in which the object does not respond to the sample guide page and conversion occurs. The fourth obtaining module includes a third obtaining sub-module and a fourth obtaining sub-module; the third obtaining sub-module is configured to perform page response conversion rate prediction on the sample object feature data to obtain a sample page response conversion rate prediction value corresponding to the sample page response path; and the fourth obtaining sub-module is configured to perform page non-response conversion rate prediction on the sample object feature data to obtain a sample page non-response conversion rate prediction value corresponding to the sample page non-response path. The second determining module includes a second determining sub-module configured to determine a target sample page conversion rate prediction value of the sample guide page based on a conditional probability method and according to the sample page response conversion rate prediction value and the sample page non-response conversion rate prediction value. 15.An electronic device, comprising: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5 or any one of claims 6-10.

16. A non-transitory computer readable storage medium having stored thereon computer instructions, wherein, The computer instructions are used to enable the computer to perform the method of any one of claims 1-5 or any one of claims 6-10. 17.A computer program product comprising a computer program which, when executed by a processor, implements the method of any one of claims 1-5 or any one of claims 6-10.

Images