Page generation method and device, electronic equipment, storage medium and program product

Abstract

The invention discloses a page generation method and device, electronic equipment, a storage medium and a program product. According to the method, interaction element identification can be performed on the to-be-identified image to obtain element information of a plurality of interaction elements; according to the element information of the plurality of interaction elements, determining interaction association among the interaction elements; on the basis of the element information of the interactive elements and the interactive association between the interactive elements, the interactive elements are converted into a task execution model of the interactive elements, and the task execution model comprises at least one task execution element used for achieving the interactive tasks corresponding to the interactive elements; determining interaction configuration information corresponding to the task execution elements; based on the interaction configuration information, an interaction page is generated, and an interaction component corresponding to the interaction configuration information is displayed on the interaction page. According to the method and the device, the efficiency of making the interactive page can be improved, and errors are reduced.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a page generation method, device, electronic device, storage medium and program product. Background Art

[0002] With the rapid development of computer technology, interactive pages (such as H5 pages, etc.) that can achieve complex user interaction and interaction are increasingly widely used in various operational activities. Especially in product promotion, product acquisition and other product operation activities, due to the diversity and immediacy of product operation activities, interactive pages have become an indispensable tool.

[0003] However, the current process of creating interactive pages usually relies on manual operations, which is not only inefficient but also prone to errors. Summary of the invention

[0004] The embodiments of the present application provide a page generation method, device, electronic device, storage medium and program product, which can improve the efficiency of making interactive pages and reduce errors.

[0005] An embodiment of the present application provides a page generation method, including: performing interactive element recognition on an image to be recognized to obtain element information of multiple interactive elements, wherein the interactive elements are image elements displayed in the image to be recognized and related to interactive tasks; determining the interactive associations between the interactive elements based on the element information of the multiple interactive elements; converting the interactive elements into task execution models of the interactive elements based on the element information of the interactive elements and the interactive associations between the interactive elements, wherein the task execution model includes at least one task execution element for implementing the interactive tasks corresponding to the interactive elements; determining interactive configuration information corresponding to the task execution element; and generating an interactive page based on the interactive configuration information, wherein the interactive page displays interactive components corresponding to the interactive configuration information, and the interactive components are used to implement the interactive tasks corresponding to the interactive elements in the interactive page.

[0006] The embodiment of the present application also provides a page generation device, including: an element recognition unit, used to perform interactive element recognition on an image to be recognized, and obtain element information of multiple interactive elements, wherein the interactive elements are image elements related to interactive tasks displayed in the image to be recognized; a relationship determination unit, used to determine the interactive associations between the interactive elements based on the element information of the multiple interactive elements; an element conversion unit, used to convert the interactive elements into task execution models of the interactive elements based on the element information of the interactive elements and the interactive associations between the interactive elements, wherein the task execution model includes at least one task execution element for implementing the interactive task corresponding to the interactive element; a configuration determination unit, used to determine the interactive configuration information corresponding to the task execution element; a page generation unit, used to generate an interactive page based on the interactive configuration information, wherein the interactive page displays interactive components corresponding to the interactive configuration information, and the interactive components are used to implement the interactive tasks corresponding to the interactive elements in the interactive page.

[0007] In some embodiments, the element information includes a task type of the interactive task corresponding to the interactive element and description information of the interactive element, and the element identification unit is used to: perform interactive element detection on the image to be identified, and determine the description information of multiple interactive elements displayed in the image to be identified; classify the interactive elements based on the description information of the interactive elements to obtain the task type of the interactive task corresponding to the interactive element.

[0008] In some embodiments, the interactive element detection on the image to be identified and determining the description information of the multiple interactive elements displayed in the image to be identified include: performing interactive element detection on the image to be identified and extracting the display information of the interactive elements in the image to be identified; based on the attribute information of the image to be identified, extracting feature information of the display information of the interactive elements to obtain the description information of the interactive elements, wherein the attribute information includes at least one of the theme information and style information of the image to be identified.

[0009] In some embodiments, the relationship determination unit is specifically used to: determine the semantic association between the element information of the interactive elements based on the element information of the multiple interactive elements; and determine the interactive association between the interactive elements based on the semantic association between the element information.

[0010] In some embodiments, the element conversion unit is specifically used to: perform semantic understanding of the element information of the interactive element and the interactive association between the interactive elements to extract a first context feature of the interactive element, wherein the first context feature represents the semantic association between the interactive elements; decode the first context feature to obtain a task execution model of the interactive element, wherein the task execution element in the task execution model is determined based on the semantic association between the interactive elements represented by the first context feature.

[0011] In some embodiments, the semantic understanding of the element information of the interactive elements and the interactive associations between the interactive elements to extract the first context features of the interactive elements includes: extracting keywords related to the interactive tasks corresponding to the interactive elements from the element information of the interactive elements; and performing semantic understanding of the keywords and the interactive associations between the interactive elements to extract the first context features of the interactive elements.

[0012] In some implementations, the element conversion unit is further used to: based on the element information of the interactive element, split the interactive element into at least one sub-element, and the sub-element is used to convert to obtain the task execution model.

[0013] In some embodiments, the configuration determination unit is specifically used to: obtain at least one guidance information, the guidance information including configuration information corresponding to a specified scene; determine the guidance information to be used corresponding to the task execution element from the at least one guidance information based on the element information of the interactive element corresponding to the task execution element; perform semantic understanding on the guidance information to be used and the task execution element to extract a second context feature of the task execution element, the second context feature representing the semantic association between the guidance information to be used and the task execution element; decode the second context feature to guide the generation of the interactive configuration information corresponding to the task execution element through the semantic association between the guidance information to be used and the task execution element, the interactive configuration information corresponding to the specified scene.

[0014] In some embodiments, the page generation device also includes a page testing unit, which is used to: call a page testing interface; test the interactive page through the page testing interface; when the interactive page test fails, modify the task execution element of the task execution model; use the modified task execution element as the task execution element of the task execution model, return to the execution step to determine the interactive configuration information corresponding to the task execution element and subsequent steps, until the interactive page test passes.

[0015] In some embodiments, when the interactive page test fails, modifying the task execution element of the task execution model includes: when the interactive page test fails, obtaining the test result; and modifying the task execution element of the task execution model based on the test result.

[0016] In some embodiments, the page generation device also includes a promotional content generation unit, which is used to: obtain a subject image in the image to be identified and attribute information of the image to be identified, wherein the subject image is a partial image in the image to be identified that is related to the subject information of the image to be identified; based on the subject image and the attribute information, generate promotional content corresponding to the interactive page, wherein the promotional content includes one or more of a promotional image and a promotional text.

[0017] An embodiment of the present application also provides an electronic device, including a processor and a memory, wherein the memory stores multiple instructions; the processor loads instructions from the memory to execute the steps in any page generation method provided in the embodiment of the present application.

[0018] The embodiment of the present application also provides a computer-readable storage medium, which stores a plurality of instructions, and the instructions are suitable for a processor to load to execute the steps in any one of the page generation methods provided in the embodiment of the present application.

[0019] The embodiment of the present application also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements the steps in any one of the page generation methods provided in the embodiment of the present application.

[0020] The embodiments of the present application can identify interactive elements of an image to be identified, and obtain element information of multiple interactive elements, wherein the interactive elements are image elements related to interactive tasks displayed in the image to be identified; determine the interactive associations between the interactive elements based on the element information of the multiple interactive elements; convert the interactive elements into task execution models of the interactive elements based on the element information of the interactive elements and the interactive associations between the interactive elements, the task execution model including at least one task execution element for implementing the interactive tasks corresponding to the interactive elements; determine the interactive configuration information corresponding to the task execution element; and generate an interactive page based on the interactive configuration information, wherein the interactive page displays the interactive components corresponding to the interactive configuration information, and the interactive components are used to implement the interactive tasks corresponding to the interactive elements in the interactive page.

[0021] In the present application, according to the interactive association of the interactive elements in the identified image, the interactive elements are converted into the task execution model of the corresponding interactive task, so as to generate an interactive page based on the interactive configuration information corresponding to the task execution model. In this process, the interactive elements are automatically converted into the task execution model of the corresponding interactive task by automatically identifying the interactive elements and determining the interactive association of the interactive elements. Then, by determining the interactive configuration information corresponding to the task execution model, the interactive page corresponding to the task flow of the interactive task is automatically generated, which reduces the need for manual intervention and greatly improves the production efficiency. Thus, an automated process is implemented to convert a static image showing image elements related to an interactive task into an interactive page that can realize dynamic interaction, which not only shortens the production time of the interactive page and improves the efficiency of producing the interactive page, but also reduces the errors caused by human operation, and ensures the accuracy and consistency of the interactive page. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1a It is a scenario diagram of the page generation method provided in the embodiment of the present application;

[0024] Figure 1b It is a flowchart of a page generation method provided in an embodiment of the present application;

[0025] Figure 1c is a schematic diagram of an image to be recognized provided in an embodiment of the present application;

[0026] Figure 1d is a schematic diagram of the identification process of the interactive element provided in the embodiment of the present application;

[0027] Figure 1e is a schematic diagram of a task execution model provided in an embodiment of the present application;

[0028] Figure 1f is a schematic diagram of interactive configuration information provided by an embodiment of the present application;

[0029] Figure 1g is a schematic diagram of a subject image and attribute information of an image to be identified provided in an embodiment of the present application;

[0030] Figure 1h is a schematic diagram of the promotion content provided by the embodiment of the present application;

[0031] Figure 2ais a schematic diagram of a page generation system provided in an embodiment of the present application;

[0032] Figure 2b is a flow chart of a page generation method provided by another embodiment of the present application;

[0033] Figure 2c is a schematic diagram of a process for generating interactive configuration information provided in an embodiment of the present application;

[0034] Figure 2d It is a flow chart of the test interactive page provided in the embodiment of the present application;

[0035] Figure 2e It is a schematic diagram of a process for generating promotional content provided by an embodiment of the present application;

[0036] Figure 3 is a structural schematic diagram of a page generating device provided in an embodiment of the present application;

[0037] Figure 4 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0039] Embodiments of the present application provide a page generation method, device, electronic device, storage medium, and program product.

[0040] The page generation device can be integrated into an electronic device, which can be a terminal, a server, etc. The terminal can be a mobile phone, a tablet computer, a smart Bluetooth device, a laptop, or a personal computer (PC); the server can be a single server or a server cluster consisting of multiple servers.

[0041] In some embodiments, the page generating device may also be integrated into multiple electronic devices. For example, the page generating device may be integrated into multiple servers, and the page generating method of the present application may be implemented by multiple servers.

[0042] In some embodiments, the terminal may also be used as a server to implement part or all of the functions of the server.

[0043] For example, refer to Figure 1aThe page generation method can be integrated in an electronic device, which can perform interactive element recognition on an image to be recognized and obtain element information of multiple interactive elements, where the interactive elements are image elements related to interactive tasks displayed in the image to be recognized; determine the interactive associations between the interactive elements based on the element information of the multiple interactive elements; based on the element information of the interactive elements and the interactive associations between the interactive elements, convert the interactive elements into task execution models of the interactive elements, where the task execution model includes at least one task execution element for implementing the interactive tasks corresponding to the interactive elements; determine the interactive configuration information corresponding to the task execution element; based on the interactive configuration information, generate an interactive page, where the interactive page displays interactive components corresponding to the interactive configuration information, where the interactive components are used to implement the interactive tasks corresponding to the interactive elements in the interactive page.

[0044] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0045] The following are detailed descriptions. It should be noted that the order of the following embodiments is not intended to limit the preferred order of the embodiments. It is understood that in the specific implementation of the present application, the data related to the images to be identified, interactive pages, promotional information, etc. of the user information, when the embodiments of the present application are applied to specific products or technologies, the user's permission or consent is required, and the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0046] In this embodiment, a page generation method is provided, such as Figure 1b As shown, the specific process of the page generation method can be as follows:

[0047] 110. Perform interactive element recognition on the image to be recognized to obtain element information of multiple interactive elements, where the interactive elements are image elements related to the interactive task displayed in the image to be recognized.

[0048] Among them, the image to be identified is an image showing interactive elements, and the image to be identified may show interactive elements of one or more interactive tasks. The interactive elements of the image to be identified can be extracted and understood (i.e., the interactive elements are identified) by an algorithm. In the embodiment of the present application, the image to be identified can be a pre-designed base map of one or more interactive tasks, and the base map can include one or more interactive elements related to each interactive task.

[0049] In the embodiment of the present application, an interactive task refers to a specific task achieved through interaction, and the specific task can be a specific goal achieved through user interaction. Figure 1c The image to be identified shown is a pre-designed base map for the operation of a certain application (i.e., product operation). The image contains multiple interactive tasks related to the operation activities, such as "log in to the application daily +1 time", "share content daily +1 time", "interact once a day +1 time" and so on in regular tasks are each an interactive task, and "add 3 new friends +3 times", "share activity content 3 times +3 times" and so on in activity tasks are each an interactive task, and "check winning records" in the image is also an interactive task.

[0050] Among them, interactive elements refer to image elements in the image to be identified that can display interactive task information. For example, interactive elements can include but are not limited to buttons, prompt texts, links, labels, graphics or other image elements that can be used to interact with users.

[0051] The element information of the interactive element refers to various attributes and / or feature information related to the interactive element. For example, the element information may include but is not limited to information in the form of text, graphics, etc. For example, the element information may include the display type of the interactive element (such as a button or a specific graphic), the description information of the interactive element displayed in the image to be identified (such as the name, prompt), the information directly identified in the image to be identified, or the information obtained by further analyzing the directly obtained information (such as feature extraction). For example, Figure 1c In the image to be identified shown, the element information of the interactive element of the interactive task "Log in to the application daily" may include text descriptions in the image "Log in to the application daily", "+1 times" and the button "Go to complete" and other related information.

[0052] In the related art, a base map of an interactive page is usually prepared first, which includes appropriate layout, colors and elements (such as buttons, icons, etc.), and then each element in the base map is sliced ​​manually, and the basic structure of the page is built based on the base map and interactive functions are added. This production method is not only inefficient, but also prone to errors. In the embodiment of the present application, an algorithm can be used to understand the interactive elements in the image to be identified, so as to automatically extract the element information of the interactive elements from the image to be identified.

[0053] In some embodiments, the task type of the interactive task corresponding to the interactive element displayed by the image to be identified and the description information of the interactive element can be determined through identification, classification and other processing. The key information of the interactive element is determined based on the image to be identified, providing an accurate and detailed judgment basis for the subsequent determination of the interactive association of the interactive element. Specifically, the element information includes the task type of the interactive task corresponding to the interactive element and the description information of the interactive element. The interactive element identification is performed on the image to be identified, and the element information of multiple interactive elements is obtained, including:

[0054] Perform interactive element detection on the image to be identified, and determine description information of multiple interactive elements displayed in the image to be identified;

[0055] Based on the description information of the interactive elements, the interactive elements are classified to obtain the task types of the interactive tasks corresponding to the interactive elements.

[0056] Among them, the task type of an interactive task refers to the category of the interactive task mode corresponding to the interactive task, that is, the category of the specific task implemented by the interactive task. The embodiment of the present application does not limit the task type, and different task types can be set according to the application scenario. For example, the task type of an interactive task may include but is not limited to one or more of a lottery, a task, a sharing, a team, a guessing game, etc. For example, the task types of "daily login + 1 time" and "daily sharing activity + 1 time" are task and sharing, respectively.

[0057] For example, Figure 1d The interactive element recognition process shown in the figure can analyze the image to be recognized by computer vision technology, locate and detect (i.e., recognize) the interactive elements (such as buttons, sliders, graphics, etc.) in the image. And extract the features of the interactive elements from the detected interactive elements through feature models such as machine learning or deep learning models (i.e., feature extraction), obtain their description information, and input the description information into the classifier to further classify the interactive elements into interactive tasks, so as to determine the type of interactive tasks corresponding to the interactive elements (i.e., obtain the interactive element recognition results).

[0058] In some embodiments, interactive elements in the image to be identified can be identified by computer vision technology such as a target detection model. The target detection model may include but is not limited to FCN (fully convolutional neural network), SegNet (deep convolutional encoder-decoder architecture for image segmentation), DeepLab (atrous convolution spatial pyramid pooling layer), Mask R-CNN (region-based convolutional neural network), U-Net (U-net-based image segmentation network), Gated SCNN (gated shape convolutional neural network for semantic segmentation), etc. One or more combinations thereof. The embodiments of the present application do not impose any restrictions on the target detection model, and different target detection models can be selected according to the application scenario.

[0059] In some embodiments, the interactive elements can be classified by a classifier to obtain the task type of the interactive task corresponding to the interactive element. The classifier can be a combination of one or more of a convolutional neural network (CNN), a recurrent neural network (RNN), a residual neural network (ResNet), a Transformer model, and an LSTM (Long Short-Term Memory Network). The embodiment of the present application does not limit the classifier, and different classifiers can be selected according to the application scenario. The classifier can be trained in advance so that it can classify the interactive elements to obtain the corresponding task type.

[0060] In some embodiments, the image to be identified may be pre-attached with attribute information such as theme information and / or style information, which may be generated by computer vision technology through image content recognition or may be manually set. Based on the theme information and / or style information of the image to be identified, the characteristics of the interactive elements may be better understood, so as to further extract features matching the theme and / or style from the information displayed by the interactive elements in the image to be identified, thereby obtaining more accurate and personalized feature information of the interactive elements. Specifically, the image to be identified is subjected to interactive element detection, and the description information of multiple interactive elements displayed in the image to be identified is determined, including:

[0061] Perform interactive element detection on the image to be identified, and extract display information of the interactive elements in the image to be identified;

[0062] Based on the attribute information of the image to be identified, feature information is extracted from the display information of the interactive element to obtain description information of the interactive element, where the attribute information includes at least one of theme information and style information of the image to be identified.

[0063] Among them, the display information of the interactive element refers to the information of the interactive element directly identified from the image to be identified, such as one or more of the display type of the interactive element (such as a button, slider or input box, etc.), the relevant information of the interactive element displayed in the image to be identified (such as name, prompt), etc.

[0064] The subject information of the image to be identified refers to the information describing the main content (i.e., theme) of the image, such as scene, object, or activity. The style information of the image to be identified refers to the information describing the form of expression or aesthetic features of the image, such as composition, artistic style, etc.

[0065] For example, when identifying interactive elements in an image to be identified, computer vision technology (such as a target detection model) can be used to extract display information of interactive elements displayed in the image to be identified, such as display type, description information of interactive elements displayed, etc. Then, the attribute information of the image to be identified and the display information of the interactive elements are spliced ​​and input into the feature extraction model. The feature extraction model can be a combination of one or more of a deep learning model (such as a multi-layer perceptron, a convolutional neural network, a Transformer network, a long short-term memory network, etc.) or a traditional machine learning model (such as a random forest, a support vector machine, etc.), and the learned weights are trained to process the input splicing features, extract higher-level abstract features, and then map the extracted abstract features to the description information of specific interactive elements through a classification layer (softmax layer), etc. The embodiment of the present application does not limit the feature extraction model, and different feature extraction models can be selected according to the application scenario. The feature extraction model can be trained in advance so that it can extract corresponding description information from the input information.

[0066] It can be understood that in the embodiment of the present application, the description information of the interactive element is information obtained by further feature extraction from the display information of the interactive element based on the attribute information of the image to be identified. In this way, during the feature extraction process, the attribute information of the image to be identified can affect the extraction result, so that the obtained description information can be the description information associated with the attribute information. For example, the theme information of the image to be identified is a game lottery, and the display type of the interactive element can be a turntable related to the game lottery, a virtual lottery box, etc. The description information of the interactive element can be information related to the game lottery based on the display information, such as "a turntable for randomly selecting game prizes, the first prize is AA, the second prize is BB, and the third prize is CC".

[0067] In some embodiments, feature information can be extracted from the display information of interactive elements through a language model (such as a conversation generation model (ChatGPT), a pre-trained language model (BERT), etc.). For example, the attribute information of the image to be identified and the display information of the interactive element are in text form. The language model can encode the text of the attribute information and display information of each interactive element, and convert the text into a high-dimensional vector. These vectors capture context-related representations, thereby generating richer and more accurate feature representations. The extracted abstract features can be predicted through an autoregressive mechanism to obtain the category and description information of the interactive task.

[0068] In some embodiments, the attribute information of the image to be identified can be extracted through a language model (such as ChatGPT or BERT, etc.), and the extracted key information can be used to extract feature information of the display information of the interactive element to obtain the task type corresponding to the interactive element and the description information of the interactive element. For example, the attribute information of the image to be identified can be input into the language model, and the language model can capture the key information in the attribute information of the image to be identified by performing semantic understanding.

[0069] 120. Determine the interactive association between the interactive elements based on the element information of the multiple interactive elements.

[0070] Among them, the interactive association of interactive elements refers to the mutual connection and influence between different interactive elements in their interactive tasks, and these interactive associations can help understand the role and function of interactive elements in the task flow of interactive tasks. In some embodiments, the interactive association of interactive elements may include but is not limited to one or more of dependency relationships, conditional relationships, and feedback relationships. Among them, dependency relationships refer to the fact that some interactive elements require the support of other interactive elements to operate effectively, such as the output of one interactive element as the input of another interactive element. Sequential relationships refer to the execution order between interactive elements, such as an interactive element must be completed before another interactive element to ensure the smooth progress of the process. Conditional relationships refer to the fact that the triggering or execution of an interactive element depends on the state or result of other interactive elements. Feedback relationships refer to the fact that there may be a feedback mechanism between interactive elements, and the result of one interactive element will affect the behavior or decision of another interactive element.

[0071] It is understandable that in practical applications, there is an interactive association between some interactive elements in the image to be identified, such as an interactive element may depend on the execution result of another interactive element. In the embodiment of the present application, the interactive association of the interactive elements can be determined based on the correlation of the element information of the interactive elements. For example, the interactive association between them can be determined by analyzing the correlation of the element information of these interactive elements. For example, the element information of interactive element A is "task type is lottery", "wheel for lottery to obtain game prizes, complete the task to obtain lottery opportunities, first prize is AA, second prize is BB, third prize is CC", and the element information of interactive element B is "task type is task", "daily login game, +1 lottery opportunity", because interactive element A and interactive element B have related information "lottery, prize, AA, BB, CC", the dependency between the description information of interactive element A and interactive element B can be analyzed by pattern matching and other methods to further determine the logical relationship (i.e., interactive association) between interactive element A and interactive element B, such as interactive element A depends on interactive element B to be completed (such as completing "daily login game") before it can be triggered.

[0072] It should be noted that, in actual applications, all interactive elements in the image to be identified may have interactive associations with other interactive elements, or only some of the interactive elements may have interactive associations. Therefore, determining the interactive associations between interactive elements may include the interactive associations between some or all interactive elements. For example, for interactive elements A and B that have interactive associations, the interactive association determined between interactive elements A and B may be "interactive element A depends on interactive element B to be completed before it can be triggered", and for interactive element C that has no interactive association with other interactive elements, the interactive association determined for interactive element C may be "empty" or "none".

[0073] In some implementations, the interactive association of the interactive elements may be determined more accurately based on the semantic association of the element information of the interactive elements. Specifically, determining the interactive association between the interactive elements based on the element information of the multiple interactive elements includes:

[0074] Determining semantic associations between element information of the interactive elements according to element information of the multiple interactive elements;

[0075] Based on the semantic association between element information, the interactive association between interactive elements is determined.

[0076] Among them, the semantic association between interactive elements refers to the semantic association of element information of the interactive elements. For example, the semantic association can be a semantic dependency relationship, such as a causal relationship, a causal dependency relationship, a temporal relationship, a conditional relationship, etc.

[0077] For example, the element information of multiple interactive elements can be spliced ​​and input into a language model (such as BERT, GPT, etc.), so that the element information of the interactive elements can be processed by the language model to output the semantic association of the interactive elements, such as the causal relationship. The language model can be trained in advance so that it can process the language model to output the semantic association of the interactive elements. And determine the interactive association of the interactive elements corresponding to the corresponding semantic association, such as the causal relationship corresponds to the dependency relationship of the interactive elements, the time relationship corresponds to the sequential relationship of the interactive elements, the conditional relationship in the semantic association corresponds to the conditional relationship of the interactive elements, and the causal dependency relationship corresponds to the feedback relationship of the interactive elements. In practical applications, a mapping relationship between the semantic association and the interactive association of the interactive elements can be established in advance, so that after determining the semantic association of the interactive elements, the corresponding interactive association can be directly determined according to the mapping relationship.

[0078] In some embodiments, in order to improve the efficiency of determining the interactive association of interactive elements, semantic features representing the semantic association of interactive elements can be extracted by encoding through a language model (such as BERT, GPT, etc.), and the interactive association of interactive elements predicted by decoding output can be obtained. The language model can be trained in advance so that it can process the interactive association of interactive elements output by the language model. In this way, the interactive association of interactive elements can be determined with only one step of model processing.

[0079] 130. Based on the element information of the interactive element and the interactive association between the interactive elements, convert the interactive element into a task execution model of the interactive element, where the task execution model includes at least one task execution element for implementing the interactive task corresponding to the interactive element.

[0080] The task execution model refers to the model used to characterize the entire process of completing an interactive task (including conditions, operations, and feedback), and the task execution elements refer to the basic components that constitute the task execution model. Figure 1e The task execution model shown in the figure can include task execution elements such as interaction conditions (Condition), operation behavior (Op, i.e. action) and reward feedback (Reward). These elements jointly influence the user's decision and action, thereby promoting the completion of the task. The task execution elements in the task execution model usually have a fixed order, such as interaction conditions → operation behavior → reward feedback.

[0081] In some ways, the conditions in the task execution model may include one or more information such as an experimental whitelist, VIP identity (i.e., VIP identity), specific time, specific behavior, and custom conditions, which can be based on the AND or NOT calculation of multiple bool expressions (i.e., Boolean expressions) as conditions for the execution of interactive tasks. The actions in the task execution model may include one or more information such as raffles, redemptions, tasks, guessing, lists, PK (i.e., battles), etc., which can build a co-constructed ecosystem based on interactive elements, improve reusability, and meet the needs of fast-changing event operations. The rewards in the task execution model may include one or more information such as gold coins, points, coupons, physical prizes, lottery opportunities, etc. (i.e., reward resources), which can build a public marketing resource library to build unified risk management and multi-dimensional abnormal data detection.

[0082] For example, in the embodiment of the present application, the task execution elements of the task execution model of each interactive element can be determined based on the element information and interactive association of the interactive element. For example, the execution condition of interactive element A depends on the execution result of interactive element B, so the execution result of interactive element B can be used as the interaction condition in the task execution model of interactive element A, and the operation behavior and reward feedback in the task execution model can be obtained from the element information of interactive element A.

[0083] In an embodiment of the present application, the interactive elements are converted into a task execution model of the interactive task based on the interactive associations of the interactive elements in the image to be identified of the identified interactive task. By automatically identifying the interactive elements and determining the interactive associations of the interactive elements, the interactive elements are automatically converted into the task execution model of the corresponding interactive task, thereby reducing the need for manual intervention and greatly improving production efficiency.

[0084] It should be noted that, for interactive elements in the image to be identified that have no interactive association with other interactive elements, they can be converted into corresponding task execution models based only on the element information of their interactive elements.

[0085] In some embodiments, for any interactive element, information related to the task execution element of the task execution model of the interactive element can be extracted from the element information of the interactive element and the element information of the interactive element that has an interactive association with the interactive element as the specific information of the corresponding task execution element. For information related to the task execution element of the task execution model of the interactive element extracted from the element information of the interactive element and the element information of the interactive element that has an interactive association with the interactive element, the extracted information can be further processed as the specific information of the corresponding task execution element, and the further processing method can include but is not limited to one or more of extracting key content from the extracted information, or supplementing or modifying the extracted information through a language model.

[0086] Since some task execution elements of an interactive element may be implicit in another interactive element, it is difficult to determine directly. Therefore, how to determine the specific task execution elements of each interactive element from the element information of many interactive elements is a difficulty. In this regard, in an embodiment of the present application, the implicit semantic associations (i.e., context features) between interactive elements are extracted through the element information and semantic associations of the interactive elements, so as to identify the semantics associated with the task execution elements of the interactive elements from the element information of many interactive elements based on the implicit semantic associations, and obtain the specific task execution elements.

[0087] Exemplarily, in some embodiments, the implicit semantic associations between interactive elements can be extracted from the element information and association information of the interactive elements through a language model, so as to predict the task execution elements of the interactive elements based on the extracted implicit semantic associations, improve the accuracy of the determined task execution elements, and improve the accuracy of converting the interactive elements into the task execution model. Specifically, based on the element information of the interactive elements and the interactive associations between the interactive elements, converting the interactive elements into the task execution models of the interactive elements includes:

[0088] Performing semantic understanding on element information of interactive elements and interactive associations between interactive elements to extract first context features of the interactive elements, where the first context features represent the semantic associations between the interactive elements;

[0089] The first context feature is decoded to obtain a task execution model of the interactive element, wherein the task execution elements in the task execution model are determined based on the semantic association between the interactive elements represented by the first context feature.

[0090] Semantic understanding is the semantic feature representation (i.e., contextual feature) of text extracted based on the contextual information of the text in the field of natural language processing. For example, one or more combinations of language models for semantic understanding such as BERT and GPT can be used to semantically understand the element information and associated information of interactive elements to obtain the contextual features (i.e., first contextual features) of the interactive elements.

[0091] For example, the element information of the interactive elements and the interactive associations between the interactive elements can be input into a language model (such as BERT, GPT, etc.) to encode the element information of the interactive elements and the interactive associations through the language model to obtain context features (i.e., first context features). Since the language model can capture the implicit semantic relationships of the information in the context when encoding the element information and the interactive associations, these semantic relationships can represent the relationships between the interactive elements, that is, they can be regarded as implicit semantic relationships between the interactive elements.

[0092] The language model can decode the first context feature to output a predicted task execution model. In the process of decoding and generating output, the language model can make full use of the information contained in the first context feature (i.e., the implicit semantic association between interactive elements) to understand the semantic relationship between the task execution element and the interactive element, so as to determine that the generated task execution element is consistent with the element information of at least one interactive element input (i.e., related), thereby extracting the information of the task execution element of another interactive element implied in the interactive element. The task execution model output by the language model includes the specific information of the corresponding task execution element, that is, the specific information of the task execution element can be obtained by the language model after encoding and decoding the element information of the interactive element and the interactive association. The language model can be trained in advance so that it can process the language model to output the task execution model. In this way, the task execution model can be determined in only one step of model processing.

[0093] In some embodiments, the element information of the interactive element is further processed and used in a language model to directly convert the interactive element into a task execution model. The further processing method may include but is not limited to extracting key content (such as keywords) from the element information. For example, keywords related to the corresponding interactive task can be extracted from the element information of the interactive element, and the keywords are used as input to the semantic model to obtain more accurate semantic features. Specifically, the element information of the interactive element and the interactive associations between the interactive elements are semantically understood to extract the first contextual features of the interactive element, including:

[0094] Extracting keywords related to the interactive tasks corresponding to the interactive elements from the element information of the interactive elements;

[0095] Semantic understanding is performed on the interactive associations between keywords and interactive elements to extract the first context features of the interactive elements.

[0096] For example, the element information of the interactive element and the task type of the interactive task corresponding to the interactive element can be spliced ​​and input into a language model (such as BERT, GPT, etc.), and the element information of the interactive element can be semantically understood by the language model, and the key points, i.e., keywords, can be output. The language model can be trained in advance so that it can process the element information of the interactive element and output the keywords of the element information. The keyword and the interactive association between the interactive elements can be input into a language model (such as BERT, GPT, etc.), and the keyword and the interactive association can be encoded by the language model to obtain the first context feature, and the predicted task execution model can be decoded and output, and the output task execution model includes the specific information of the corresponding task execution element.

[0097] In some implementations, the interactive elements may be split into sub-elements, and the sub-elements may be converted into task execution models of the interactive elements based on the element information of the interactive elements and the interactive associations between the interactive elements, further comprising:

[0098] Based on the element information of the interactive element, the interactive element is split into at least one sub-element, and the sub-element is used to convert to obtain the task execution model.

[0099] For example, an interactive element can be split into multiple sub-questions. For example, an interactive task for a lottery includes two sub-tasks: a turntable and a highlight. In this way, the interactive element of the interactive task can be split into a sub-element of the sub-task turntable and a sub-element of the sub-task highlight. And the information related to each sub-element in the element information of the interactive element is used as the element information of the sub-element. The element information of the sub-element and the interactive association of the interactive element corresponding to the sub-element can be semantically understood to extract the third context feature of the sub-element, and the third context feature represents the semantic association between the sub-elements; the third context feature is decoded to obtain a task execution model of the sub-element, and the task execution model includes at least one task execution element for implementing the interactive task corresponding to the sub-element. The task execution element in the task execution model is determined based on the semantic association between the sub-elements represented by the third context feature.

[0100] In some embodiments, the element information of the interactive element can be input into a language model (such as BERT, GPT, etc.) to split the interactive element through the language model to obtain multiple sub-elements and their element information. The language model can be trained in advance so that it can process the language model output and split it to obtain sub-elements and their element information.

[0101] 140. Determine the interactive configuration information corresponding to the task execution elements.

[0102] The interactive configuration information refers to the specific settings and parameters generated for the task execution elements (such as interactive conditions, operation behaviors, and reward feedback, etc.). For example, the interactive configuration information corresponding to the interactive conditions, the interactive configuration information corresponding to the operation behaviors, and the interactive configuration information corresponding to the reward feedback of any task execution model can constitute an interactive task corresponding to the task execution model.

[0103] For example, a task execution model may include task execution elements such as interactive conditions, operating behaviors, and reward feedback. Based on the application scenario, specific types and parameters can be defined for the task execution elements of the task execution model, and configurations can be generated based on the defined types and parameters. In the embodiment of the present application, for any interactive element, three types of interactive configuration information corresponding to the interactive conditions, operating behaviors, and reward feedback in its task execution model can be generated.

[0104] For example, by defining the condition type of the interaction condition (such as time, score, specific event) and the corresponding parameters (such as time threshold, score range), the interaction configuration information {"type":"score","threshold":1000} corresponding to the interaction condition is generated. By defining the operation type of the operation behavior (such as increase, decrease, change state) and the corresponding parameters (such as how many points to increase, which state to change to), the interaction configuration information {"operation type":"increase","quantity":50} corresponding to the operation behavior is generated. By defining the reward type of the reward feedback (such as experience points, items, gold coins) and the corresponding parameters (such as the specific content and quantity of the reward), the interaction configuration information {"type":"gold coins","quantity":10} corresponding to the reward feedback is generated.

[0105] In some embodiments, the interactive configuration information corresponding to the task execution elements can be directly predicted by the neural network model. For example, the specific information of each task execution element of the task execution model can be input into the neural network model, and the interactive configuration information corresponding to each task execution element in the task execution model can be output after encoding and decoding. The neural network model may include but is not limited to a combination of one or more of a multilayer perceptron, a convolutional neural network, a Transformer network, a long short-term memory network, GPT, BERT, etc., and the neural network model can be trained in advance so that it can process the specific information of the task execution elements and output the corresponding interactive configuration information. The embodiment of the present application does not limit the neural network model, and different neural network models can be selected according to the application scenario.

[0106] Since the model is usually predicted based on past training data, it does not have the ability to understand the latest information and knowledge of the activity field in the application scenario in real time. Especially when the model is applied in an offline system, it is impossible to access the Internet or update data, and it is impossible to obtain external knowledge by itself, which means that it can only rely on existing knowledge and information. In addition, the working memory of the model is limited and cannot process too much information at one time. Therefore, it is necessary to use other tools to enhance its functions. This makes it difficult to meet the actual application scenario by directly using the model to generate interactive components. In this regard, in an embodiment of the present application, by adding guidance information as a knowledge supplement in the actual application scenario during the process of generating interactive components, the language model can also give reasonable answers under the influence of guidance information without being trained by the latest data (i.e., the latest business knowledge), that is, generate interactive configuration information that meets the needs of the application scenario.

[0107] Exemplarily, in some implementations, determining the interaction configuration information corresponding to the task execution element includes:

[0108] Obtain at least one piece of guidance information, where the guidance information includes configuration information corresponding to a specified scenario;

[0109] Determine guidance information to be used corresponding to the task execution element from at least one guidance information;

[0110] Performing semantic understanding on the guidance information to be used and the task execution elements to extract second context features of the task execution elements, where the second context features represent the semantic association between the guidance information to be used and the task execution elements;

[0111] The second context feature is decoded to guide the generation of interaction configuration information corresponding to the task execution element through the semantic association between the guidance information to be used and the task execution element, and the interaction configuration information corresponds to the specified scene.

[0112] Among them, guidance information refers to information used to guide the generation of interactive configuration information. That is, guidance information can be used to guide (i.e. tell) the model how to generate interactive configuration information corresponding to the task execution elements. The specified scenario is the application scenario of the activity page determined according to the actual acquisition needs, such as the specific operation scenario of any product. For example, the guidance information can be integrated with information related to the specific settings and parameters of the interactive configuration in the application scenario (such as the specified scenario) (i.e. the business knowledge in the scenario), so that this information affects the model to give answers related to this information and applicable to the application scenario (i.e. the interactive configuration information). For example, in the product operation scenario, the interactive configuration information corresponding to the interactive conditions can be the settings and parameters of user guidance, functions, personalized settings, etc. for the launch of a new product (such as the launch of a new version of the application). In the game operation scenario, the interactive configuration information corresponding to the interactive conditions can be the settings and parameters of new skins, strategies, etc. when the game is launched.

[0113] For example, multiple guidance information can be pre-set according to the application scenario, and the multiple guidance information can be used to guide different task execution elements. For example, in a game scenario, condition guidance information, action guidance information, reward guidance information, game information guidance information, etc. can be pre-set, and sub-guidance information corresponding to a variety of different specific scenarios (such as specific game types, or operation scenarios) can also be set for these types of guidance information. According to the correlation (such as similarity) between each guidance information and the task execution element, the guidance information with the highest correlation with the task execution element can be determined as the guidance information to be used. The model then identifies the specific settings and parameters integrated in the guidance information to be used as the interactive configuration information corresponding to the task execution element.

[0114] Specifically, the information of the guidance information to be used and the task execution elements in the specified scene can be input into a language model (such as BERT, GPT, etc.) to encode the information input by the language model to obtain context features (i.e., second context features). Since the language model can capture the implicit semantic relationships of the information in the context when encoding the guidance information to be used and the task execution elements, these semantic relationships can characterize the relationship between the guidance information to be used and the task execution elements, that is, they can be regarded as implicit semantic associations between the guidance information to be used and the task execution elements.

[0115] The language model can decode the second context feature to output the predicted task execution model. In the process of decoding and generating output, the language model can make full use of the information contained in the second context feature (i.e., the implicit semantic association between the guidance information to be used and the task execution elements), understand the semantic relationship between the guidance information to be used and the task execution elements, so as to determine that the generated interactive configuration information is consistent with the input guidance information to be used (i.e., related), thereby extracting the information implicitly related to the task execution elements in the guidance information to be used, so that the interactive configuration information output by the language model is related to the specified scene. The language model can be trained in advance so that it can process the language model to output the interactive configuration information. In this way, the task execution model can be determined in just one step of model processing.

[0116] For example, Figure 1c For the image to be identified as shown in the figure, the multiple interactive elements in the image can generate interactive configuration information related to the product operation scenario. For example, for the multiple interactive elements displayed in the image to be identified, the corresponding Figure 1f The interactive configuration information shown includes interactive configuration information corresponding to task execution elements (such as reward feedback) corresponding to multiple interactive elements, such as receiving gift packages (task type), sharing, links, gifts and lucky draws, etc. In actual applications, these interactive configuration information may include specific settings and parameters.

[0117] In some implementations, the guidance information may be a plug-in for providing expansion capabilities for the model, so as to serve as an independent module attached to the model to enhance the model's ability to generate interactive configuration information.

[0118] In some embodiments, the guidance information to be used corresponding to the task execution element can be determined from at least one guidance information through a language model. For example, the specific information of the task execution element and the at least one guidance information obtained can be input into a language model (such as BERT, GPT, etc.) to encode and decode the specific information of the task execution element and the guidance information through the language model to output the predicted guidance information to be used with the highest correlation with the task execution element, and the output task execution model includes the specific information of the corresponding task execution element. The language model can be trained in advance so that it can process and output the guidance information to be used with the highest correlation with the task execution element.

[0119] In some embodiments, the interactive configuration information corresponding to the task execution element can be generated by a language model. For example, the specific information of any task execution element and the corresponding guidance information to be used can be input into a language model (such as BERT, GPT, etc.), so that the specific information of the task execution element and the guidance information to be used can be encoded and decoded by the language model to output the predicted interactive configuration information corresponding to the task execution element. The language model can be trained in advance so that it can process and output the interactive configuration information corresponding to the task execution element.

[0120] 150. Generate an interactive page based on the interactive configuration information. The interactive page displays interactive components. The interactive components are used to implement interactive tasks corresponding to the interactive elements in the interactive page.

[0121] Among them, the interactive component refers to the component used for user interaction in the interactive page to achieve interactive tasks, such as but not limited to one or more of user interface elements (such as buttons, menus, input boxes, multimedia elements), feedback mechanisms (such as prompts, notifications), and functional modules that support users to complete specific tasks. The embodiment of the present application does not limit the form of the interactive component, and different forms of interactive components can be set according to the application scenario.

[0122] Among them, an interactive page refers to a page used for user interaction to achieve interactive tasks, such as an H5 page, a mobile application page, or a web application page, etc. In the embodiment of the present application, an interactive page may include one or more pages. The embodiment of the present application does not limit the form of the interactive page, and different forms of interactive pages can be set according to the application scenario.

[0123] For example, by parsing the interactive configuration information of interactive conditions, operation behaviors, and reward feedback, the front-end framework can obtain the corresponding UI elements and generate corresponding interactive components, such as score display, buttons, animations of operations and rewards, or other visual effects, etc. These interactive components are displayed in one or more pages to form an interactive page. By determining the interactive configuration information corresponding to the task execution model, the interactive page corresponding to the task flow of the interactive task is automatically generated. In this way, an automated process is implemented to convert a static image that displays image elements related to the interactive task into an interactive page that can realize dynamic interaction, which not only shortens the production time of the interactive page and improves the efficiency of producing the interactive page, but also reduces the errors caused by human operation, ensuring the accuracy and consistency of the interactive page.

[0124] In some implementations, the interactive page can be tested to continuously modify the task execution elements according to the test results until a reliable interactive page is generated, so as to reduce errors after the interactive page is launched and improve the user experience. Among them, the task execution elements are modified each time the test fails, so as to adjust the specific details without changing the overall interactive logic of the interactive page, reduce new errors that may be caused by major changes or reconstruction, and make the interactive process implemented by the interactive page always consistent, without affecting the implementation of the interactive task. Specifically, based on the interactive configuration information, after the interactive page is generated, it also includes:

[0125] Call the page test interface;

[0126] Test the interactive page through the page test interface;

[0127] When the interactive page test fails, modify the task execution elements of the task execution model;

[0128] The modified task execution element is used as the task execution element of the task execution model, and the execution step is returned to determine the interactive configuration information and subsequent steps corresponding to the task execution element until the interactive page test passes.

[0129] The page test interface refers to the interface used to call the page test function.

[0130] For example, the page test function can be integrated in the background, and the process of generating an interactive page based on the interactive configuration information can be integrated in the front end, so that after the interactive page is generated, the front end can call the page test interface of the background, such as the mock interface, to call the page test function of the background after the interactive page is generated by the front end to perform automated testing on the interactive page. When the test passes, the interactive page can be delivered according to the delivery plan. When the test fails, the specific information of one or more task execution elements in the task execution model can be adjusted to generate new interactive configuration information based on the specific information of the adjusted task execution elements.

[0131] In some implementations, the specific information of the task execution elements may be reorganized, replaced, added, deleted, or modified to obtain the modified task execution elements. For example, the operation instruction information of the modification operation and the specific information of the task elements may be input into a language model (such as BERT, GPT, etc.) so that the language model performs the above modification operation on the specific information of the task elements to obtain the modified task execution elements.

[0132] In some implementations, the task execution elements can be modified based on the test results to improve the accuracy of the modification process. Specifically, when the interactive page test fails, the task execution elements of the task execution model are modified, including:

[0133] When the interactive page test fails, get the test result;

[0134] Based on the test results, modify the task execution elements of the task execution model.

[0135] For example, the test results can be fed back to the language model so that the language model modifies the task execution elements based on the fed-back test results. For example, the test results, the element information of the interactive elements, and the interactive associations between the interactive elements can be spliced ​​and input into a language model (such as BERT, GPT, etc.), so that the language model can encode the input information to obtain semantic features and decode the output predicted task execution model (i.e., the modified task execution model), and the modified output task execution model includes the specific information of the corresponding modified task execution elements.

[0136] In some embodiments, the ReAct framework (reactive framework) can be used to implement the process of converting interactive elements into task execution models of interactive elements, determining interactive configuration information corresponding to task execution elements, generating interactive pages based on interactive configuration information, and testing interactive pages. The ReAct framework is Reason + Action (Reason + Act), and its principles include autonomous understanding, planning, providing large model expansion capabilities, executing complex tasks, etc. The ReAct framework can obtain the final result through N cycles of reasoning through questions (i.e., multiple interactive elements) - thinking (i.e., determining interactive associations, and converting to obtain task execution models) - execution expansion capabilities (i.e., guided by guidance information to be used) - observation results (i.e., testing interactive pages, and feeding back test results to the model).

[0137] In some implementations, accurate and personalized promotional content related to the interactive page can be generated based on the subject image and attribute information of the image to be identified, so as to attract users' attention to the interactive page through the promotional content, increase the click-through rate and conversion rate of the interactive page, and thus improve the marketing effect. Specifically, the method also includes:

[0138] Acquire a subject image in the image to be identified and attribute information of the image to be identified, wherein the subject image is a partial image in the image to be identified that is related to the subject information of the image to be identified;

[0139] Based on the theme image and attribute information, promotional content corresponding to the interactive page is generated, and the promotional content includes one or more of a promotional image and a promotional text.

[0140] The theme image refers to a portion of the image to be identified that displays the main content (i.e., theme) of the image. For example, the image to be identified may be composed of multiple sub-images of the interactive task, and the theme image may be the cover of the multiple sub-images.

[0141] Among them, promotional content refers to the content used to promote the interactive page, and the promotional content includes one or more forms of content in images and texts. The promotional image can be an image such as a poster, a cover image, an infographic, etc. used to convey the theme of the interactive page, and the promotional text can be a text such as a slogan, a promotional copy, etc. used to convey the theme of the interactive page. In the embodiment of the present application, the interactive page and the promotional content can be linked to each other, such as embedding the interactive page into the promotional content, or the interactive page and the promotional content can be separately delivered, such as on different channels or platforms.

[0142] For example, the key information in the subject image can be identified and extracted through computer vision technology (image recognition model), and the extracted key information and the attribute information of the image to be identified are input into a Vincent text model (such as BERT, GPT, etc.) or a Vincent graph model (such as the Vincent graph function of GPT, Stable Diffusion (diffusion model), DALL-E, etc.) to generate promotional text or promotional images respectively. The Vincent text model or Vincent graph model can be trained in advance so that it can process the key information in the subject image and the attribute information of the image to be identified (theme information and / or style information) to output a promotional image or promotional text. The embodiments of the present application do not limit the Vincent text model and the Vincent graph model, and different Vincent text models and Vincent graph models can be selected according to the application scenario.

[0143] For example, the embodiments of the present application can be based on Figure 1g The subject image and attribute information of the image to be identified are generated as shown in Figure 1h The promotional content shown. Among them, the theme image of the image to be identified is the cover of the image to be identified, and the attribute information includes the title and style of the image to be identified. The promotional text in the promotional content can be used as the operation copy content, and the promotional image can be used as the promotion cover.

[0144] In some implementations, the attribute information includes style information, so that promotional content corresponding to the style information can be generated.

[0145] In some embodiments, the key information extracted from the extracted subject image and the attribute information of the image to be identified can be input into the model through a language model (such as BERT, GPT, etc.) to generate a model of a Vincent text model or a Vincent image model Prompt to generate a promotional text or promotional image. The language model can be trained in advance so that it can output a Vincent text model or a Vincent image model Prompt of the generation model. The embodiment of the present application does not limit the language model, and different language models can be selected according to the application scenario.

[0146] In some embodiments, the promotional content is associated with the interactive page, such as the promotional content can be used to access the interactive page. For example, the promotional content and the interactive page can be placed in different channels, different platforms, or different locations on the same platform. When a user views the promotional content, the interactive page can be redirected or the user can be guided to access the interactive page.

[0147] The page generation scheme provided by the embodiment of the present application can be applied in various interactive page generation scenarios. For example, taking various types of operation activities as an example, interactive elements are identified for the image to be identified, and element information of multiple interactive elements is obtained, where the interactive elements are image elements related to the interactive tasks displayed in the image to be identified; based on the element information of multiple interactive elements, the interactive associations between the interactive elements are determined; based on the element information of the interactive elements and the interactive associations between the interactive elements, the interactive elements are converted into task execution models of the interactive elements, and the task execution model includes at least one task execution element for implementing the interactive tasks corresponding to the interactive elements; the interactive configuration information corresponding to the task execution element is determined; based on the interactive configuration information, an interactive page is generated, and the interactive page displays the interactive components corresponding to the interactive configuration information, and the interactive components are used to implement the interactive tasks corresponding to the interactive elements in the interactive page.

[0148] As can be seen from the above, the embodiments of the present application can convert the interactive elements into the task execution model of the corresponding interactive task according to the interactive association of the interactive elements in the identified image, so as to generate an interactive page based on the interactive configuration information corresponding to the task execution model. In this process, the interactive elements are automatically converted into the task execution model of the corresponding interactive task by automatically identifying the interactive elements and determining the interactive association of the interactive elements. Then, by determining the interactive configuration information corresponding to the task execution model, the interactive page corresponding to the task flow of the interactive task is automatically generated, which reduces the need for manual intervention and greatly improves the production efficiency. Thus, an automated process of converting a static image showing image elements related to an interactive task into an interactive page that can realize dynamic interaction is realized, which not only shortens the production time of the interactive page and improves the efficiency of producing the interactive page, but also reduces the errors caused by human operation, and ensures the accuracy and consistency of the interactive page.

[0149] The method described in the above embodiment will be further described in detail below.

[0150] In this embodiment, the method of the embodiment of the present application will be described in detail by taking the application in a game operation scenario as an example.

[0151] The embodiments of the present application can be applied to a page generation system, such as Figure 2a The page generation system shown in the figure includes an interactive element recognition module, an interactive page generation module and a promotional content generation module. The interactive element recognition module is used for preprocessing processes, such as extracting element information of interactive elements, the interactive page generation module is used for understanding, deducing, merging, and generating content of element information, such as generating interactive pages based on the extracted element information, and the promotional content generation module is used for understanding, deducing, merging, and generating content of input images, themes and styles of element information, such as generating promotional content (including promotional text and images).

[0152] like Figure 2b As shown, the specific process of a page generation method is as follows:

[0153] 210. Perform interactive element recognition on the image to be recognized to obtain element information of multiple interactive elements, where the interactive elements are image elements related to the interactive task displayed in the image to be recognized.

[0154] For example, Figure 2a The page generation system shown can input the operation content for the game operation scene into the page generation system. The operation content includes the activity page picture (i.e., the image to be identified) produced by any activity and the operation copy (such as activity theme, style, etc. keywords) and cover picture of the activity page. The activity page picture may include one or more interactive elements of interactive tasks. The interactive element recognition module can be used to perform natural language processing (NLP for short) on the activity page picture and the keywords and themes of the activity page to understand the feature extraction style information. And locate and detect (i.e., identify) the interactive elements in the image (such as buttons, sliders, graphics, etc.) through computer vision technology, and extract the features of the interactive elements (such as the task type corresponding to the interactive element, the description content corresponding to the interactive element) based on the corresponding keyword characteristics (i.e., the extracted style information). Then, the interactive elements are classified based on their features by a classifier to determine the task type of the interactive task corresponding to the interactive element (such as lottery, task, sharing, team, guessing, etc.).

[0155] For example, the specific process may include: (1) the operator initiates a request; (2) the image of the activity page is loaded into the component extraction system; (3) the image interactive element detection; (4) the image interactive element preprocessing; (5) the characteristics of the interactive element, and obtain the description information of the interactive element; (6) the classifier is trained to classify and obtain the task type of the corresponding interactive task; (7) the result is returned to the system.

[0156] 220. Determine the interactive association between the interactive elements based on the element information of the multiple interactive elements.

[0157] In an embodiment of the present application, the activity page image may include a variety of gameplays (i.e., interactive tasks), such as lottery, tasks, sharing, team formation, guessing and other different gameplays, and the rewards for each gameplay are different, and the number of times you can participate is also different. And there is a correlation between gameplays, and the combination between gameplays is ever-changing. For example, a user shares an activity, and after sharing it many times, he completes the task to get points, and the points can be used to draw a lottery. Therefore, the key issue in this application is how to construct and identify the relationship between each interactive element, how to generate an activity page and how to criticize the correctness of the generated activity page when GPT does not understand the business-defined operational activities.

[0158] In response to the above-mentioned problem of constructing the relationship between interactive elements, the interactive page generation module understands each interactive element and the description of the interactive element through chatgpt, extracts the relationship between the interactive elements and the interactive element dependency (i.e., interactive association), and the main keyword of the reward issued (the keyword is extracted based on the description information of the interactive element) as a feature. Among them, the relationship and dependency between interactive elements refers to the relationship between multiple interactive elements that realize the same gameplay (i.e., the association relationship triggered by the interactive elements, etc., the association relationship between operations and reality), such as completing the task first to obtain the chance to draw a lottery, that is, there is an association relationship between the task and the lottery.

[0159] For example, the interactive page generation module can determine the interactive association between interactive elements based on the semantics of the element information of the interactive elements. For example, the keywords can be determined first, and then the relationship between interactive elements can be extracted based on the keywords. For example, if the element information records "Complete the AA task, increase the chance of lottery three times", and "Lottery depends on AA, the number of lottery draws", these words can be used to clarify the relationship and prove that there is an association relationship (i.e., interactive association) between the interactive elements.

[0160] 230. Based on the element information of the interactive element and the interactive association between the interactive elements, convert the interactive element into a task execution model of the interactive element, where the task execution model includes at least one task execution element for implementing the interactive task corresponding to the interactive element.

[0161] After clarifying the relationship between the interactive elements, the configuration corresponding to the interactive elements can be generated. In the embodiment of the present application, in order to generate interactive elements faster and more accurately, such as Figure 2a The page generation system shown can abstract the interactive elements of the gameplay (i.e., interactive tasks) through the interactive page generation module, and abstract them into a condition+op+reward model (i.e., task execution model) corresponding to each interactive element. In all gameplays, configuration information is generated for the three task execution elements in the task execution model respectively.

[0162] 240. Determine the interactive configuration information corresponding to the task execution element.

[0163] 250. Generate an interactive page based on the interactive configuration information. The interactive page displays interactive components corresponding to the interactive configuration information. The interactive components are used to implement interactive tasks corresponding to the interactive elements in the interactive page.

[0164] 260. Test the interactive page.

[0165] In the embodiment of the present application, in order to avoid the disadvantages of large models, Figure 2a In the page generation system shown, the interactive page generation module uses the Prompt optimization mode ReAct framework (reactive framework) to generate interactive configuration information, wherein the display of the React prompt can be realized by a custom prompt dialog box component. The ReAct framework is Reasoning + Action (Reason + Act), and its principle is: autonomous understanding, planning, providing large model expansion capabilities, and executing complex tasks, and the final result is obtained by looping N times of reasoning through questions-thinking-executing expansion capabilities-observing results. Three large model plug-ins (i.e., guide information) are provided here to provide expansion capabilities: conditional plug-ins, action plug-ins, reward plug-ins, and game information plug-ins. In the embodiment of the present application, three large model expansion plug-ins are provided: gameplay template components (i.e., conditional plug-ins, action plug-ins, and reward plug-ins), which mainly generate corresponding configurations by identifying the types of interactive elements, rewards, flow control and other information. These plug-ins can be subdivided into a variety of different plug-ins; game information plug-ins, which generate configurations related to game information; automated testing plug-ins, which are used for automated testing.

[0166] like Figure 2cThe specific execution process of the ReAct framework is as follows: 1. Input the relationship between interactive elements and the element information of interactive elements (such as description information) into the big model (i.e., language model 1); 2. The big model (i.e., language model 1) splits the interactive elements of an activity into n sub-problems, i.e., sub-elements of n sub-game modes, and builds multiple task execution models. One task execution model corresponds to one interactive task, and the task execution elements in the task execution model correspond to the sub-elements; 3. Tell GPT (i.e., language model 2) which extension capability plug-ins based on business knowledge (such as game Game information plug-in, condition plug-in, action plug-in, reward plug-in), language model 2 combines the sub-problem to select an available plug-in (such as combining sub-problem A to select the corresponding plug-in); 4. The interactive page generation module calls the plug-in of the selected large model, and language model 2 generates the corresponding configuration (such as calling the extension plug-in to get the result of A); 5. Language model 2 calls the automated testing plug-in to perform automated testing to determine whether the interactive page corresponding to the generated configuration (such as an H5 page) meets the delivery requirements. If the test passes, the process ends and the final interactive page for delivery is obtained. If it fails, repeat steps 2 to 4.

[0167] Among them, Figure 2d In the process of testing the interactive page shown, step 260 may include the following steps 261 to 264:

[0168] 261. Call the page test interface;

[0169] 262. Test the interactive page through the page test interface;

[0170] 263. When the interactive page test fails, modify the task execution elements of the task execution model;

[0171] 264. Use the modified task execution element as the task execution element of the task execution model, return to the execution step to determine the interactive configuration information and subsequent steps corresponding to the task execution element, until the interactive page test passes.

[0172] For example, the interactive page generation module introduces an automated testing plug-in into the ReAct framework. Each time an activity page is generated, it is automatically tested to determine whether it has been generated correctly. The large model determines whether it meets the delivery requirements. If the test passes, the process ends. If it fails, the test result is used as input and fed back to the large model (i.e., language model 1) to correct the configuration.

[0173] 270. Generate promotional content corresponding to the interactive page.

[0174] like Figure 2e In the process of generating promotional content shown, step 270 may include the following steps 271-272:

[0175] 271. Obtain a subject image in the image to be identified and attribute information of the image to be identified, where the subject image is a partial image in the image to be identified that is related to the subject information of the image to be identified.

[0176] 272. Generate promotional content corresponding to the interactive page based on the theme image and attribute information, where the promotional content includes one or more of a promotional image and a promotional text.

[0177] For example, after the event h5 page, you can efficiently generate promotional content such as a cover that matches the event page and eye-catching operation copy based on the information on the event page. Specifically, you can generate promotional content through the following process: 1. Operation input reference pictures, titles, and styles; 2. Extract operation content tags (i.e. titles and styles), as well as elements of uploaded pictures; 3. Figure 2a The page generation system shown in the figure uses the style extracted from the activity page image and the information obtained by aggregating the title as the prompt of the text-generated-text model, so as to generate the corresponding copy (i.e., promotional text) by utilizing the text-generated-image capability of GPT; 4. The information obtained by aggregating the copy and image tags is used as the prompt of the text-generated-image model (such as GPT), and the corresponding cover image (i.e., promotional image) is generated by utilizing the image-generated-image capability of GPT; 5. The copy and image are uploaded to the server for delivery.

[0178] As can be seen from the above, the embodiment of the present application understands the content of the activity page image based on the information such as the activity picture and operation copy input by the operation. Based on the results of content understanding, combined with the plug-in of the language model and model extension, such as the plug-in of game information and game props, the H5 page of the operation activity is generated. Finally, the generated page H5 is automatically tested, and after the test is completed, the generated page is animated. Thus, the embodiment of the present application provides an interactive content processing pipeline based on intelligent content understanding, which realizes comprehensive processing links such as preprocessing, understanding, reasoning, processing and content generation. Thus, the operation only needs to input the operation copy and the activity picture, that is, intelligent identification and automatic generation of the H5 page and promotion content of the activity page delivery. Among them, through the large model and the plug-in of the model extension, the activity page with complex dependencies and component types is understood, which greatly improves the efficiency and configuration error rate of the configuration. Through the technical means such as Wenshengtu and Wenshengwen, different types of personalized covers and contents are generated according to the user population based on style information, so as to improve the delivery efficiency and conversion rate of the activity.

[0179] In order to better implement the above method, the embodiment of the present application also provides a page generation device, which can be integrated in an electronic device, and the electronic device can be a terminal, a server, etc. Among them, the terminal can be a mobile phone, a tablet computer, a smart Bluetooth device, a laptop, a personal computer, etc.; the server can be a single server or a server cluster composed of multiple servers.

[0180] For example, in this embodiment, the method of the embodiment of the present application will be described in detail by taking the page generation device being specifically integrated into an electronic device as an example.

[0181] For example, Figure 3 As shown, the page generation device may include an element identification unit 310, a relationship determination unit 320, an element conversion unit 330, a configuration determination unit 340, and a page generation unit 350, as follows:

[0182] (I) Element Identification Unit

[0183] It is used to identify interactive elements in the image to be identified and obtain element information of multiple interactive elements. The interactive elements are image elements related to the interactive tasks displayed in the image to be identified.

[0184] In some embodiments, the element information includes the task type of the interactive task corresponding to the interactive element and the description information of the interactive element. The element identification unit is used to: perform interactive element detection on the image to be identified, and determine the description information of multiple interactive elements displayed in the image to be identified; classify the interactive elements based on the description information of the interactive elements, and obtain the task type of the interactive task corresponding to the interactive element.

[0185] In some embodiments, interactive element detection is performed on the image to be identified to determine the description information of multiple interactive elements displayed in the image to be identified, including: performing interactive element detection on the image to be identified to extract the display information of the interactive elements in the image to be identified; based on the attribute information of the image to be identified, feature information is extracted from the display information of the interactive elements to obtain the description information of the interactive elements, the attribute information includes at least one of the theme information and style information of the image to be identified.

[0186] 2. Relationship determination unit

[0187] Used to determine the interaction relationship between interactive elements based on element information of multiple interactive elements.

[0188] In some embodiments, the relationship determination unit is specifically used to: determine the semantic association between element information of the interactive elements based on element information of multiple interactive elements; and determine the interactive association between the interactive elements based on the semantic association between the element information.

[0189] (III) Element conversion unit

[0190] The method is used to convert an interactive element into a task execution model of the interactive element based on element information of the interactive element and the interactive association between the interactive elements. The task execution model includes at least one task execution element for implementing an interactive task corresponding to the interactive element.

[0191] In some embodiments, the element conversion unit is specifically used to: perform semantic understanding of element information of interactive elements and interactive associations between interactive elements to extract first context features of the interactive elements, the first context features representing the semantic associations between the interactive elements; decode the first context features to obtain a task execution model of the interactive elements, the task execution elements in the task execution model being determined based on the semantic associations between the interactive elements represented by the first context features.

[0192] In some embodiments, semantic understanding is performed on element information of interactive elements and interactive associations between interactive elements to extract first context features of the interactive elements, including: extracting keywords related to interactive tasks corresponding to the interactive elements from the element information of the interactive elements; and semantic understanding is performed on the keywords and interactive associations between interactive elements to extract first context features of the interactive elements.

[0193] In some implementations, the element conversion unit is further used to: split the interactive element into at least one sub-element based on the element information of the interactive element, and the sub-element is used to convert to obtain the task execution model.

[0194] (IV) Configuration determination unit

[0195] Used to determine the interactive configuration information corresponding to the task execution elements.

[0196] In some embodiments, the configuration determination unit is specifically used to: obtain at least one guidance information, the guidance information including configuration information corresponding to a specified scenario; determine the guidance information to be used corresponding to the task execution element from at least one guidance information based on the element information of the interactive element corresponding to the task execution element; perform semantic understanding on the guidance information to be used and the task execution element to extract a second context feature of the task execution element, the second context feature representing the semantic association between the guidance information to be used and the task execution element; decode the second context feature to guide the generation of interactive configuration information corresponding to the task execution element through the semantic association between the guidance information to be used and the task execution element, the interactive configuration information corresponding to the specified scenario.

[0197] (V) Page Generation Unit

[0198] It is used to generate an interactive page based on the interactive configuration information. The interactive page displays the interactive components corresponding to the interactive configuration information. The interactive components are used to implement the interactive tasks corresponding to the interactive elements in the interactive page.

[0199] In some embodiments, the page generation device also includes a page testing unit, which is used to: call a page testing interface; test the interactive page through the page testing interface; when the interactive page test fails, modify the task execution element of the task execution model; use the modified task execution element as the task execution element of the task execution model, return to the execution step to determine the interactive configuration information and subsequent steps corresponding to the task execution element, until the interactive page test passes.

[0200] In some embodiments, when the interactive page test fails, the task execution elements of the task execution model are modified, including: when the interactive page test fails, obtaining the test results; based on the test results, modifying the task execution elements of the task execution model.

[0201] In some embodiments, the page generation device also includes a promotional content generation unit, which is used to: obtain a subject image in the image to be identified and attribute information of the image to be identified, the subject image being a partial image in the image to be identified that is related to the subject information of the image to be identified; based on the subject image and the attribute information, generate promotional content corresponding to the interactive page, the promotional content including one or more of a promotional image and a promotional text.

[0202] In specific implementation, the above units can be implemented as independent entities, or can be arbitrarily combined to be implemented as the same or several entities. The specific implementation of the above units can refer to the previous method embodiments, which will not be repeated here.

[0203] As can be seen from the above, the page generation device of this embodiment includes an element recognition unit, a relationship determination unit, an element conversion unit, a configuration determination unit and a page generation unit. Among them, the element recognition unit is used to perform interactive element recognition on the image to be recognized, and obtain element information of multiple interactive elements, where the interactive elements are image elements related to interactive tasks displayed in the image to be recognized; the relationship determination unit is used to determine the interactive associations between the interactive elements based on the element information of the multiple interactive elements; the element conversion unit is used to convert the interactive elements into task execution models of the interactive elements based on the element information of the interactive elements and the interactive associations between the interactive elements, and the task execution model includes at least one task execution element for implementing the interactive tasks corresponding to the interactive elements; the configuration determination unit is used to determine the interactive configuration information corresponding to the task execution element; the page generation unit is used to generate an interactive page based on the interactive configuration information, and the interactive page displays the interactive components corresponding to the interactive configuration information, and the interactive components are used to implement the interactive tasks corresponding to the interactive elements in the interactive page.

[0204] Thus, the embodiment of the present application can convert the interactive elements into the task execution model of the corresponding interactive task according to the interactive association of the interactive elements in the identified image, so as to generate an interactive page based on the interactive configuration information corresponding to the task execution model. In this process, the interactive elements are automatically converted into the task execution model of the corresponding interactive task by automatically identifying the interactive elements and determining the interactive association of the interactive elements. Then, by determining the interactive configuration information corresponding to the task execution model, the interactive page corresponding to the task flow of the interactive task is automatically generated, which reduces the need for manual intervention and greatly improves the production efficiency. Thus, an automated process for converting a static image showing image elements related to an interactive task into an interactive page that can realize dynamic interaction is realized, which not only shortens the production time of the interactive page and improves the efficiency of producing the interactive page, but also reduces the errors caused by human operation, and ensures the accuracy and consistency of the interactive page.

[0205] The embodiment of the present application also provides an electronic device, which can be a terminal, a server, etc. The terminal can be a mobile phone, a tablet computer, a smart Bluetooth device, a laptop, a personal computer, etc. The server can be a single server or a server cluster composed of multiple servers, etc.

[0206] In some embodiments, the page generating device may also be integrated into multiple electronic devices. For example, the page generating device may be integrated into multiple servers, and the page generating method of the present application may be implemented by multiple servers.

[0207] In this embodiment, the electronic device of this embodiment is a server as an example for detailed description, for example, Figure 4As shown, it shows a schematic diagram of the structure of the server involved in the embodiment of the present application, specifically:

[0208] The server may include one or more processing core processors 410, one or more computer-readable storage media memories 420, a power supply 430, an input module 440, and a communication module 450. Those skilled in the art will appreciate that Figure 4 The server structure shown in the figure does not constitute a limitation on the server, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. Among them:

[0209] The processor 410 is the control center of the server, and uses various interfaces and lines to connect various parts of the entire server. It executes various functions of the server and processes data by running or executing software programs and / or modules stored in the memory 420, and calling data stored in the memory 420. In some embodiments, the processor 410 may include one or more processing cores; in some embodiments, the processor 410 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 410.

[0210] The memory 420 can be used to store software programs and modules. The processor 410 executes various functional applications and data processing by running the software programs and modules stored in the memory 420. The memory 420 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the server, etc. In addition, the memory 420 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory 420 may also include a memory controller to provide the processor 410 with access to the memory 420.

[0211] The server also includes a power supply 430 for supplying power to various components. In some embodiments, the power supply 430 can be logically connected to the processor 410 through a power management system, so that the power management system can manage charging, discharging, and power consumption. The power supply 430 can also include any components such as one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, and power status indicators.

[0212] The server may further include an input module 440, which may be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function controls.

[0213] The server may further include a communication module 450. In some embodiments, the communication module 450 may include a wireless module. The server may perform short-range wireless transmission through the wireless module of the communication module 450, thereby providing wireless broadband Internet access to the user. For example, the communication module 450 may be used to help the user send and receive emails, browse web pages, and access streaming media.

[0214] Although not shown, the server may further include a display unit, etc., which will not be described in detail herein. Specifically in this embodiment, the processor 410 in the server will load the executable files corresponding to the processes of one or more application programs into the memory 420 according to the following instructions, and the processor 410 will run the application programs stored in the memory 420, thereby realizing various functions, as follows:

[0215] Interactive elements are identified on the image to be identified to obtain element information of multiple interactive elements, where the interactive elements are image elements related to the interactive tasks displayed in the image to be identified; based on the element information of the multiple interactive elements, the interactive associations between the interactive elements are determined; based on the element information of the interactive elements and the interactive associations between the interactive elements, the interactive elements are converted into task execution models of the interactive elements, where the task execution model includes at least one task execution element for implementing the interactive tasks corresponding to the interactive elements; interactive configuration information corresponding to the task execution elements is determined; based on the interactive configuration information, an interactive page is generated, where the interactive page displays interactive components corresponding to the interactive configuration information, where the interactive components are used to implement the interactive tasks corresponding to the interactive elements in the interactive page.

[0216] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0217] As can be seen from the above, the embodiment of the present application can convert the interactive elements into the task execution model of the corresponding interactive task according to the interactive association of the interactive elements in the identified image, so as to generate an interactive page based on the interactive configuration information corresponding to the task execution model. In this process, the interactive elements are automatically converted into the task execution model of the corresponding interactive task by automatically identifying the interactive elements and determining the interactive association of the interactive elements. Then, by determining the interactive configuration information corresponding to the task execution model, the interactive page corresponding to the task flow of the interactive task is automatically generated, which reduces the need for manual intervention and greatly improves the production efficiency. Thus, an automated process is implemented to convert a static image showing image elements related to an interactive task into an interactive page that can realize dynamic interaction, which not only shortens the production time of the interactive page and improves the efficiency of producing the interactive page, but also reduces the errors caused by human operation, and ensures the accuracy and consistency of the interactive page.

[0218] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0219] To this end, an embodiment of the present application provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions can be loaded by a processor to execute the steps in any of the page generation methods provided in the embodiments of the present application. For example, the instructions can execute the following steps:

[0220] Interactive elements are identified on the image to be identified to obtain element information of multiple interactive elements, where the interactive elements are image elements related to the interactive tasks displayed in the image to be identified; based on the element information of the multiple interactive elements, the interactive associations between the interactive elements are determined; based on the element information of the interactive elements and the interactive associations between the interactive elements, the interactive elements are converted into task execution models of the interactive elements, where the task execution model includes at least one task execution element for implementing the interactive tasks corresponding to the interactive elements; interactive configuration information corresponding to the task execution elements is determined; based on the interactive configuration information, an interactive page is generated, where the interactive page displays interactive components corresponding to the interactive configuration information, where the interactive components are used to implement the interactive tasks corresponding to the interactive elements in the interactive page.

[0221] The storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0222] According to one aspect of the present application, a computer program product or computer program is provided, including a computer program or instruction, which, when executed by a processor, implements the steps in the method provided in various optional implementations provided in the above embodiments. The computer program / instruction is stored in a computer-readable storage medium. A processor of an electronic device reads the computer program / instruction from the computer-readable storage medium, and the processor executes the computer program / instruction, so that the electronic device performs the method provided in various optional implementations provided in the above embodiments.

[0223] Since the instructions stored in the storage medium can execute the steps in any one of the page generation methods provided in the embodiments of the present application, the beneficial effects that can be achieved by any one of the page generation methods provided in the embodiments of the present application can be achieved. For details, please refer to the previous embodiments and will not be repeated here.

[0224] The above is a detailed introduction to a page generation method, device, electronic device, storage medium and program product provided in the embodiments of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for technical personnel in this field, according to the ideas of the present application, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A page generation method, characterized in that: include: Perform interactive element recognition on the image to be recognized to obtain element information of multiple interactive elements, wherein the interactive elements are image elements related to the interactive task displayed in the image to be recognized; Determining, according to the element information of the plurality of interactive elements, interactive associations between the interactive elements; Based on the element information of the interactive element and the interactive association between the interactive elements, convert the interactive element into a task execution model of the interactive element, wherein the task execution model includes at least one task execution element for implementing the interactive task corresponding to the interactive element; Determining interactive configuration information corresponding to the task execution element; Based on the interaction configuration information, an interaction page is generated, and the interaction page displays the interaction components corresponding to the interaction configuration information. The interaction components are used to implement the interaction tasks corresponding to the interactive elements in the interaction page.

2. The page generation method according to claim 1, characterized in that: The element information includes the task type of the interactive task corresponding to the interactive element and the description information of the interactive element. The interactive element recognition of the image to be recognized is performed to obtain the element information of the plurality of interactive elements, including: Performing interactive element detection on the image to be identified, and determining description information of a plurality of the interactive elements displayed in the image to be identified; Based on the description information of the interactive element, the interactive element is classified to obtain the task type of the interactive task corresponding to the interactive element.

3. The page generation method according to claim 2, characterized in that: The detecting interactive elements of the image to be identified and determining description information of a plurality of the interactive elements displayed in the image to be identified includes: Performing interactive element detection on the image to be identified, and extracting display information of the interactive element in the image to be identified; Based on the attribute information of the image to be identified, feature information is extracted from the display information of the interactive element to obtain description information of the interactive element, wherein the attribute information includes at least one of theme information and style information of the image to be identified.

4. The page generation method according to claim 1, characterized in that: The step of determining the interactive association between the interactive elements according to the element information of the plurality of interactive elements includes: Determining semantic associations between the element information of the interactive elements according to the element information of the multiple interactive elements; Based on the semantic association between the element information, the interactive association between the interactive elements is determined.

5. The page generation method according to claim 1, characterized in that: The converting the interactive element into a task execution model of the interactive element based on the element information of the interactive element and the interactive association between the interactive elements includes: Performing semantic understanding on the element information of the interactive elements and the interactive associations between the interactive elements to extract first context features of the interactive elements, where the first context features represent the semantic associations between the interactive elements; The first context feature is decoded to obtain a task execution model of the interactive element, wherein the task execution elements in the task execution model are determined based on the semantic associations between the interactive elements represented by the first context feature.

6. The page generation method according to claim 5, characterized in that: The semantically understanding the element information of the interactive element and the interactive association between the interactive elements to extract the first context feature of the interactive element includes: Extracting keywords related to the interactive task corresponding to the interactive element from the element information of the interactive element; Semantic understanding is performed on the keywords and the interactive associations between the interactive elements to extract first context features of the interactive elements.

7. The page generation method according to claim 5, characterized in that: The step of converting the interactive element into a task execution model of the interactive element based on the element information of the interactive element and the interactive association between the interactive elements further includes: Based on the element information of the interactive element, the interactive element is split into at least one sub-element, and the sub-element is used to convert and obtain the task execution model.

8. The page generation method according to claim 1, characterized in that: The determining of the interactive configuration information corresponding to the task execution element includes: Acquire at least one piece of guidance information, where the guidance information includes configuration information corresponding to a specified scenario; Based on the element information of the interactive element corresponding to the task execution element, determining the to-be-used guidance information corresponding to the task execution element from the at least one guidance information; Performing semantic understanding on the guidance information to be used and the task execution element to extract a second context feature of the task execution element, wherein the second context feature represents a semantic association between the guidance information to be used and the task execution element; The second context feature is decoded to generate the interaction configuration information corresponding to the task execution element through the semantic association between the guidance information to be used and the task execution element, and the interaction configuration information corresponds to the designated scene.

9. The page generation method according to claim 1, characterized in that: After generating the interactive page based on the interactive configuration information, the method further includes: Call the page test interface; Testing the interactive page through the page testing interface; When the interactive page test fails, modifying the task execution element of the task execution model; The modified task execution element is used as the task execution element of the task execution model, and the execution step is returned to determine the interactive configuration information corresponding to the task execution element and subsequent steps until the interactive page test passes.

10. The page generation method according to claim 9, characterized in that: When the interactive page test fails, modifying the task execution element of the task execution model includes: When the interactive page test fails, obtaining the test result; Based on the test results, the task execution elements of the task execution model are modified.

11. The page generation method according to any one of claims 1 to 10, characterized in that: The method further comprises: Acquire a subject image in the image to be identified and attribute information of the image to be identified, wherein the subject image is a partial image in the image to be identified that is related to the subject information of the image to be identified; Based on the theme image and the attribute information, promotional content corresponding to the interactive page is generated, and the promotional content includes one or more of a promotional image and a promotional text.

12. A page generating device, characterized in that: include: An element recognition unit, used to perform interactive element recognition on the image to be recognized, and obtain element information of a plurality of interactive elements, wherein the interactive elements are image elements related to the interactive task displayed in the image to be recognized; a relationship determination unit, configured to determine an interactive relationship between the interactive elements according to the element information of the plurality of interactive elements; an element conversion unit, configured to convert the interactive element into a task execution model of the interactive element based on the element information of the interactive element and the interactive association between the interactive elements, wherein the task execution model includes at least one task execution element for realizing the interactive task corresponding to the interactive element; A configuration determination unit, used to determine the interactive configuration information corresponding to the task execution element; A page generation unit is used to generate an interactive page based on the interactive configuration information, wherein the interactive page displays an interactive component corresponding to the interactive configuration information, and the interactive component is used to implement the interactive task corresponding to the interactive element in the interactive page.

13. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a plurality of instructions; the processor loads instructions from the memory to execute the steps in the page generation method according to any one of claims 1 to 11.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the steps in the page generation method according to any one of claims 1 to 11.

15. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, the steps in the page generation method according to any one of claims 1 to 11 are implemented.

Images