Landing page processing method and device, landing page generating method and device, electronic equipment and readable storage medium

Abstract

The invention provides a landing page processing method and device, a landing page generating method and device, electronic equipment and a readable storage medium, and relates to the field of data processing, in particular to the technical field of artificial intelligence or front-end pages. According to the specific implementation scheme, a plurality of candidate landing pages related to a current user are obtained; on the basis of user portrait features of the current user, first page content features and second page content features, conversion rate predicted values of all the candidate landing pages are determined, and the first page content features are page content features extracted on the basis of all the candidate landing pages; the second page content feature is a page content feature extracted based on a related historical browsing page, and the related historical browsing page is a page meeting a preset condition in the historical browsing pages of the current user; a target landing page is determined from the candidate landing pages based on the conversion rate predictor. According to the scheme, the target landing page with high correlation with the user can be screened out from the candidate landing pages, so that a good putting effect is achieved, and the user conversion rate is effectively improved.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to the field of artificial intelligence or front-end page technology. Specifically, the present disclosure relates to a landing page processing, generation method, device, electronic device and readable storage medium. Background Art

[0002] A landing page is a detail page used to carry marketing activities and express key information in marketing activities. It plays an important role in user conversion.

[0003] The landing pages provided in the related art have low relevance to users. Summary of the invention

[0004] In order to solve at least one of the above-mentioned defects, the present disclosure provides a landing page processing and generation method, device, electronic device and readable storage medium.

[0005] According to a first aspect of the present disclosure, a landing page processing method is provided, the method comprising:

[0006] Get multiple candidate landing pages related to the current user;

[0007] Determine the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature and the second page content feature, where the first page content feature is the page content feature extracted based on each candidate landing page, and the second page content feature is the page content feature extracted based on the relevant historical browsing page, where the relevant historical browsing page is the page that meets the preset conditions in the historical browsing page of the current user;

[0008] Determine the target landing page from the candidate landing pages based on the conversion rate prediction value.

[0009] According to a second aspect of the present disclosure, a landing page generation method is provided, the method comprising:

[0010] Obtain historical search content of historical users and determine candidate objects related to the historical search content;

[0011] Based on the candidate objects, determine the landing page template corresponding to the candidate objects;

[0012] Acquire page data, the page data including object data of the candidate object;

[0013] Pages are assembled based on the landing page template and page data to obtain candidate landing pages. The candidate landing pages are used to filter out target landing pages, which are then displayed to users.

[0014] According to a third aspect of the present disclosure, a landing page processing device is provided, the device comprising:

[0015] A candidate landing page acquisition module is used to acquire multiple candidate landing pages related to the current user;

[0016] A conversion rate prediction module is used to determine the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature and the second page content feature, wherein the first page content feature is a page content feature extracted based on each candidate landing page, and the second page content feature is a page content feature extracted based on related historical browsing pages, wherein the related historical browsing pages are pages in the historical browsing pages of the current user that meet preset conditions;

[0017] The target landing page determination module is used to determine the target landing page from the candidate landing pages based on the conversion rate prediction value.

[0018] According to a fourth aspect of the present disclosure, a landing page generation device is provided, the device comprising:

[0019] A historical search content acquisition module is used to acquire historical search content of historical users and determine candidate objects related to the historical search content;

[0020] A landing page template determination module is used to determine the landing page template corresponding to the candidate object based on the candidate object;

[0021] A page data acquisition module, used to acquire page data, the page data including object data of candidate objects;

[0022] The candidate landing page generation module is used to assemble pages based on landing page templates and page data to obtain candidate landing pages. The candidate landing pages are used to filter out target landing pages, and the target landing pages are used to display to users.

[0023] According to a fifth aspect of the present disclosure, an electronic device is provided, the electronic device comprising:

[0024] at least one processor; and

[0025] A memory communicatively connected to the at least one processor; wherein,

[0026] The memory stores instructions that can be executed by the above-mentioned at least one processor, and the instructions are executed by the above-mentioned at least one processor so that the above-mentioned at least one processor can execute the above-mentioned landing page processing or generation method.

[0027] According to a sixth aspect 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 execute the above-mentioned landing page processing or generation method.

[0028] According to a seventh aspect of the present disclosure, a computer program product is provided, comprising a computer program, which implements the above-mentioned landing page processing or generation method when executed by a processor.

[0029] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0031] Figure 1 It is a flowchart of a landing page processing method provided by an embodiment of the present disclosure;

[0032] Figure 2 It is a schematic diagram of a specific implementation of the landing page processing method provided by the embodiment of the present disclosure;

[0033] Figure 3 It is a schematic diagram of a specific implementation of the landing page generation and processing method provided in the embodiment of the present disclosure;

[0034] Figure 4 It is a flowchart of a landing page generation method provided by an embodiment of the present disclosure;

[0035] Figure 5 It is a structural schematic diagram of a landing page processing device provided by an embodiment of the present disclosure;

[0036] Figure 6 It is a structural schematic diagram of a landing page generating device provided by an embodiment of the present disclosure;

[0037] Figure 7 It is a block diagram of an electronic device used to implement the landing page processing or generation method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0038] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0039] In the search scenario, after the user performs a search operation, guidance information will be displayed on the search results page, and the user can reach the landing page by clicking on the guidance information.

[0040] In the related art, keywords are usually bound to landing pages. When the user's search content hits the keyword, the landing page corresponding to the hit keyword will be provided to the user. The landing page provided in this solution is related to the keyword, but the relevance to the user is low, which affects the delivery effect of the landing page. If a landing page with high relevance to the user can be provided, the delivery effect of the landing page can be effectively improved and the user conversion rate can be increased.

[0041] In the related art, the promoter usually designs and creates landing pages manually. When the number of landing pages required is large, the labor cost of creating landing pages will be high. In addition, the created landing pages need to be bound to keywords, and this binding operation will also consume a lot of operating costs.

[0042] The landing page processing, generation method, device, electronic device and readable storage medium provided by the embodiments of the present disclosure are intended to solve at least one of the above-mentioned technical problems in the prior art.

[0043] Figure 1 A flowchart of a landing page processing method provided by an embodiment of the present disclosure is shown. Figure 1 As shown in , the method can mainly include:

[0044] Step S110: Acquire multiple candidate landing pages related to the current user;

[0045] Step S120: Determine the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature, and the second page content feature, where the first page content feature is a page content feature extracted based on each candidate landing page, and the second page content feature is a page content feature extracted based on related historical browsing pages, where the related historical browsing pages are pages that meet preset conditions in the historical browsing pages of the current user;

[0046] Step S130: Determine a target landing page from candidate landing pages based on the predicted conversion rate value.

[0047] Among them, candidate landing pages are created in advance. When the landing page needs to be displayed to the current user, candidate landing pages that have a certain relevance to the current user can be obtained, and the target landing page can be filtered out from these candidate landing pages and provided to the current user.

[0048] User portrait features may include multi-dimensional user features to achieve accurate description of users. When extracting user portrait features in the embodiments of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved shall comply with the provisions of relevant laws and regulations and shall not violate public order and good customs.

[0049] Page features are used to characterize the page content of a page. As an example, page keywords can be extracted from the text content of a page, and page features can be constructed based on the page keywords. The first page content feature is used to characterize the page content of a candidate landing page. The second page content feature is used to characterize the page content of a related historical browsing page.

[0050] As an example, the historical browsing pages of the current user may be determined based on the behavior log of the current user.

[0051] The relevant historical browsing pages may be pages in the historical browsing pages that have a high correlation with the current user. The second page content feature extracted based on the page content of the relevant historical browsing pages can reflect the user's preference and browsing intention.

[0052] The preset conditions can be set according to actual needs. When a historical browsing page meets the preset conditions, it means that it has a high relevance to the current user, and the historical browsing page belongs to the relevant historical browsing page.

[0053] In the disclosed embodiment, since the user portrait feature can represent some inherent characteristics of the user, the second page content feature can reflect the user's preference and browsing intention, and the first page content feature can reflect the page content of the candidate landing page, therefore, based on the user portrait feature of the current user, the first page content feature and the second page content feature, the conversion rate prediction value of each candidate landing page can be determined.

[0054] The conversion rate prediction value is the predicted probability that the user will convert after the candidate landing page is provided to the user. Based on the conversion rate prediction value, the target landing page can be screened out from the candidate landing pages. The target landing page has a strong correlation with the current user. Providing the target landing page to the current user can achieve better delivery effects and improve user conversion rate.

[0055] As an example, the candidate landing page with the highest predicted conversion rate value may be determined as the target landing page.

[0056] The method provided by the disclosed embodiment obtains multiple candidate landing pages related to the current user; determines the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature and the second page content feature, wherein the first page content feature is a page content feature extracted based on each candidate landing page, and the second page content feature is a page content feature extracted based on relevant historical browsing pages, and the relevant historical browsing pages are pages in the historical browsing pages of the current user that meet preset conditions; determines the target landing page from the candidate landing pages based on the conversion rate prediction value. Based on this solution, it is possible to screen out the target landing page that has a strong correlation with the user from the candidate landing pages, and to deliver the target landing page, which can achieve a better delivery effect and effectively improve the user conversion rate.

[0057] In an optional manner of the present disclosure, based on the user portrait feature of the current user, the first page content feature, and the second page content feature, determining the conversion rate prediction value of each candidate landing page includes:

[0058] The user portrait features of the current user and the second page content features are respectively concatenated with the first page content features as input features;

[0059] The input features are input into the pre-trained conversion rate prediction model to obtain the conversion rate prediction value of each candidate landing page output by the conversion rate prediction model.

[0060] In the disclosed embodiment, the user portrait feature, the first page content feature, and the second page content feature can all be expressed in vector form. The user portrait feature of the current user and the second page content feature are respectively concatenated with each first page content feature, which can be concatenating the feature vector of the user portrait feature, the feature vector of the second page content feature, and the feature vector of each first page content feature to obtain a feature vector of the input feature.

[0061] In the disclosed embodiment, the input features corresponding to each candidate landing page may be input into a conversion rate prediction model to obtain a conversion rate prediction value of each candidate landing page output by the conversion rate prediction model.

[0062] As an example, the user portrait feature A and the second page content feature B can be respectively spliced ​​with the first page content feature C1 of the candidate landing page c1 and the first page content feature C2 of the candidate landing page c2 to obtain input features D1 and D2, wherein the input feature D1 corresponds to the candidate landing page c1, and the input feature D1 is input into the conversion rate prediction model to obtain the conversion rate prediction value of the candidate landing page c1. The input feature D2 corresponds to the candidate landing page c2, and the input feature D3 is input into the conversion rate prediction model to obtain the conversion rate prediction value of the candidate landing page c2.

[0063] In the disclosed embodiment, the conversion rate prediction model may adopt a deep neural network (DNN) model.

[0064] In the disclosed embodiment, based on the conversion rate prediction model, the target landing page can be effectively determined according to the user's browsing intention and user preference, which helps to ensure that the target landing page has a strong correlation with the user.

[0065] In an optional manner of the present disclosure, the conversion rate prediction model is trained in the following manner:

[0066] Obtain a training sample set, the training sample set includes sample input features and conversion labels corresponding to the sample input features, wherein the sample input features are obtained by concatenating the user portrait features of the sample user and the sample second page content features with the second page content features of each sample, respectively; the sample user is associated with multiple sample candidate landing pages; the sample first page content features are page content features extracted based on each sample candidate landing page; the sample second page content features are page content features extracted based on sample-related historical browsing pages; the sample-related historical browsing pages are pages in the sample user's historical browsing pages that meet preset conditions; and the conversion labels are used to characterize the conversion status of the sample user based on each sample candidate landing page;

[0067] The initial conversion rate prediction model is trained based on the training sample set to obtain a conversion rate prediction model.

[0068] In the disclosed embodiment, a training sample set may be constructed to perform model training on an initial conversion rate prediction model to obtain a conversion rate prediction model.

[0069] In the disclosed embodiment, user portrait features of sample users in the training sample set may be extracted from a user portrait feature database.

[0070] The historical browsing pages of the sample users may be determined based on the behavior logs of the sample users.

[0071] The conversion status may include whether the sample user has a conversion behavior on the sample candidate landing page, and whether the sample user has no conversion behavior on the sample candidate landing page. The conversion status of the sample candidate landing page may be determined based on the conversion log.

[0072] The user portrait features of the sample users and the content features of the sample second pages are respectively concatenated with the content features of the second pages of each sample to obtain the sample input features.

[0073] For any sample input feature, it corresponds to the second page content feature used for feature concatenation to obtain the sample input feature, and corresponds to the sample candidate landing page corresponding to the second page content feature. The conversion label corresponding to the sample input feature is used to characterize the conversion status of the sample user based on the sample candidate landing page corresponding to the sample input feature. When the sample user has a conversion behavior on the sample candidate landing page, the corresponding conversion label is converted. When the sample user has no conversion behavior on the sample candidate landing page, the corresponding conversion label is not converted.

[0074] The first page content feature and the second page content feature may be pre-extracted and stored in a page content feature library, and extracted from the page content feature library when needed.

[0075] In the disclosed embodiment, when performing model training, the sample input features can be input into the initial conversion rate prediction model, and the conversion labels corresponding to the sample input features can be used as training targets. Multiple iterations of training are performed until the preset training stop conditions are reached to obtain the conversion rate prediction model.

[0076] In an optional manner of the present disclosure, obtaining multiple candidate landing pages related to the current user includes:

[0077] In response to detecting a search operation of the current user, obtaining current search content input by the current user;

[0078] Extract current search keywords based on current search content;

[0079] Based on the current search keyword, multiple candidate landing pages corresponding to the current keyword are determined.

[0080] In an optional method of the present disclosure, a candidate landing page may be determined based on the current search content of the current user, so that the determined candidate landing page is relevant to the search intent of the current user.

[0081] Specifically, the correspondence between search keywords and candidate landing pages can be pre-established, and when the current search keyword is extracted based on the current search content, multiple candidate landing pages corresponding to the current search keyword can be determined based on the current search keyword and the correspondence between the search keyword and the candidate landing pages.

[0082] In the embodiment of the present disclosure, the current search keyword may be a semantic keyword obtained by semantically understanding the current search content.

[0083] In an optional manner of the present disclosure, the preset conditions include:

[0084] The historical browsing page is the search result display page corresponding to the search operation;

[0085] The historical browsing pages are pages visited within a preset period of time before the search operation is performed, or the historical browsing pages are a preset number of pages visited before the search operation is performed.

[0086] In the disclosed embodiment, after the current user performs a search operation, search results may be determined based on the current search content, and the search results may be displayed through a search result display page.

[0087] In the disclosed embodiment, the search results displayed in the search results display page can effectively reflect the user's current search intention, so the search results display page can be used as a related history browsing page.

[0088] In the disclosed embodiments, the historical browsing pages are pages visited within a preset time period before a search operation is performed, or the historical browsing pages are a preset number of pages visited before a search operation is performed, that is, pages browsed by the current user immediately before a search operation. Since the user may have browsed pages related to the actual search intent before performing the current search operation, or performed a search operation related to the actual search intent, these pages browsed by the current user immediately before the search operation can reflect the user's search intent, and thus these historical browsing pages can be treated as relevant historical browsing pages.

[0089] The preset time period and the preset number can be set according to actual experience. As an example, the preset time period can be five minutes, and the relevant historical browsing pages are the pages visited by the current user within five minutes before the search operation. The preset number can be ten, and the historical browsing pages of the current user can be sorted according to the order of the time of access, and the ten historical browsing pages before the search operation in the sorting are determined as the relevant historical browsing pages.

[0090] In the embodiments of the present disclosure, different users have different page browsing frequencies. To ensure the validity of the relevant historical search pages and users, a preset time period or preset number can be effectively determined based on the user's page browsing frequency. As an example, when the user's page browsing frequency is greater than the first preset frequency, it means that the user's page browsing frequency is high. For example, the historical browsing pages accessed within the preset time period before the search operation is performed are taken as relevant historical browsing pages. Since the user's page browsing frequency is high, there will be a large number of relevant historical browsing pages within the preset time period. Too many relevant historical browsing pages will not be convenient for effective analysis of the user's browsing intention. At this time, based on the corresponding relationship between the page browsing rate and the first preset number, the first preset number corresponding to the user's page browsing frequency can be determined, and the first preset number of historical browsing pages before the search operation is performed are taken as relevant historical browsing pages.

[0091] As an example, when the user's page browsing frequency is less than the second preset frequency, it means that the user's page browsing frequency is low. For example, a preset number of historical browsing pages visited before the search operation is performed are taken as relevant historical browsing pages. Since the user's page browsing frequency is low, the preset number of related historical browsing pages may contain pages that were visited a long time ago and have low relevance to the user's current browsing intention, which is not convenient for effective analysis of the user's browsing intention. At this time, based on the correspondence between the page browsing rate and the first preset time period, the first preset time period corresponding to the user's page browsing frequency can be determined, and the historical browsing pages visited within the first preset time period before the search operation are taken as relevant historical browsing pages.

[0092] As an example, Figure 2 A schematic diagram of a specific implementation of the landing page processing method provided by an embodiment of the present disclosure is shown in the figure.

[0093] like Figure 2 As shown in , user portrait data is used to extract user portrait features. User behavior logs are used to extract historical browsing pages. Conversion log data is used to extract the conversion status of candidate landing pages and generate conversion tags. Landing page features are the page content features of the landing page.

[0094] Feature engineering, that is, extracting user portrait features, first page content features, and second page content features respectively.

[0095] In the feature library, the context feature is the second page content feature of the relevant historical browsing page, and the page content feature extracted from the candidate download landing page is the first page content feature of the candidate landing page.

[0096] Feature concatenation is to concatenate the user portrait features and the second page content features with the first page content features as input features.

[0097] The discrete DNN optimization network model, that is, the conversion rate prediction model, can input input features into the conversion rate prediction model to obtain the conversion rate prediction value of the candidate landing page.

[0098] The online landing page optimization service initiates the task of screening candidate landing pages when the current user triggers the landing page through a search operation. The conversion rate prediction model outputs the conversion rate prediction value of each candidate landing page, so that the target landing page can be screened out from the candidate landing pages based on the conversion rate prediction value. For example, the candidate landing page with the highest conversion rate prediction value is selected as the target landing page.

[0099] The preferred download landing page is the target landing page selected from the candidate landing pages.

[0100] In an optional manner of the present disclosure, after determining the target landing page from the candidate landing pages based on the conversion rate prediction value, the method further includes:

[0101] In response to detecting that the current user has triggered an operation on a guidance information control in a search result display page corresponding to the search operation, a target landing page is displayed to the current user.

[0102] In the disclosed embodiment, after the current user performs a search operation, the search results can be displayed through the search result display page. The guide information control is used to display the guide information, which may include but is not limited to advertising information. The user can trigger an access request to the landing page by clicking the guide information control, and the page can jump to the target landing page and display the target landing page to the current user.

[0103] In an optional manner of the present disclosure, the above method further includes:

[0104] Obtain historical search content of historical users and determine candidate objects related to the historical search content;

[0105] Based on the candidate objects, determine the landing page template corresponding to the candidate objects;

[0106] Acquire page data, the page data including object data of the candidate object;

[0107] Pages are assembled based on the landing page template and page data to obtain candidate landing pages.

[0108] In the disclosed embodiment, candidate landing pages can be created in advance and stored in a candidate landing page database. When the landing pages need to be displayed, the candidate landing pages can be obtained from the candidate landing page database, and then the target landing pages can be filtered based on the candidate landing pages.

[0109] In the disclosed embodiment, candidate landing pages may be constructed based on historical search rows of historical users.

[0110] The candidate objects may include but are not limited to the promoted products. The object data of the candidate objects may be product data, which may include multi-dimensional data such as product name, product details, and product image.

[0111] In the disclosed embodiment, a landing page template may be set for different candidate objects, and an association relationship between the landing page template and the candidate objects may be established.

[0112] As an example, if the candidate object is a commodity, you can configure landing page templates for different types of commodities, and associate each type of commodity with the landing page template of that type. Commodity types may include ticketing, recruitment, real estate, video, novel, game, etc.

[0113] In the disclosed embodiment, the page data may be assembled into a landing page template to obtain a candidate landing page. Specifically, the page data may include object data of a candidate object, for example, the page data includes product data of a product. In a product list consisting of multiple products in a landing page template, the page data may include product data of multiple products, and these product data may be assembled into the product list of the landing page template.

[0114] The method provided by the embodiment of the present disclosure obtains the historical search content of historical users to determine the candidate objects related to the historical search content; based on the candidate objects, determines the landing page template corresponding to the candidate objects; obtains page data, the page data includes the object data of the candidate objects; assembles the page based on the landing page template and the page data to obtain the candidate landing page. The basic scheme can realize the automatic creation of landing pages and reduce the labor cost in the process of generating landing pages.

[0115] In an optional manner of the present disclosure, obtaining page data includes:

[0116] constructing an object representation vector of the candidate object based on object data of the candidate object, the object data including an object name of the candidate object;

[0117] Construct a search content representation vector based on historical search content;

[0118] Determining the similarity between the object representation vector of the candidate object and the search content representation vector;

[0119] Object data corresponding to the target object is obtained as page data, wherein the target object is a candidate object whose similarity between the corresponding object representation vector and the search content representation vector satisfies a preset similarity condition.

[0120] In the disclosed embodiment, an object representation vector can be constructed based on the object data of the candidate object. In addition to the object name, some other dimensional data in the object data can also be used when constructing the object representation vector. For example, if the candidate object is a commodity, and the specific commodity type is a real estate commodity, in this case, in addition to the name of the real estate, the name of the city where the real estate is located can also be used to construct the object representation vector, that is, the city name and the real estate name are spliced, and the object representation vector is constructed based on the spliced ​​text.

[0121] In addition to the object name, other dimensional data used in the object representation vector can be determined according to the situation of the candidate object and the distribution of the search keywords corresponding to the candidate object.

[0122] In the disclosed embodiment, semantic understanding can be performed on historical search content to extract semantic keywords, and a search content representation vector can be constructed based on the extracted semantic keywords.

[0123] The similarity between the object representation vector of the candidate object and the search content representation vector can characterize the correlation between the historical search content and the candidate object, and improve the effectiveness of the candidate object by screening out target objects with strong correlation with the historical search content.

[0124] As an example, a graph-based Ernie-sage model can be constructed to perform vector representation on candidate objects and historical search content, and then the Approximate Nearest Neighbor (ANN) algorithm is used to calculate the vector distance between the object representation vector of the candidate object and the search content representation vector, and the similarity between the object representation vector and the search content representation vector is measured by the vector distance.

[0125] In an optional manner of the present disclosure, the page data further includes page guide information, and the page guide information is determined based on historical search content and / or object data of the candidate object.

[0126] In the disclosed embodiment, the page data also includes page guidance information, such as the title and header image of the landing page, guidance text and footer image for guiding related download operations, etc.

[0127] In the disclosed embodiments, page guidance information can be generated based on the semantic understanding results of historical search content, for example: a landing page title that describes the landing page as a whole can be generated based on the semantic understanding results of historical search content. When the historical search content contains less information, page guidance information can also be determined based on the object data of the candidate object. For example, a guidance copy is generated based on the product name and location of the first product in the product list.

[0128] As an example, the user's historical search content is "How to make scrambled eggs with tomatoes", and the title of the landing page determined at this time can be "If you want to learn how to make scrambled eggs with tomatoes, come to xx website."

[0129] As another example, when the user's historical search content is "hotel", the title of the first product "xx hotel" can be extracted from the product list, and the guiding copy determined at this time can be "If you want to know more about xx hotel, please download xx application."

[0130] In the disclosed embodiment, the candidate landing page may include a local dynamic page, a global dynamic page, and a pure static page, wherein the local dynamic page is a candidate landing page obtained by assembling the object data of the candidate object into the landing page template. The global dynamic page is a candidate landing page obtained by assembling the object data of the candidate object and the page guide information into the landing page template. A pure static page is a pre-created page, which can be provided by the promoter and can also be used as a candidate landing page.

[0131] Since the global dynamic page is assembled from multiple page data and landing page templates, the global dynamic page is more relevant to the current user's current search content. When screening the target landing page from the candidate landing pages, you can prioritize screening from the global dynamic page. If you cannot screen out a valid target landing page from the global dynamic page (for example, the predicted conversion rate of the screened target landing page is too low), you can screen the target landing page from the local dynamic page. If you cannot screen out a valid target landing page from the local dynamic page (for example, the predicted conversion rate of the screened target landing page is too low), you can screen the target landing page from the pure static page.

[0132] By setting the screening priority of global dynamic pages, local dynamic pages and pure static pages, it helps to effectively screen the target landing pages.

[0133] In an optional manner of the present disclosure, page assembly is performed based on a landing page template and page data to obtain a candidate landing page, including:

[0134] Get the historical conversion rate of candidate landing pages generated based on page data;

[0135] Determine the page data whose historical conversion rate of the corresponding candidate landing page meets the preset conversion rate condition as the target page data;

[0136] Based on the landing page template, the page is assembled with the target page data to determine the candidate landing pages.

[0137] In the disclosed embodiment, the historical conversion rates of candidate landing pages generated by the page data may be obtained, and the page data may be screened based on the historical conversion rates.

[0138] Specifically, when the historical conversion rate of the candidate landing page corresponding to the page data does not meet the preset conversion rate condition, it means that the historical conversion rate corresponding to the page data is low, and the landing page generated based on the page data has an unsatisfactory delivery effect. In this case, this part of the page data can be filtered out. When the historical conversion rate of the candidate landing page corresponding to the page data meets the preset conversion rate condition, it means that the historical conversion rate corresponding to the page data is high, and the landing page generated based on the page data has a good delivery effect. In this case, this part of the page data is determined as the target page data, and the candidate landing page is generated based on the target page data.

[0139] As an example, Figure 3 A schematic diagram of a specific implementation of the landing page generation and processing method provided in an embodiment of the present disclosure is shown in FIG.

[0140] The solution includes two stages: page generation and page optimization.

[0141] Industry templates are landing page templates. Marketing plans are used to provide pure static pages. Structured materials are page data. Ad libraries and view logs can provide landing page delivery effect data.

[0142] Content understanding, that is, semantic understanding of the user's search content and building a search content vector.

[0143] Content generation is to extract semantic keywords based on semantic understanding of search content and construct search content vectors based on semantic keywords.

[0144] Structured recall refers to recalling products corresponding to product vectors that have a high correlation with the search content vector, and assembling the product data of these products with the landing page template.

[0145] Combine and assemble, assemble pages based on product data and landing page templates to obtain candidate landing pages.

[0146] In this example, candidate landing pages may include: purely static pages, local dynamic pages, and global dynamic pages.

[0147] In the page generation stage, you can build multi-dimensional product data and a landing page template. Use the similarity between the product vector and the search content vector to recall highly effective products, assemble the product data of the retrieved products with the landing page template corresponding to the product, and get a candidate landing page.

[0148] User query refers to the current search content of the current user. User portrait refers to the user portrait features.

[0149] In the page optimization stage, candidate landing pages related to the current user can be obtained based on the current search content of the current user, and the target landing page can be optimized from the candidate landing pages based on user portrait features, first page content features, and second page content features, and based on a click-through rate prediction model.

[0150] Figure 4 A schematic diagram of a method for generating a landing page provided by an embodiment of the present disclosure is shown. Figure 4 As shown in , the method can mainly include:

[0151] Step S410: Obtain the historical search content of the historical user and determine the candidate objects related to the historical search content;

[0152] Step S420: Based on the candidate object, determine the landing page template corresponding to the candidate object;

[0153] Step S430: acquiring page data, the page data including object data of the candidate object;

[0154] Step S440: Assemble the page based on the landing page template and the page data to obtain candidate landing pages. The candidate landing pages are used to filter out target landing pages, and the target landing pages are used to display to the user.

[0155] In the disclosed embodiment, candidate landing pages can be created in advance and stored in a candidate landing page database. When the landing pages need to be displayed, the candidate landing pages can be obtained from the candidate landing page database, and then the target landing pages can be filtered based on the candidate landing pages.

[0156] In the disclosed embodiment, candidate landing pages may be constructed based on historical search rows of historical users.

[0157] The candidate objects may include but are not limited to the promoted products. The object data of the candidate objects may be product data, which may include multi-dimensional data such as product name, product details, and product image.

[0158] In the disclosed embodiment, a landing page template may be set for different candidate objects, and an association relationship between the landing page template and the candidate objects may be established.

[0159] As an example, if the candidate object is a commodity, you can configure landing page templates for different types of commodities, and associate each type of commodity with the landing page template of that type. Commodity types may include ticketing, recruitment, real estate, video, novel, game, etc.

[0160] In the disclosed embodiment, the page data may be assembled into a landing page template to obtain a candidate landing page. Specifically, the page data may include object data of a candidate object, for example, the page data includes product data of a product. In a product list consisting of multiple products in a landing page template, the page data may include product data of multiple products, and these product data may be assembled into the product list of the landing page template.

[0161] The method provided by the embodiment of the present disclosure obtains the historical search content of historical users to determine the candidate objects related to the historical search content; based on the candidate objects, determines the landing page template corresponding to the candidate objects; obtains page data, the page data includes the object data of the candidate objects; assembles the page based on the landing page template and the page data to obtain the candidate landing page. The basic scheme can realize the automatic creation of landing pages and reduce the labor cost in the process of generating landing pages.

[0162] In an optional manner of the present disclosure, obtaining page data includes:

[0163] constructing an object representation vector of the candidate object based on object data of the candidate object, the object data including an object name of the candidate object;

[0164] Construct a search content representation vector based on historical search content;

[0165] Determining the similarity between the object representation vector of the candidate object and the search content representation vector;

[0166] Object data corresponding to the target object is obtained as page data, wherein the target object is a candidate object whose similarity between the corresponding object representation vector and the search content representation vector satisfies a preset similarity condition.

[0167] In the disclosed embodiment, an object representation vector can be constructed based on the object data of the candidate object. In addition to the object name, some other dimensional data in the object data can also be used when constructing the object representation vector. For example, if the candidate object is a commodity, and the specific commodity type is a real estate commodity, in this case, in addition to the name of the real estate, the name of the city where the real estate is located can also be used to construct the object representation vector, that is, the city name and the real estate name are spliced, and the object representation vector is constructed based on the spliced ​​text.

[0168] In addition to the object name, other dimensional data used in the object representation vector can be determined according to the situation of the candidate object and the distribution of the search keywords corresponding to the candidate object.

[0169] In the disclosed embodiment, semantic understanding can be performed on historical search content to extract semantic keywords, and a search content representation vector can be constructed based on the extracted semantic keywords.

[0170] The similarity between the object representation vector of the candidate object and the search content representation vector can characterize the correlation between the historical search content and the candidate object, and improve the effectiveness of the candidate object by screening out target objects with strong correlation with the historical search content.

[0171] As an example, a graph-based Ernie-sage model can be constructed to perform vector representation on candidate objects and historical search content, and then the Approximate Nearest Neighbor (ANN) algorithm is used to calculate the vector distance between the object representation vector of the candidate object and the search content representation vector, and the similarity between the object representation vector and the search content representation vector is measured by the vector distance.

[0172] In an optional manner of the present disclosure, the page data further includes page guide information, and the page guide information is determined based on historical search content and / or object data of the candidate object.

[0173] In the disclosed embodiment, the page data also includes page guidance information, such as the title and header image of the landing page, guidance text and footer image for guiding related download operations, etc.

[0174] In the disclosed embodiments, page guidance information can be generated based on the semantic understanding results of historical search content, for example: a landing page title that describes the landing page as a whole can be generated based on the semantic understanding results of historical search content. When the historical search content contains less information, page guidance information can also be determined based on the object data of the candidate object. For example, a guidance copy is generated based on the product name and location of the first product in the product list.

[0175] As an example, the user's historical search content is "How to make scrambled eggs with tomatoes", and the title of the landing page determined at this time can be "If you want to learn how to make scrambled eggs with tomatoes, come to xx website."

[0176] As another example, when the user's historical search content is "hotel", the title of the first product "xx hotel" can be extracted from the product list, and the guiding copy determined at this time can be "If you want to know more about xx hotel, please download xx application."

[0177] In the disclosed embodiment, the candidate landing page may include a local dynamic page, a global dynamic page, and a pure static page, wherein the local dynamic page is a candidate landing page obtained by assembling the object data of the candidate object into the landing page template. The global dynamic page is a candidate landing page obtained by assembling the object data of the candidate object and the page guide information into the landing page template. A pure static page is a pre-created page, which can be provided by the promoter and can also be used as a candidate landing page.

[0178] Since the global dynamic page is assembled from multiple page data and landing page templates, the global dynamic page is more relevant to the current user's current search content. When screening the target landing page from the candidate landing pages, you can prioritize screening from the global dynamic page. If you cannot screen out a valid target landing page from the global dynamic page (for example, the predicted conversion rate of the screened target landing page is too low), you can screen the target landing page from the local dynamic page. If you cannot screen out a valid target landing page from the local dynamic page (for example, the predicted conversion rate of the screened target landing page is too low), you can screen the target landing page from the pure static page.

[0179] By setting the screening priority of global dynamic pages, local dynamic pages and pure static pages, it helps to effectively screen the target landing pages.

[0180] In an optional manner of the present disclosure, determining a candidate landing page based on a landing page template and page data includes:

[0181] Get the historical conversion rate of candidate landing pages generated based on page data;

[0182] Determine the page data whose historical conversion rate of the corresponding candidate landing page meets the preset conversion rate condition as the target page data;

[0183] Determine candidate landing pages based on the landing page template and target page data.

[0184] In the disclosed embodiment, the historical conversion rates of candidate landing pages generated by the page data may be obtained, and the page data may be screened based on the historical conversion rates.

[0185] Specifically, when the historical conversion rate of the candidate landing page corresponding to the page data does not meet the preset conversion rate condition, it means that the historical conversion rate corresponding to the page data is low, and the landing page generated based on the page data has an unsatisfactory delivery effect. In this case, this part of the page data can be filtered out. When the historical conversion rate of the candidate landing page corresponding to the page data meets the preset conversion rate condition, it means that the historical conversion rate corresponding to the page data is high, and the landing page generated based on the page data has a good delivery effect. In this case, this part of the page data is determined as the target page data, and the candidate landing page is generated based on the target page data.

[0186] Based on Figure 1 The same principle as shown in the method, Figure 5 A schematic diagram of the structure of a landing page processing device provided by an embodiment of the present disclosure is shown. Figure 5 As shown, the landing page processing device 50 may include:

[0187] A candidate landing page acquisition module 510 is used to acquire multiple candidate landing pages related to the current user;

[0188] The conversion rate prediction module 520 is used to determine the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature and the second page content feature, wherein the first page content feature is a page content feature extracted based on each candidate landing page, and the second page content feature is a page content feature extracted based on related historical browsing pages, wherein the related historical browsing pages are pages in the historical browsing pages of the current user that meet the preset conditions;

[0189] The target landing page determination module 530 is used to determine the target landing page from the candidate landing pages based on the conversion rate prediction value.

[0190] The device provided by the disclosed embodiment obtains multiple candidate landing pages related to the current user; determines the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature and the second page content feature, wherein the first page content feature is a page content feature extracted based on each candidate landing page, and the second page content feature is a page content feature extracted based on relevant historical browsing pages, and the relevant historical browsing pages are pages in the historical browsing pages of the current user that meet preset conditions; determines the target landing page from the candidate landing pages based on the conversion rate prediction value. Based on this solution, it is possible to screen out the target landing page that has a strong correlation with the user from the candidate landing pages, and to deliver the target landing page, which can achieve a better delivery effect and effectively improve the user conversion rate.

[0191] Optionally, when determining the predicted conversion rate value of each candidate landing page based on the user portrait feature of the current user, the first page content feature, and the second page content feature, the conversion rate prediction module is specifically used to:

[0192] The user portrait features of the current user and the second page content features are respectively concatenated with the first page content features as input features;

[0193] The input features are input into the pre-trained conversion rate prediction model to obtain the conversion rate prediction value of each candidate landing page output by the conversion rate prediction model.

[0194] Optionally, the conversion rate prediction model is trained in the following manner:

[0195] Obtain a training sample set, the training sample set includes sample input features and conversion labels corresponding to the sample input features, wherein the sample input features are obtained by concatenating the user portrait features of the sample user and the sample second page content features with the second page content features of each sample, respectively; the sample user is associated with multiple sample candidate landing pages; the sample first page content features are page content features extracted based on each sample candidate landing page; the sample second page content features are page content features extracted based on sample-related historical browsing pages; the sample-related historical browsing pages are pages in the sample user's historical browsing pages that meet preset conditions; and the conversion labels are used to characterize the conversion status of the sample user based on each sample candidate landing page;

[0196] The initial conversion rate prediction model is trained based on the training sample set to obtain a conversion rate prediction model.

[0197] Optionally, the candidate landing page acquisition module is specifically used to:

[0198] In response to detecting a search operation of the current user, obtaining current search content input by the current user;

[0199] Extract current search keywords based on current search content;

[0200] Based on the current search keyword, multiple candidate landing pages corresponding to the current keyword are determined.

[0201] Optionally, the preset conditions include:

[0202] The historical browsing page is the search result display page corresponding to the search operation;

[0203] The historical browsing pages are pages visited within a preset period of time before the search operation is performed, or the historical browsing pages are a preset number of pages visited before the search operation is performed.

[0204] Optionally, the above device further includes:

[0205] The landing page display module is used to display the target landing page to the current user after determining the target landing page from the candidate landing pages based on the conversion rate prediction value, in response to detecting the current user's triggering operation on the guidance information control in the search result display page corresponding to the search operation.

[0206] Optionally, the above device further includes:

[0207] A historical search content acquisition module is used to acquire historical search content of historical users and determine candidate objects related to the historical search content;

[0208] A landing page template determination module is used to determine the landing page template corresponding to the candidate object based on the candidate object;

[0209] A page data acquisition module, used to acquire page data, the page data including object data of candidate objects;

[0210] The candidate landing page generation module is used to assemble pages based on landing page templates and page data to obtain candidate landing pages.

[0211] Optionally, the page data acquisition module is specifically used for:

[0212] constructing an object representation vector of the candidate object based on object data of the candidate object, the object data including an object name of the candidate object;

[0213] Construct a search content representation vector based on historical search content;

[0214] Determining the similarity between the object representation vector of the candidate object and the search content representation vector;

[0215] Object data corresponding to the target object is obtained as page data, wherein the target object is a candidate object whose similarity between the corresponding object representation vector and the search content representation vector satisfies a preset similarity condition.

[0216] Optionally, the page data further includes page guide information, and the page guide information is determined based on historical search content and / or object data of the candidate object.

[0217] Optionally, the candidate landing page generation module is specifically used to:

[0218] Get the historical conversion rate of candidate landing pages generated based on page data;

[0219] Determine the page data whose historical conversion rate of the corresponding candidate landing page meets the preset conversion rate condition as the target page data;

[0220] Determine candidate landing pages based on the landing page template and target page data.

[0221] It can be understood that the above modules of the landing page processing device in the embodiment of the present disclosure have the function of realizing Figure 1 The functions of the corresponding steps of the landing page processing method in the embodiment shown in . This function can be implemented by hardware, or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions. The above modules can be software and / or hardware, and the above modules can be implemented separately or integrated with multiple modules. For the functional description of each module of the above landing page processing device, please refer to Figure 1 The corresponding description of the landing page processing method in the embodiment shown in will not be repeated here.

[0222] Based on Figure 4 The same principle as shown in the method, Figure 6 A schematic diagram of the structure of a landing page generating device provided by an embodiment of the present disclosure is shown. Figure 6 As shown, the landing page generating device 60 may include:

[0223] A historical search content acquisition module 610 is used to acquire historical search content of historical users and determine candidate objects related to the historical search content;

[0224] A landing page template determination module 620 is used to determine a landing page template corresponding to a candidate object based on the candidate object;

[0225] A page data acquisition module 630 is used to acquire page data, where the page data includes object data of the candidate object;

[0226] The candidate landing page generation module 640 is used to assemble pages based on landing page templates and page data to obtain candidate landing pages. The candidate landing pages are used to filter out target landing pages, and the target landing pages are used to display to users.

[0227] The device provided by the embodiment of the present disclosure obtains the historical search content of the historical user to determine the candidate object related to the historical search content; based on the candidate object, determines the landing page template corresponding to the candidate object; obtains page data, the page data includes the object data of the candidate object; assembles the page based on the landing page template and the page data to obtain the candidate landing page. The basic scheme can realize the automatic creation of the landing page and reduce the labor cost in the process of generating the landing page.

[0228] Optionally, the page data acquisition module is specifically used for:

[0229] constructing an object representation vector of the candidate object based on object data of the candidate object, the object data including an object name of the candidate object;

[0230] Construct a search content representation vector based on historical search content;

[0231] Determining the similarity between the object representation vector of the candidate object and the search content representation vector;

[0232] Object data corresponding to the target object is obtained as page data, wherein the target object is a candidate object whose similarity between the corresponding object representation vector and the search content representation vector satisfies a preset similarity condition.

[0233] Optionally, the page data further includes page guide information, and the page guide information is determined based on historical search content and / or object data of the candidate object.

[0234] Optionally, the candidate landing page generation module is specifically used to:

[0235] Get the historical conversion rate of candidate landing pages generated based on page data;

[0236] Determine the page data whose historical conversion rate of the corresponding candidate landing page meets the preset conversion rate condition as the target page data;

[0237] Determine candidate landing pages based on the landing page template and target page data.

[0238] It can be understood that the above modules of the landing page generation device in the embodiment of the present disclosure have the function of realizing Figure 4 The functions of the corresponding steps of the landing page generation method in the embodiment shown in . This function can be implemented by hardware, or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions. The above modules can be software and / or hardware, and the above modules can be implemented separately or integrated with multiple modules. For the functional description of each module of the above landing page generation device, please refer to Figure 4 The corresponding description of the landing page generation method in the embodiment shown in will not be repeated here.

[0239] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0240] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0241] The electronic device includes: 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 so that the at least one processor can execute a landing page processing or generation method as provided in an embodiment of the present disclosure.

[0242] Compared with the prior art, this electronic device obtains multiple candidate landing pages related to the current user; determines the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature and the second page content feature, the first page content feature is the page content feature extracted based on each candidate landing page, the second page content feature is the page content feature extracted based on the relevant historical browsing page, the relevant historical browsing page is the page in the historical browsing page of the current user that meets the preset conditions; determines the target landing page from the candidate landing pages based on the conversion rate prediction value. Based on this solution, it is possible to screen out the target landing page that has a strong correlation with the user from the candidate landing pages, and put the target landing page on the market, which can achieve a better delivery effect and effectively improve the user conversion rate.

[0243] The readable storage medium is a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute a landing page processing or generation method as provided in an embodiment of the present disclosure.

[0244] Compared with the prior art, the readable storage medium obtains multiple candidate landing pages related to the current user; determines the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature and the second page content feature, wherein the first page content feature is a page content feature extracted based on each candidate landing page, and the second page content feature is a page content feature extracted based on related historical browsing pages, and the related historical browsing pages are pages in the current user's historical browsing pages that meet preset conditions; determines the target landing page from the candidate landing pages based on the conversion rate prediction value. Based on this solution, it is possible to screen out the target landing page that has a strong correlation with the user from the candidate landing pages, and to deliver the target landing page, which can achieve a better delivery effect and effectively improve the user conversion rate.

[0245] The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the landing page processing or generation method provided in the embodiment of the present disclosure.

[0246] Compared with the prior art, this computer program product obtains multiple candidate landing pages related to the current user; determines the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature and the second page content feature, the first page content feature is the page content feature extracted based on each candidate landing page, the second page content feature is the page content feature extracted based on the relevant historical browsing page, and the relevant historical browsing page is the page that meets the preset conditions in the historical browsing page of the current user; determines the target landing page from the candidate landing pages based on the conversion rate prediction value. Based on this solution, it is possible to screen out the target landing page that has a strong correlation with the user from the candidate landing pages, and put the target landing page into service, which can achieve a better delivery effect and effectively improve the user conversion rate.

[0247] Figure 7 A schematic block diagram of an example electronic device 70 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0248] like Figure 7 As shown, the electronic device 70 includes a computing unit 710, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 720 or a computer program loaded from a storage unit 780 into a random access memory (RAM) 730. In the RAM 730, various programs and data required for the operation of the device 70 can also be stored. The computing unit 710, the ROM 720, and the RAM 730 are connected to each other via a bus 740. An input / output (I / O) interface 750 is also connected to the bus 740.

[0249] A number of components in the device 70 are connected to the I / O interface 750, including: an input unit 760, such as a keyboard, a mouse, etc.; an output unit 770, such as various types of displays, speakers, etc.; a storage unit 780, such as a disk, an optical disk, etc.; and a communication unit 790, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 790 allows the device 70 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0250] The computing unit 710 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 710 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 710 performs the landing page processing or generation method provided in the embodiments of the present disclosure. For example, in some embodiments, the execution of the landing page processing or generation method provided in the embodiments of the present disclosure may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 780. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 70 via the ROM 720 and / or the communication unit 790. When the computer program is loaded into the RAM 730 and executed by the computing unit 710, one or more steps of the landing page processing or generation method provided in the embodiments of the present disclosure may be performed. Alternatively, in other embodiments, the computing unit 710 may be configured in any other appropriate manner (for example, by means of firmware) to execute the landing page processing or generation method provided in the embodiments of the present disclosure.

[0251] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0252] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0253] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0254] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the 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 acoustic input, voice input, or tactile input).

[0255] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0256] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0257] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0258] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A landing page processing method, comprising: Get multiple candidate landing pages related to the current user; Determine the conversion rate prediction value of each of the candidate landing pages based on the user portrait feature of the current user, the first page content feature, and the second page content feature, wherein the first page content feature is a page content feature extracted based on each of the candidate landing pages, and the second page content feature is a page content feature extracted based on related historical browsing pages, and the related historical browsing pages are pages in the historical browsing pages of the current user that meet preset conditions; A target landing page is determined from the candidate landing pages based on the conversion rate prediction value.

2. The method according to claim 1, wherein: The step of determining the predicted conversion rate value of each of the candidate landing pages based on the user portrait feature of the current user, the first page content feature, and the second page content feature includes: splicing the user portrait feature of the current user and the second page content feature with each of the first page content features as input features; The input features are input into a pre-trained conversion rate prediction model to obtain a conversion rate prediction value of each candidate landing page output by the conversion rate prediction model.

3. The method according to claim 2, wherein: The conversion rate prediction model is trained in the following way: Obtain a training sample set, the training sample set including sample input features and conversion labels corresponding to the sample input features, wherein the sample input features are obtained by respectively concatenating user portrait features of sample users and sample second page content features with the second page content features of each sample, the sample user is associated with multiple sample candidate landing pages, the sample first page content features are page content features extracted based on each of the sample candidate landing pages, the sample second page content features are page content features extracted based on sample-related historical browsing pages, the sample-related historical browsing pages are pages in the sample user's historical browsing pages that meet preset conditions, and the conversion labels are used to characterize the conversion status of the sample user based on each of the sample candidate landing pages; The initial conversion rate prediction model is trained based on the training sample set to obtain the conversion rate prediction model.

4. The method according to any one of claims 1 to 3, wherein: Get multiple candidate landing pages related to the current user, including: In response to detecting a search operation of a current user, obtaining current search content input by the current user; Extracting current search keywords based on the current search content; Based on the current search keyword, multiple candidate landing pages corresponding to the current keyword are determined.

5. The method according to claim 4, wherein: The preset conditions include: The historical browsing page is a search result display page corresponding to the search operation; The historical browsing pages are pages accessed within a preset period of time before the search operation is performed, or the historical browsing pages are a preset number of pages accessed before the search operation is performed.

6. The method according to claim 4 or 5, after determining the target landing page from the candidate landing pages based on the conversion rate prediction value, the method further comprises: In response to detecting that the current user has triggered an operation on a guide information control in a search result display page corresponding to the search operation, the target landing page is displayed to the current user.

7. The method according to any one of claims 1 to 6, further comprising: Acquire historical search content of historical users, and determine candidate objects related to the historical search content; Based on the candidate object, determine a landing page template corresponding to the candidate object; Acquire page data, wherein the page data includes object data of the candidate object; The page is assembled based on the landing page template and the page data to obtain the candidate landing page.

8. The method according to claim 7, wherein: The obtaining of page data includes: constructing an object representation vector of the candidate object based on object data of the candidate object, wherein the object data includes an object name of the candidate object; Constructing a search content representation vector based on the historical search content; Determining the similarity between the object representation vector of the candidate object and the search content representation vector; Object data corresponding to a target object is obtained as page data, wherein the target object is a candidate object whose similarity between a corresponding object representation vector and the search content representation vector satisfies a preset similarity condition.

9. The method according to claim 7 or 8, wherein: The page data further includes page guide information, which is determined based on the historical search content and / or the object data of the candidate object.

10. The method according to any one of claims 7 to 9, wherein: The step of assembling a page based on the landing page template and the page data to obtain the candidate landing page includes: Obtain historical conversion rates of candidate landing pages generated based on the page data; Determine the page data whose historical conversion rate of the corresponding candidate landing page meets the preset conversion rate condition as the target page data; The page is assembled based on the landing page template and the target page data to determine a candidate landing page.

11. A landing page generation method, comprising: Acquire historical search content of historical users, and determine candidate objects related to the historical search content; Based on the candidate object, determine a landing page template corresponding to the candidate object; Acquire page data, wherein the page data includes object data of the candidate object; The page is assembled based on the landing page template and the page data to obtain the candidate landing page, and the candidate landing page is used to filter and obtain the target landing page, and the target landing page is used to display to the user.

12. The method according to claim 11, wherein: The obtaining of page data includes: constructing an object representation vector of the candidate object based on object data of the candidate object, wherein the object data includes an object name of the candidate object; Constructing a search content representation vector based on the historical search content; Determining the similarity between the object representation vector of the candidate object and the search content representation vector; Object data corresponding to a target object is obtained as page data, wherein the target object is a candidate object whose similarity between a corresponding object representation vector and the search content representation vector satisfies a preset similarity condition.

13. The method according to claim 11 or 12, wherein: The page data further includes page guide information, and the page guide information is determined by the historical search content and / or the object data of the candidate object.

14. The method according to any one of claims 11 to 13, wherein: The step of assembling a page based on the landing page template and the page data to obtain the candidate landing page includes: Obtain historical conversion rates of candidate landing pages generated based on the page data; Determine the page data whose historical conversion rate of the corresponding candidate landing page meets the preset conversion rate condition as the target page data; The page is assembled based on the landing page template and the target page data to determine a candidate landing page.

15. A landing page processing device, comprising: A candidate landing page acquisition module is used to acquire multiple candidate landing pages related to the current user; A conversion rate prediction module, used to determine the conversion rate prediction value of each candidate landing page based on the user portrait feature of the current user, the first page content feature and the second page content feature, wherein the first page content feature is a page content feature extracted based on each candidate landing page, and the second page content feature is a page content feature extracted based on related historical browsing pages, and the related historical browsing pages are pages in the historical browsing pages of the current user that meet preset conditions; A target landing page determination module is used to determine a target landing page from the candidate landing pages based on the conversion rate prediction value.

16. The device according to claim 15, wherein: When the conversion rate prediction module determines the conversion rate prediction value of each of the candidate landing pages based on the user portrait feature of the current user, the first page content feature, and the second page content feature, it is specifically used to: splicing the user portrait feature of the current user and the second page content feature with each of the first page content features as input features; The input features are input into a pre-trained conversion rate prediction model to obtain a conversion rate prediction value of each candidate landing page output by the conversion rate prediction model.

17. A landing page generation device, comprising: A historical search content acquisition module is used to acquire historical search content of historical users and determine candidate objects related to the historical search content; A landing page template determination module, used to determine a landing page template corresponding to the candidate object based on the candidate object; A page data acquisition module, used to acquire page data, wherein the page data includes object data of the candidate object; The candidate landing page generation module is used to assemble pages based on the landing page template and the page data to obtain the candidate landing page. The candidate landing page is used to filter and obtain the target landing page, and the target landing page is used to display to the user.

18. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed 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 according to any one of claims 1 to 14.

19. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-14.

20. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 14.

Images