Interaction method of travel service application, electronic device and storage medium

Abstract

The application provides an interaction method of a travel service application, an electronic device and a storage medium, relates to the technical field of computer Internet, and the method comprises the following steps: the travel service application displays a first AI component matched with a current function page, so that the AI entrance is flexibly adapted to different scenes; in response to a component form change instruction or a page switching event, the first AI component is switched to a second AI component, form flexible switching is realized; and if there is an AI dialogue session at present, the AI components before and after the switching share the session, and the user does not need to reinitiate interaction. Through the comprehensive mechanism of form switching and session sharing, not only is the obstruction and operation interference of the AI entrance to the page content reduced, but also the dialogue continuity is continuously maintained across pages and forms, the scene adaptation flexibility and service continuity of AI interaction are improved, and the operation burden and decision cost of the user are reduced.

Description

Technical Field

[0001] This application relates to the field of computer and internet technology, and in particular to an interactive method for a travel service application, an electronic device, and a readable storage medium. Background Technology

[0002] With the development of computer and internet technology, travel transaction service applications have become important online tools for users to purchase travel service products. Travel transaction service applications refer to online applications that provide users with transaction services for travel service products, such as OTP (Online Travel Platform) and OTA (Online Travel Agency) applications; travel service products refer to goods related to travel services such as transportation (e.g., air travel, train travel), hotel accommodation, and scenic spot visits, such as air tickets, train tickets, hotel room types, and scenic spot tickets.

[0003] For travel transaction service applications, they can offer users a wide variety of travel service products, such as different hotels, different transportation tickets, different attractions, different travel routes, and other similar but different service products. Faced with such a diverse range of travel service products, due to the complex information structure of non-standardized products, the core attributes valued vary significantly across different service scenarios. For example, users prioritizing efficiency may prefer transportation tickets with relatively short travel times, while users prioritizing cost-effectiveness may prefer affordable tickets. Therefore, users need to choose the service products suitable for their travel needs from a wide variety of options. With the development of AI technology, users can use AI to assist them in information filtering and decision-making.

[0004] However, related travel service applications have significant shortcomings in AI interaction functions that assist users in decision-making: First, there is a serious disconnect between AI interactive components and complex service scenarios. For example, the pages presented by travel service applications include diverse functional pages such as homepage, search page, order details page, and product details page. Regardless of the page, the AI ​​entry point is displayed in a fixed form. As a result, on the one hand, the fixed design is difficult to adapt to the ever-changing travel scenarios, and on the other hand, the AI ​​entry point can easily obscure the key information presented on the page, interfering with the user's core operations. Furthermore, on pages that require guidance, the AI ​​entry point lacks sufficient information carrying capacity to demonstrate its service scope.

[0005] Secondly, the continuity of cross-page services between different pages is poor. For example, during the process of travel planning, users often need to jump between different pages (such as from the homepage consultation to a specific hotel page). While browsing different pages, it is often difficult to quickly and easily activate AI, and the conversation context is easily lost during page switching. Users can only access the corresponding historical conversation context by jumping to a specific page (such as the historical conversation area) after switching pages. This makes it impossible to achieve seamless intelligent assistance from global search to specific product booking, increasing the user's decision-making cost and operational burden. Summary of the Invention

[0006] This application provides an interactive method, electronic device, and readable storage medium for travel service applications, in order to solve or partially solve the problems of single AI interaction entry form, inability to adapt to different functional pages, and poor service continuity across pages in travel service applications.

[0007] This application discloses an interaction method for a travel service application, which provides a graphical user interface (GUI) through an electronic device. The GUI presents content that includes at least the currently viewed traditional function pages of the travel service application. The method includes: The first AI component displays the target component form that matches the current traditional function page; In response to a component form change instruction or a page switching event for the first AI component, the first AI component is switched to a second AI component with another component form; If an AI dialogue session exists, the first AI component and the second AI component share the AI ​​dialogue session.

[0008] This application also discloses an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory is used to store computer programs; When the processor executes a program stored in the memory, it implements the method described in the embodiments of this application.

[0009] This application also discloses a readable storage medium storing instructions that, when executed by one or more processors, cause the processors to perform the methods described in this application.

[0010] The embodiments of this application have the following advantages: In this embodiment, the travel service application displays a first AI component with a target component form matching the current traditional function page. This allows the AI ​​component to be presented in a component form adapted to the current scenario within the current traditional function page, improving the flexibility and scenario adaptability of the AI ​​entry point presentation. In response to a component form change command or page switching event for the first AI component, the application switches the first AI component to a second AI component with a different component form. This enables flexible switching between different component forms of the AI ​​component. Furthermore, if an AI dialogue session exists, the first and second AI components share this AI dialogue session. This allows users to continue interacting within the same AI dialogue session even after a component form change or page switching, without needing to worry about component changes. By switching forms or jumping between pages to re-initiate AI interactions or redescribe needs, a comprehensive mechanism combining component form switching and session sharing is achieved. This not only enables flexible form switching of AI components under different scenario requirements, reducing the obstruction of page content and interference with user operations by AI entry points, but also ensures the continuity of AI dialogue sessions across forms before and after component form changes or page switching. Furthermore, by sharing the same AI dialogue session before and after switching, users can continue the interaction progress of the same AI dialogue in different component forms. The overall design of the above-mentioned solution in this application improves the adaptability of AI interaction in different scenarios and the service continuity across pages and forms in travel service applications, reducing the operational burden and decision-making costs for users in the process of travel planning and decision-making. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating the steps of an interaction method for a travel service application provided in this embodiment of the application. Figure 2 This is a schematic diagram of the AI ​​component provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the application interface provided in the embodiments of this application; Figure 4 This is a schematic diagram of component form switching provided in the embodiments of this application; Figure 5 This is a schematic diagram of the AI ​​session processing flow provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the AI ​​component provided in an embodiment of the present invention; Figure 7 This is a schematic diagram of the AI ​​component provided in an embodiment of the present invention; Figure 8 This is a schematic diagram of the AI ​​component provided in an embodiment of the present invention. Detailed Implementation

[0012] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0013] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0014] As examples, in travel transaction service applications, users can browse various traditional functional pages such as homepage, list page, search page, product details page, order details page, personal center page, and message page on electronic devices through a graphical user interface. During this process, they can use the AI-assisted functions provided by the application for travel planning, product filtering, and decision-making. However, these travel service applications have significant shortcomings in AI interaction functionality. First, the component form of the AI ​​interaction component is uniform and fixed across different functional pages. Regardless of the functional page, the same component form is used to display the AI ​​entry point, causing the AI ​​entry point to obscure page content or interfere with user operation in some of the presented functional pages. Second, the interactive experience between AI interaction components before and after page switching lacks continuity. When a user initiates an AI dialogue through the AI ​​entry point on one page and switches to another page, the AI ​​dialogue from the previous page cannot continue in the AI ​​interaction component after the page switch. Users need to jump to a specific page (such as the history conversation area) to access the corresponding history conversation context, resulting in an incoherent AI assistance process across pages.

[0015] In response, this application constructs a component form library containing various component forms and a cross-component session sharing mechanism. When an electronic device detects that a user has opened or switched to a function page in a travel service application, it can obtain the page context information of that function page and determine the component form matching that function page from the component form library based on the page context information. Then, it displays the first AI component matching that traditional function page on the current traditional function page. Through this method, the AI ​​component can be presented in an information-carrying form adapted to the current page scenario requirements on different function pages. This not only reduces the obstruction of key information and interference with core operation paths by the AI ​​entry point, but also provides sufficient information carrying capacity to demonstrate the scope of AI services on pages requiring functional guidance, improving the flexibility and scenario adaptability of the AI ​​entry point presentation.

[0016] Furthermore, by introducing a cross-component form-switching mechanism, when a user executes a component form change command on the currently displayed first AI component, or when a page switching event is detected, the application can switch the first AI component to a second AI component with a different form. This enables the AI ​​component to proactively switch in response to user intent within the same functional page, and to automatically switch in response to the target page context during page transitions. Through these methods, the AI ​​component can not only respond to the user's personalized needs for information capacity at different interaction stages, but also automatically adapt to the scene characteristics of the target page during page jumps, without requiring manual adjustments from the user. This enhances the interactive flexibility and cross-page form adaptation capabilities of the AI ​​component.

[0017] Furthermore, during the switch from the first AI component to the second AI component, if an AI dialogue session already exists, the first and second AI components share the same AI dialogue session. The application transfers the session identifier and dialogue context from the AI ​​component before the switch to the AI ​​component after the switch, allowing users to continue the same AI dialogue before and after the switch. This eliminates the need to re-initiate AI interactions or re-describe requirements due to component form changes or page jumps, thus ensuring the continuity of the AI ​​dialogue session during switches across component forms and functional pages. This avoids context loss and interaction interruptions caused by switching, improving the continuity of intelligent assistance across pages and forms.

[0018] Furthermore, in page switching scenarios, travel service applications can respond to page switching events, obtain the target page context information of the target traditional function page, and determine the second target component form that matches the target traditional function page based on the target page context information. By comparing the current AI component form with the second target component form, if they match, the current form is maintained; if they do not match, the application switches to the second AI component corresponding to the second target component form. This allows the AI ​​component to adaptively adjust its form as the user browses different function pages, always presenting itself in the form most suitable for the current page scenario. At the same time, it maintains the continuity of the AI ​​dialogue conversation throughout the entire page switching process, ensuring that the AI-assisted service always matches the page scenario the user is currently browsing. This improves the user's seamless cross-page experience during travel planning and decision-making, and reduces operational burden and decision-making costs.

[0019] To enable those skilled in the art to better understand the technical solutions in the embodiments of this application, some technical features involved in the embodiments of this application are explained and described below: In this embodiment, the graphical user interface (GUI) is a visual interface environment presented on the display screen of an electronic device, through which the user interacts with the travel service application. When the electronic device is running the travel service application, the GUI can display the currently viewed traditional function pages of the travel service application, such as the homepage, list page, search page, product details page, order details page, personal center page, message page, and itinerary planning page. These traditional function pages are the main carriers for the travel service application to display travel service product information, transaction entry points, and user operation paths to the user, constituting the main scenario for the user to perform operations such as travel planning, product browsing, and booking decisions.

[0020] In this application embodiment, AI UI refers to the collective term for visual interface elements and interaction logic displayed in conjunction with traditional function pages in a graphical user interface, used to provide users with opportunities to interact with the AI ​​system. The methods of this associated display include, but are not limited to: AI UI embedded in the view hierarchy of traditional function pages, floating above the view hierarchy of traditional function pages, and parallel display with traditional function pages in a split-screen format, etc., which this application does not limit.

[0021] Furthermore, the AI ​​UI includes at least AI components and AI dialogue windows. AI components serve as the entry point to the AI ​​UI, representing the AI ​​functionality within traditional functional pages and enabling user access. AI dialogue windows act as the interactive interface, displaying the dialogue content within AI conversations and receiving user input. Moreover, AI components and AI dialogue windows are interconnected at the conversation level; the same AI dialogue session can be created and invoked within AI components of different forms and continued within the AI ​​dialogue window. The dialogue content within the AI ​​dialogue window can be presented as a summary in the associated area of ​​the AI ​​component or in the conversation storage control after the AI ​​dialogue window is collapsed. Users can quickly restore the AI ​​dialogue window and continue the same AI dialogue session by triggering the AI ​​component or the conversation storage control.

[0022] It should be noted that the AI ​​component has multiple switchable component forms in the graphical user interface, with different component forms having different size parameters and information carrying capacity. In the embodiments of this application, the AI ​​UI can adaptively determine the target component form based on the page context information of the current traditional function page, and present it in the current traditional function page in the target component form, so that the AI ​​UI can be presented in different traditional function pages with an information carrying capacity that adapts to the needs of the current page scenario. At the same time, the AI ​​UI can switch the AI ​​component from the current component form to another component form in response to the user's command to change the form of the AI ​​component, or in response to a page switching event, and maintain the continuity of the AI ​​dialogue session before and after the switch. That is, the AI ​​component after the switch shares the same AI dialogue session as the AI ​​component before the switch, and the dialogue content displayed in the AI ​​dialogue window remains consistent before and after the switch, so the user does not need to re-initiate AI interaction due to changes in the form of the AI ​​component or page switching.

[0023] Reference Figure 1 This document illustrates a flowchart of the interaction method for a travel service application provided in an embodiment of this application. The method utilizes an electronic device to provide a graphical user interface (GUI). The GUI presents content that includes at least the currently viewed traditional function pages of the travel service application, and may specifically include the following steps: Step 101: Display the first AI component that matches the target component form of the current traditional function page; In this embodiment, the application can be applied to a client application of a travel transaction service application (hereinafter referred to as the application). This client application runs on an electronic device (e.g., a smartphone, tablet, wearable device, laptop, or desktop computer) and provides a graphical user interface to the user through the electronic device's display screen. The content presented in the graphical user interface includes at least the currently viewed traditional function pages of the travel service application, such as the homepage, list page, search page, product details page, order details page, personal center page, message page, and itinerary planning page. This application does not impose any limitations on this.

[0024] The interactive depth and information carrying capacity of AI components vary across different traditional functional pages. On high-level navigation pages like the homepage, users expect extensive service guidance and travel inspiration from AI, thus requiring AI components with high information carrying capacity to display multiple function entry points and recommended content. On operation-focused pages like product detail pages or order detail pages, users' core attention is concentrated on product information and purchase decisions, requiring AI components to be presented in a lightweight form to avoid obscuring core content or interfering with the user's operation path. Based on this, the application can pre-build a component form library, defining multiple AI components with different information carrying forms to adapt to the scenario requirements of different functional pages. Furthermore, when a user enters or opens a functional page, the application can obtain the page context information and determine the target component form matching that functional page, thus displaying the corresponding first AI component on the current traditional functional page, ensuring that the form of the AI ​​component matches the scenario requirements of the current traditional functional page.

[0025] In some feasible implementations, the target component form matching the current traditional function page can be determined first, and then the first AI component corresponding to the target component form can be rendered in the current traditional function page. Optionally, by obtaining the first page context information corresponding to the current traditional function page, and then extracting at least the first page parameters from the first page context information, the first page parameters include at least one of the first page type and the first page information density value. Then, based on at least one of the first page type and the first page information density value, the first target component form matching the current traditional function page can be determined from a preset component form library. Finally, the first AI component corresponding to the target component form can be rendered in the current traditional function page.

[0026] In its implementation, each functional page in a travel service application corresponds to a unique view controller or page route identifier in the client application. The application can determine the page type of the current traditional functional page by parsing its route stack information or view controller identifier. For example, in iOS, the current page type can be determined by obtaining the class name of the top-level UIViewController or its custom route identifier; in Android, it can be determined by obtaining the class name of the current Activity or Fragment or its route annotation information. Furthermore, the application can extract page layout parameters by parsing the view hierarchy of the current traditional functional page and calculate the page information density value based on these layout parameters.

[0027] The page information density value reflects the density of information content in a traditional functional page. It can be calculated using at least one of the following parameters: the number of interactive elements (e.g., buttons, clickable cards, input boxes), the length of text content (e.g., the total number of text characters on the page or the percentage of visible text area), and the number of image elements (e.g., images, carousels). For example, a product details page contains a large amount of product descriptions, specifications, user reviews, etc., with relatively dense interactive elements and text content, resulting in a relatively high page information density value. Conversely, a message page typically consists of a structured message list with a relatively regular and limited number of interactive elements, leading to a relatively low page information density value. Optionally, the page information density value can be calculated based on the above parameters using weighted summation or normalization methods, such as a normalized value ranging from 0 to 1. This application does not impose any limitations on this.

[0028] Furthermore, after obtaining the first page context information corresponding to the current traditional function page, the application extracts page parameters for subsequent component form matching decisions. The page type identifies the page category of the current traditional function page, such as homepage, list page, search results page, product details page, order details page, personal center page, message page, trip planning page, etc. The page information density value quantifies the density of information content on the current traditional function page, serving as a supplementary decision dimension for page type. It is particularly suitable for distinguishing different content density states within the same page type; for example, the page information density value may differ significantly between products with less text and images and products with more text and images on a product details page.

[0029] Based on the extracted first page type and first page information density value, the application can determine the target component form matching the current traditional functional page from the component form library according to preset matching rules. The component form library may include at least one of a first component form, a second component form, and a third component form. Optionally, the first component form corresponds to a first size parameter and includes a global floating trigger icon, or a global floating trigger icon and guidance prompts. This form presents the AI ​​entry point with minimal interface space, suitable for functional pages with high information density or focused user operations. The second component form corresponds to a second size parameter and includes input controls. This form provides basic AI interaction capabilities, allowing users to input text, voice, or images, suitable for functional pages with medium information density. The third component form corresponds to a third size parameter and includes a functional category area and a guidance prompt area. This form has the highest information carrying capacity, suitable for functional pages that need to demonstrate the scope of AI services and provide rich functional guidance to users.

[0030] In this design, the first size parameter is smaller than the second size parameter, and the second size parameter is smaller than the third size parameter. For example, on a mobile device, the first size parameter can be a circular or rounded rectangular area with a diameter of 48 to 64 pixels; the second size parameter can be an input control with a width of 85% to 95% of the current traditional function page and a height of 48 to 64 pixels; and the third size parameter can be a panel with a width of 85% to 95% of the current traditional function page and a height of 40% to 70% of the visible area height of the current traditional function page. Embodiments of this invention include, but are not limited to, the above examples. Through the design of the above component form library, the application can present AI components with different information carrying capacities on different function pages, realizing a progressive information presentation of the AI ​​entry point from a "lightweight entry point" to a "complete panel."

[0031] In the specific implementation, for the three component forms in the component form library, the application can define corresponding view component classes and rendering parameters for each form. Optionally, the view component corresponding to the first component form can be a floating bubble view, which contains a global floating trigger icon (such as a preset Q icon, where the Q icon is the unified visual anchor point for the AI ​​service of the application). Optionally, a short guiding prompt message, such as "Ask a question" or "AI assistant" text prompts, can also be displayed in the vicinity of the global floating trigger icon (as shown below or to the right). Correspondingly, the view component corresponding to the second component form can be an input control view, which contains input controls. The input controls can include at least one of the following: voice input control (such as a microphone icon button, which allows the user to record voice and convert it into text after clicking), text input box (where the user can enter text content), image input control (such as a camera or album icon button, which allows the user to select or take an image as input content after clicking), and extended control (such as a "+" icon button, which allows the user to expand more input methods or function options after clicking).

[0032] In addition, the view component corresponding to the third component form can be a complete panel view, which includes a function category area and a guidance prompt area. The function category area contains at least one function scene label and task control (such as labels such as "Help me think", "Help me order", "Help me do", etc., and task control such as "View all tasks"). The guidance prompt area displays one of the following: task information, guidance prompt information, and itinerary card associated with the currently selected function scene label.

[0033] Through the aforementioned component form library, the application provides a variety of AI component forms, ranging from lightweight to complete, for different functional pages. This allows AI components to be presented in the most suitable form for different page scenarios, thereby ensuring the accessibility of AI functions while minimizing interference with users' browsing and operation of core content. Furthermore, the three forms share the same session management mechanism, laying the foundation for subsequent cross-form session sharing.

[0034] In some feasible implementations, the process of determining the first target component form matching the current traditional function page based on the first page parameters and the component form library can be as follows: If the page type is a message page, the first component form is used as the first target component form. If the page type is a product details page, order details page, or product information density page, and the page information density value is greater than a first preset threshold, the first component form is used as the first target component form. If the page type is a list page, search results page, or itinerary planning page, the second component form is used as the first target component form. If the page type is a homepage or personal center page, the third component form is used as the first target component form. Furthermore, if the page type does not belong to the preset page type set, the first target component form is determined based on the page information density value, including: if the page information density value is greater than a first preset threshold, the first component form is used as the first target component form; if the page information density value is less than or equal to a first preset threshold but greater than a second preset threshold, the second component form is used as the first target component form; if the page information density value is less than or equal to a second preset threshold, the third component form is used as the first target component form, where the second preset threshold is less than the first preset threshold.

[0035] In practical implementation, the application can determine the form of the target component according to the following preset matching rules: For the message page, which primarily features a message list, the core user action is viewing and replying to messages. Therefore, the AI ​​entry point should be presented in the most lightweight way possible to avoid interfering with browsing the message list. Thus, the first component form (i.e., a floating bubble containing only a global floating trigger icon) can be used as the matching form for the message page. This allows users to activate the AI ​​function by clicking the floating bubble when needed, without affecting the normal browsing of the message content.

[0036] For product detail pages, order detail pages, and similar functional pages, these are all operation-focused pages. Users' core attention on these pages is concentrated on product information, product parameters, or order status, and they need to perform key operations such as adding to cart, purchasing immediately, and confirming the order. Furthermore, when the page information density value exceeds a first preset threshold, it indicates that the current page content is very rich (e.g., containing multiple product carousels, detailed specifications, numerous user reviews, etc.). In this case, the AI ​​entry should be presented in a lightweight form to avoid obscuring key content and interfering with the user's core operation path. Therefore, the first component format can still be considered the matching format for this type of page.

[0037] For functional pages such as list pages, search results pages, and itinerary planning pages, list pages (e.g., product lists, hotel lists) and search results pages are used to display multiple candidate options to users. User behavior on these pages involves browsing and comparing, requiring AI assistance for filtering and recommendations. Itinerary planning pages are used for users to plan and edit travel routes, requiring AI assistance to provide suggestions and solutions. These pages have a moderate information density and operational complexity, requiring AI to have some interactive capabilities (such as receiving user questions through input controls) without needing all the guidance functions of a full panel. Therefore, the second component form (i.e., an input control containing input controls) can be considered the appropriate form for this type of page.

[0038] For homepages or user profile pages, the homepage serves as the entry point for travel service applications, showcasing the application's core functions and recommended content. Users on this page typically browse and explore, expecting broad service guidance. Presenting AI components in a complete panel format allows for a comprehensive demonstration of the AI's service scope (displaying various service categories through functional scenario tags) and recommended content (showing specific actionable tasks or questions through guiding prompts), helping users quickly understand the types of assistance AI can provide. Similarly, the user profile page displays personal information, order records, and saved content. Users on this page may require AI assistance with order-related issues or personalized recommendations. AI components presented in a complete panel format provide rich functional entry points and guiding content, enhancing the discoverability and usability of AI on the user profile page. Therefore, the third component format (i.e., a complete panel format including functional category areas and guiding prompt areas) can be considered the appropriate format for this type of page.

[0039] Furthermore, in practical applications, travel service applications may contain some low-frequency access or non-core functional pages that are not explicitly listed in the aforementioned preset rules. For these pages, the application can perform hierarchical matching based on the page information density value: if the page information density value is greater than a first preset threshold (indicating dense page content, such as a help center page or event landing page containing a large amount of text and interactive elements), then the first component form is used as the first target component form; if the page information density value is less than or equal to the first preset threshold but greater than the second preset threshold (indicating moderate page content density), then the second component form is used as the first target component form; if the page information density value is less than or equal to the second preset threshold (indicating sparse page content, such as a blank status page or a simple settings page), then the third component form is used as the first target component form. The second preset threshold is less than the first preset threshold. Through this hierarchical matching strategy, the application can provide reasonable component form adaptation for functional pages that are not explicitly categorized, ensuring that the AI ​​entry point is presented in a form that adapts to the current page content density on any page.

[0040] Optionally, the first preset threshold and the second preset threshold can be adjusted according to the actual application scenario. For example, the first preset threshold can be set to 0.7, and the second preset threshold can be set to 0.3, etc. In other examples, the first preset threshold can be set to any value between 0.6 and 0.8, and the second preset threshold can be set to any value between 0.2 and 0.4. This application does not impose any restrictions on this.

[0041] By employing the comprehensive matching rules between page types and page information density values, the application can adapt AI component formats to differentiated forms for all functional pages within the travel service application. Specifically, pages requiring focused browsing or core operations, such as message pages and detail pages, adopt a lightweight first-component format to minimize the distraction of user attention from the AI ​​entry point; pages with high functional guidance needs, such as the homepage and personal center pages, adopt a complete third-component format to maximize the discoverability and guidance capabilities of AI services; pages requiring interactive assistance, such as list pages and search results pages, adopt a centrally located second-component format, balancing AI interactivity and content browsing space; and uncategorized pages are matched according to information density levels to ensure full coverage. This achieves a refined matching of AI components with the needs of different functional page scenarios, reducing the obstruction of key information and interference with core operations by the AI ​​entry point, while providing sufficient information capacity to showcase the scope of AI services on pages requiring guidance.

[0042] In some application scenarios, let's take the homepage scenario as an example. After a user opens the homepage of a travel service application, the application obtains the page context information of the homepage, determines that the page type is a homepage, and determines the third component form as the target component form that matches the homepage according to the matching rules. Then, it renders the AI ​​component (complete panel form) in the third component form on the homepage, which includes a functional category area (containing three functional scenario tags such as "Help Me Think", "Help Me Book", and "Help Me Apply") and a guidance prompt area (containing guidance prompts or tasks associated with the currently selected tag), so as to fully demonstrate the service scope of AI and guide the user to use it.

[0043] In other application scenarios, let's take the product details page as an example. After a user clicks on a product from the homepage to enter the product details page, the application obtains the page context information of the product details page, determines that the page type is a product details page, and calculates the current page information density value of the product details page (e.g., 0.82). If the first preset threshold is 0.7, then the current page information density value is greater than the first preset threshold. The application determines the first component form as the target component form that matches the product details page according to the matching rules, and renders the AI ​​component (floating bubble form) in the product details page, which only contains the global floating trigger icon (Q icon). This minimizes the visual footprint of the AI ​​entry point and avoids obstructing product image and text browsing and purchase operations. If the user needs AI assistance while browsing the product details page, they can click the Q icon at any time to expand the AI ​​component for interaction.

[0044] Step 102: In response to a component form change instruction or a page switching event for the first AI component, the first AI component is switched to a second AI component with another component form. In this embodiment, after the first matching AI component has been displayed in step 101 based on the page context information of the current traditional function page, the user may need to change the form of the AI ​​component during use, or the form of the current AI component may not match the scenario requirements of the new page when the user switches between pages. For example, the user may want to expand the lightweight first component form into a more feature-rich third component form to obtain more guidance, or the user may want to compress the complete third component form into a lightweight first component form to free up screen space. As another example, when the user switches from the homepage (matching the third component form) to the product details page (matching the first component form), the form of the AI ​​component needs to switch from the third component form to the first component form to adapt to the scenario requirements of the new page, etc.

[0045] Therefore, the application can respond to the user's command to change the form of the first AI component or the detected page switching event, and switch the first AI component to a second AI component with another form. This enables flexible switching of AI components between different scenarios, supporting both manual switching of forms by users to meet personalized interaction preferences and automatic adaptation of forms by the system to meet the needs of different page scenarios based on page switching.

[0046] In some feasible implementations, during the process of switching the first AI component to a second AI component with a different component form based on the component form change command, the application can first identify the current component form of the currently displayed first AI component, and then execute the corresponding switching operation according to the type of form change command input by the user. Optionally: When the first AI component is in its first component form, in response to a trigger operation on the global floating trigger icon (e.g., the user clicks the Q icon), the global floating trigger icon remains displayed in the graphical user interface. The input controls for the second component form of the AI ​​component, or the functional category area for the third component form, or the functional category area and guidance prompt area for the third component form of the AI ​​component are displayed in a preset adjacent area of ​​the global floating trigger icon. That is, when the AI ​​component is currently in its first component form (only containing the floating bubble form of the global floating trigger icon), clicking the global floating trigger icon will trigger the expansion of the AI ​​component, expanding it from the minimized first component form to the second or third component form. During expansion, the global floating trigger icon remains visible, acting as a visual anchor. Simultaneously, in its preset adjacent area (e.g., immediately above, below, to the left, or to the right of the global floating trigger icon; the specific expansion direction can be adaptively determined based on the current position of the global floating trigger icon in the graphical user interface), the input controls for the second component form, or the functional category area for the third component form (only displaying the functional category labels), or the functional category area and guidance prompt area for the third component form (the complete panel) are displayed. In the specific implementation, the specific expansion target form can be further determined based on the page type and / or page information density value of the current traditional function page determined in step 101 above, or it can be determined based on the user's historical interaction behavior, so that when the user expands from the first component form, he / she can reach the form level that is most suitable for the current scenario and user habits.

[0047] When the first AI component is an AI component in the second component form, in response to a shape compression command for the second component form (e.g., the user clicks the collapse button on the input control, or performs a downward swipe operation on the input control), the input control is hidden, meaning the AI ​​component is compressed back from the second component form to the first component form; or, in response to a shape expansion command for the second component form (e.g., the user clicks the expand button on the input control, or performs an upward swipe operation on the input control), the input control is hidden, and the functional category area of ​​the third component form, or the functional category area and guidance prompt area of ​​the third component form, is displayed in a preset adjacent area of ​​the global floating trigger icon, meaning the AI ​​component expands upward from the second component form to the third component form. During the expansion from the second component form to the third component form, the input control is hidden (or collapsed and merged into the bottom area of ​​the third component form), and the functional category area of ​​the third component form, or the functional category area and guidance prompt area, is presented in a preset adjacent area of ​​the global floating trigger icon, allowing the AI ​​component to be presented to the user in the form of a complete panel, providing richer functional guidance and interactive options.

[0048] When the first AI component is an AI component in the third component form, in response to a shape compression command for the third component form (e.g., the user clicks the collapse button on the full panel, performs a downward swipe on the panel, or clicks a blank area outside the panel), the function category area, or the function category area and the guidance prompt area, are hidden, and an input control is displayed in a preset adjacent area of ​​the global floating trigger icon, or only the global floating trigger icon is displayed. That is, the AI ​​component is compressed from the third component form back to the second component form or the first component form. Specifically, if the shape compression command indicates compression to the second component form, the function category area and the guidance prompt area are hidden, and an input control is displayed in a preset adjacent area of ​​the global floating trigger icon, so that the AI ​​component is presented as an input control; if the shape compression command indicates compression to the first component form, only the global floating trigger icon is displayed, and all other UI elements are hidden. In the specific implementation, the target shape of compression can be determined by judging the type of shape compression command (e.g., whether the user clicks the "Compress to Input Control" button or the "Close" button), or the degree of compression can be determined by detecting the swipe distance or gesture speed of the user's operation.

[0049] Among them, the shape expansion command is any one of the expansion operation or the upward swipe operation applied to the currently displayed AI component; the shape compression command is any one of the collapse operation, the downward swipe operation, or the click operation on the blank area outside the component applied to the currently displayed AI component.

[0050] In some examples, the shape expansion and shape compression commands can also be other preset operation types, such as double-click operations and long-press operations, which are not limited in this application. By defining multiple types of shape change commands, the application can provide users with flexible and natural ways to switch the shape of AI components. Users can choose different interaction methods according to their personal operating habits and usage scenarios, thereby improving the interaction flexibility and user experience of AI components in different usage scenarios.

[0051] Optionally, when the first AI component is in its first component form (such as a floating bubble), the application also needs to determine whether clicking the global floating icon expands the AI ​​component to a second or third component form. Specifically, this switching process may include: In response to a trigger operation on the global floating trigger icon in the first AI component, the time difference between the most recent time the AI ​​component in the first component form was collapsed and the current time is obtained. Here, the most recent time it was collapsed refers to the moment when the AI ​​component in the first component form was most recently compressed back to the first component form from the second or third component form, that is, the moment when the AI ​​component was most recently collapsed by the user from an expanded state or automatically collapsed by the system to a state where only the global floating trigger icon is displayed. The application can obtain this information by recording the timestamp of each form-switching event in the state manager of the AI ​​component.

[0052] Furthermore, if the time difference is less than or equal to a preset time threshold, the AI ​​component in the first component form is switched to the AI ​​component in the second component form. If the time difference is greater than the preset time threshold, the AI ​​component in the first component form is switched to the AI ​​component in the third component form.

[0053] In practical applications, the judgment logic can be determined based on the following user behavior patterns: If the user has just collapsed or closed the AI ​​component (e.g., within 5 minutes), it indicates that the user is still in the process of continuous AI interaction. They may have only temporarily collapsed the panel to view page content or perform other operations. In this case, the user's main need to quickly resume AI interaction is to click the Q icon again. Therefore, it is only necessary to expand to the second component form (input control) so that the user can quickly input content, without loading the full third component form panel, thereby reducing loading time and improving response speed. If a longer period of time has passed since the user last collapsed the AI ​​component (e.g., more than 5 minutes), it indicates that the user may have ended the previous interaction and is now initiating a new AI interaction. In this case, the user may need richer functional guidance to help them clarify their needs. Therefore, expand to the third component form (full panel form), displaying functional category labels and guiding prompts to help the user quickly understand the services that AI can provide and select the appropriate question or task.

[0054] Optionally, the preset time threshold can be adjusted based on the actual application scenario and user behavior statistics. For example, the preset time threshold can be set to 5 minutes. In other examples, the preset time threshold can be set to any value between 3 and 10 minutes. If the application detects that the user frequently folds and unfolds the AI ​​component repeatedly in a short period of time, a shorter time threshold can be used; if the user tends to use the AI ​​component for a longer period of time, a longer time threshold can be used. Through this time difference judgment mechanism, the application can infer the user's interaction intent based on the user's recent operation behavior, thereby adaptively determining the unfolding depth of the AI ​​component. This avoids the loading wait caused by unfolding to the full panel every time, while ensuring that sufficient functional display is provided when the user needs guidance.

[0055] In the above process, the component form of the AI ​​component is switched based on the component form change instruction. In addition, the component form of the AI ​​component can also be switched based on the page switching event triggered by the change of the function page. Optionally, in response to switching the current traditional function page to the target traditional function page, the application can obtain the second page context information of the target traditional function page, then extract the second page parameters from the second page context information. The second page parameters include at least the second page type and / or the second page information density value. Then, based on the second page type and / or the second page information density value, the application can determine the second target component form that matches the target traditional function page from the component form library. If the first target component form is consistent with the second target component form, the component form of the first AI component is maintained in the target traditional function page; if the first target component form is inconsistent with the second target component form, the first AI component is switched to the second AI component corresponding to the second target component form in the target traditional function page.

[0056] In practical implementation, applications can listen for page navigation events in the graphical user interface in real time. These page switching events can occur when a user clicks a page navigation control on the current traditional function page (such as a back button, breadcrumb navigation, bottom tab switching, sidebar menu items, etc.) to jump from the current traditional function page to the target traditional function page; they can also occur when a user performs a page return operation (such as a system-level back gesture or physical back button) to return from the current traditional function page to the target traditional function page; or they can occur by clicking on internal links or recommended cards within the page.

[0057] When a switch from the current traditional function page to the target traditional function page is detected, the application can obtain the second page type and / or the second page information density value of the target traditional function page. The method of obtaining this information is the same as the method described above for obtaining the page context information of the current traditional function page. Then, based on the second page type and / or the second page information density value, a second target component form matching the target traditional function page can be determined from the component form library. The determination rule here is consistent with the rule for determining the first target component form described in step 101, that is, the component form that the target traditional function page should present is determined according to the same preset matching rule (the combination rule of page type and page information density value). The relevant process can be referred to the aforementioned embodiments and will not be repeated here.

[0058] If the form of the first target component is consistent with the form of the second target component, then the form of the first AI component will be maintained in the target traditional function page. That is, if the form of the current AI component is already consistent with the form of the target traditional function page, there is no need to perform a switching operation, and the same form of AI component can continue to be displayed in the target traditional function page, thereby avoiding unnecessary view redrawing and component reconstruction, improving the smoothness of page switching, and saving computing resources.

[0059] If the form of the first target component is inconsistent with the form of the second target component, the first AI component will be switched to the second AI component corresponding to the form of the second target component on the target traditional function page. That is, if the form of the current AI component is inconsistent with the form matched by the target traditional function page, a switching operation will be performed to switch the first AI component to the second AI component (i.e., the AI ​​component instance corresponding to the form matched by the target traditional function page or the updated state of the same instance).

[0060] In practical implementation, the switching method of the AI ​​component during page transitions can vary depending on the type of page transition. For example, when a user navigates between different main pages using the bottom tabs, the AI ​​component can execute a smooth view transition animation along with the page content; when the user enters the next level details page, the AI ​​component can first disappear with a fade-out animation, and then fade in again in the target page with a matching appearance. Through this form adaptation mechanism during page transitions, the form of the AI ​​component can automatically adapt to changes in the user's browsing page, ensuring that the AI ​​entry point is presented in the most suitable form for the current scenario on any functional page, achieving dynamic synchronization between the AI ​​component and the page context. Furthermore, since the same AI dialogue session is shared before and after the transition, the user can directly continue the previous dialogue after the page transition without needing to initiate the interaction again.

[0061] Step 103: If an AI dialogue session exists, the first AI component and the second AI component share the AI ​​dialogue session.

[0062] In this embodiment, after switching the first AI component to the second AI component through the above process, if an AI dialogue session currently exists (i.e., the user initiated an AI dialogue through the AI ​​component before the switch), the first AI component and the second AI component share the AI ​​dialogue session. Specifically, during the switch from the first AI component to the second AI component, the application transfers the session ID and dialogue context (including the historical dialogue message list, current dialogue state, user preference settings, etc.) of the current AI dialogue session from the first AI component to the second AI component. This allows the second AI component to continue the AI ​​dialogue session in the first AI component before the switch, enabling the user to continue viewing historical dialogue content and sending new messages in the second AI component. The AI ​​system can also provide context-aware responses based on the complete dialogue history. During this process, the AI ​​dialogue session itself is not destroyed or rebuilt due to the form change of the AI ​​component or the page switch. Instead, it is always kept in the active session list of the application session manager. The AI ​​component, as the "view carrier" of the session, only changes its presentation when switching between different forms, while the session data it carries remains unchanged. This achieves the continuous maintenance of the AI ​​dialogue session during the form change and page switch of the AI ​​component, without requiring the user to re-initiate AI interaction or redescribe the needs.

[0063] It should be noted that the session sharing mechanism defined in step 103 is executed based on step 102. That is, whether the switch is triggered by a component form change command or a page switch event, the transfer and sharing of session state must be performed after the switch is completed. In addition, if there is no AI dialogue session at present (i.e., the user has not yet initiated a dialogue through the AI ​​component, or an existing session has ended / been cleared), the session sharing step in step 103 can be omitted. In this case, after the first AI component switches to the second AI component, the second AI component is in a "waiting to interact" state, waiting for the user to initiate a new AI dialogue.

[0064] In some feasible implementations, the user may not have established an AI dialogue session before step 103, therefore the AI ​​dialogue session needs to be created before the switch occurs. Optionally, the application can create an AI dialogue session in response to various user interactions with the first AI component: When the first AI component is in its first component form (i.e., floating bubble form), the application can respond to a trigger operation on the global floating trigger icon in the first AI component (such as when the user clicks the Q icon) by switching the first AI component to a second component form or a third component form on the current traditional function page. Subsequently, if the second component form AI component contains at least one guidance prompt message, then in response to a trigger operation on the guidance prompt message, an AI dialogue session corresponding to the guidance prompt message is created; or, in response to a trigger operation on the input control in the second component form AI component, the target input information corresponding to the trigger operation is obtained, and an AI dialogue session corresponding to the target input information is created. If the second component form AI component does not contain guidance prompt messages, then in response to a trigger operation on the input control in the second component form AI component, the target input information corresponding to the trigger operation is obtained, and an AI dialogue session corresponding to the target input information is created. Alternatively, after switching to the third component mode, in response to triggering operations on task information, guidance prompts, or itinerary cards within the AI ​​component of the third component mode, an AI dialogue session corresponding to the task information, guidance prompts, or itinerary card is created. After creating the AI ​​dialogue session, an AI dialogue window corresponding to the AI ​​dialogue session is displayed on the current traditional function page, and the dialogue content associated with the AI ​​dialogue session is displayed in the AI ​​dialogue window.

[0065] In practical applications, the AI ​​dialogue window can be a semi-transparent background panel superimposed on the current traditional function page as a floating layer, or a half-screen dialogue interface presented as a bottom pop-up (Bottom Sheet), or a dialogue page that covers the entire screen. This application does not limit this.

[0066] Accordingly, when the first AI component is in the form of a second component (i.e., an input control), if the AI ​​component in the form of a second component contains at least one guidance prompt, then in response to a trigger operation on the guidance prompt, an AI dialogue session corresponding to the guidance prompt is created; or, in response to a trigger operation on the input control in the AI ​​component in the form of a second component, the target input information corresponding to the trigger operation is obtained, and an AI dialogue session corresponding to the target input information is created. If the AI ​​component in the form of a second component does not contain guidance prompt, then in response to a trigger operation on the input control in the AI ​​component in the form of a second component, the target input information corresponding to the trigger operation is obtained, and an AI dialogue session corresponding to the target input information is created. Furthermore, an AI dialogue window corresponding to the AI ​​dialogue session is displayed on the current conventional function page, and dialogue content associated with the AI ​​dialogue session is displayed in the AI ​​dialogue window.

[0067] Accordingly, when the first AI component is in the form of the third component (i.e., the complete panel form), in response to the triggering operation of the task information, guidance prompts, or itinerary card in the AI ​​component in the form of the third component, an AI dialogue session corresponding to the task information, guidance prompts, or itinerary card is created; and an AI dialogue window corresponding to the AI ​​dialogue session is displayed on the current traditional function page, and the dialogue content associated with the AI ​​dialogue session is displayed in the AI ​​dialogue window.

[0068] Through the above methods, users can initiate AI dialogue sessions via multiple interaction paths in different component formats. For the first and second component formats, users can quickly initiate dialogues by clicking on pre-set questions or by entering custom content through input controls, providing two parallel paths: pre-set quick interactions and custom interactions. For the third component format, users can quickly initiate dialogues by clicking on task information, pre-set questions, or itinerary cards. These pre-generated interactions are based on the current page context or user behavior, helping users quickly clarify their needs and efficiently initiate dialogues. This diverse session creation method lowers the cognitive threshold for new users (providing pre-set questions through pre-set questions) while also allowing advanced users greater freedom (entering custom content through input controls), improving the accessibility and ease of use of AI functions across different user groups.

[0069] In some application scenarios, let's take the product details page as an example. After a user clicks on a product from the homepage to enter the product details page, the application determines the first component form (floating bubble form) as the matching AI component form based on the page type and information density value of the product details page. When the user clicks the Q icon in the floating bubble, the application obtains the time difference between the last time the AI ​​component in the first component form was collapsed and the current time. If the time difference is less than or equal to a preset time threshold (e.g., 5 minutes), it expands into the second component form (input control), and displays a guiding prompt generated based on the current product above the input control, such as "Does this hotel have a gym?", "What attractions are nearby?", "Is this hotel suitable for families with children?". When the user clicks on one of the guiding prompts (such as "Does this hotel have a gym?"), the application uses this text as input information to create an AI dialogue session and displays the AI ​​dialogue window on the product details page. The AI ​​system can provide targeted responses to the user based on the product details information. If the time difference exceeds the preset time threshold, it will expand into the third component form (complete panel form), displaying a function category area (including labels such as "Help me think", "Help me order", "Help me do" and a guidance prompt area (including guidance content corresponding to each label). Users can switch labels to browse different categories of guidance content and initiate conversations.

[0070] Furthermore, when the first AI component is in the form of a third component (i.e., a complete panel), the application can also create corresponding AI dialogue sessions through interaction with functional scene tags. Optionally: The first AI component displays a first scene label, a second scene label, and a third scene label. In this embodiment, taking a travel service application as an example, the first scene label can be a "Help me think" label, used to guide users in AI dialogues related to travel inspiration and itinerary planning; the second scene label can be a "Help me book" label, used to guide users in AI dialogues related to booking travel service products; and the third scene label can be a "Help me handle" label, used to guide users in AI dialogues related to handling to-do items in the current itinerary and processing travel services.

[0071] In its implementation, the application can respond to the selection of a first scenario label (e.g., the user clicking the "Help Me Think" label) by displaying at least one travel task to be decided or executed, and / or at least one first guidance prompt related to travel inspiration or itinerary planning, in the guidance prompt area. Before displaying the at least one travel task to be decided or executed, and / or at least one first guidance prompt related to travel inspiration or itinerary planning, the application needs to perform the following preprocessing steps: The system retrieves the user's current travel information, which may include information on booked transportation products (such as air tickets, train tickets, etc.), hotel accommodation products (such as booked hotels and check-in / check-out dates), and scenic spot tour products (such as purchased tickets and entry dates). This information can be obtained from the order system or itinerary management module of the travel service application.

[0072] Next, the application can extract to-do list information and / or decision-making information from the dialogue content in the AI ​​conversation, as well as identify related itinerary nodes in the current itinerary information. In specific implementation, the application can use natural language processing technology to semantically analyze the dialogue content in the AI ​​conversation, identifying target travel service products, target itinerary nodes, and / or target actions involved in the dialogue content. For example, if a user states in a conversation, "I want to book a hotel in Sanya, with a budget of around 800, and it needs to be near the beach," the application can extract key information such as "book a hotel," "Sanya," "budget 800," and "near the beach" from the dialogue content, identifying the to-do list "book a hotel in Sanya." Similarly, if a user says, "Help me check if there are any earlier return flights," the application can extract the decision-making item "modify return flight." Simultaneously, the application can also identify related itinerary nodes from the current itinerary information, such as identifying the "waiting to check in" node from booked flight information and the "waiting to check in" node from booked hotel information.

[0073] Furthermore, based on the extracted content, the application can generate at least one trip task to be decided or executed corresponding to a to-do list, and / or a decision-making item, and / or a trip node. Specifically, the application can generate corresponding trip tasks based on the extracted to-do list information, decision-making item information, and identified trip nodes. The trip tasks to be decided or executed may include at least one of the following: a task to confirm a booked travel service (e.g., "confirm hotel booking in Sanya"), a task to confirm a payment order (e.g., "confirm flight ticket payment"), a task to confirm modified trip information (e.g., "confirm whether the return date needs to be modified"), a task to confirm supplemented user information (e.g., "supplement passenger identification information"), a task to confirm a selected travel plan (e.g., "select transportation from the airport to the hotel"), and a task to confirm a departure time (e.g., "confirm whether the departure time is 9:00 AM"). These trip tasks are displayed in an interactive list or card format in the guidance prompt area, and users can create a corresponding AI dialogue session by clicking on any task. Through the above methods, the application can automatically extract the trip items that users need to process or make decisions about and present them in a centralized form as tasks. Users do not need to manually recall and sort out the to-do items, which reduces the cognitive burden in the trip management process.

[0074] Simultaneously, the application can also use the first page context to determine the target scenario type of the current traditional function page, and extract at least one candidate guidance prompt that matches the target scenario type from a preset guidance information library, generating at least one first guidance prompt related to travel inspiration or itinerary planning. In specific implementation, the application extracts the page type identifier and / or page content keywords from the page context information of the current traditional function page, and determines the target scenario type corresponding to the current traditional function page based on the page type identifier and / or page content keywords. The target scenario type may include at least one of destination recommendation scenario, itinerary planning scenario, product booking scenario, and travel preparation scenario.

[0075] For example, if the current traditional function page is the homepage, and its page type is identified as "Homepage," then the target scenario type can be determined as "Itinerary Planning Scenario" or "Destination Recommendation Scenario." If the current traditional function page is a travel guide page for a specific destination, and its page content keywords include "Sanya," "guide," and "attraction recommendation," then the target scenario type can be determined as "Destination Recommendation Scenario." The application filters at least one candidate guidance prompt that matches the target scenario type from a pre-set guidance information database and generates the first guidance prompt. During the filtering process, it can further obtain the user's historical behavior data (such as historical search records, historical browsing records, historical collection records, and historical order records). Based on the historical behavior data and the target scenario type, it filters at least one candidate guidance prompt that matches the target scenario type and is associated with the user's historical behavior data from the pre-set guidance information database, achieving personalized guidance message recommendations. For example, in the homepage scenario, guidance prompts such as "Help me plan a 7-day self-driving tour around Hainan Island," "Where to go for a self-driving tour to see cherry blossoms around Hangzhou this weekend?" and "Recommended family trips for May Day" can be generated.

[0076] Alternatively, in response to the selection of a second scenario tab (e.g., the user clicking the "Subscribe for me" tab), at least one second onboarding prompt message is displayed in the onboarding prompt area. Before displaying at least one second onboarding prompt message, the application needs to perform the following preprocessing steps: The app acquires user behavior data within travel service applications. This data includes at least one of the following: search behavior, browsing behavior, add-to-cart behavior, and favorites behavior. For example, the app can acquire a user's search history over a recent period (e.g., the last 7 or 30 days) (e.g., the user searched for keywords such as "Sanya hotels" or "Wuzhizhou Island tickets"), browsing history (e.g., the user viewed multiple hotel and attraction details pages in the Sanya area and the duration of their stay), add-to-cart history (e.g., the user added a hotel to their cart but did not complete the booking), and favorites history (e.g., the user saved a travel itinerary or a hotel).

[0077] Based on user behavior data, the application can determine the target products or destinations that a user is currently interested in. The application can comprehensively determine a user's current target products (e.g., a hotel in the Haitang Bay area of ​​Sanya) or target destinations (e.g., Sanya City) based on the frequency and freshness of search records, page dwell time and number of views in browsing history, product types in add-to-cart records, and content types in favorites records. For example, if a user has searched for "Sanya hotels" multiple times in the past 3 days and viewed the details pages of several Sanya hotels, and added one of the hotels to their shopping cart, the application can determine that the user's current target products are that hotel or other similar hotels in the same area.

[0078] The system selects at least one guiding prompt related to the target product or destination from a pre-defined guidance information database as the second guiding prompt. This is applied after the target product or destination is determined, by selecting guiding prompts associated with it from the database. For example, if the target product is the Sheraton Sanya Haitang Bay Hotel, the second guiding prompt might include "Book the Sheraton Sanya Haitang Bay Hotel," "View user reviews of this hotel," or "Compare other nearby hotels." If the target destination is Sanya City, the second guiding prompt might include "Purchase tickets for Wuzhizhou Island," "Book a day trip to popular Sanya attractions," or "View Sanya travel guides." Through this method, the second guiding prompt accurately reflects the user's current consumption intentions and interests, shortening the decision-making path from browsing to booking and reducing the user's operational burden during product selection and booking.

[0079] In addition, the application can also respond to the selection of a third scenario tab (e.g., when the user clicks the "Help me with this" tab) by displaying a trip card corresponding to the target trip node in the user's current trip in the onboarding prompt area. Before displaying the trip card corresponding to the target trip node in the user's current trip, the application needs to perform the following preprocessing steps: The application obtains the user's current travel information, which includes at least one of the following: booked transportation, hotel accommodation, and / or sightseeing tours. The application can retrieve all travel-related product information booked by the user from the travel service application's order system or travel management module.

[0080] Next, the application can filter out target trip nodes in the user's current trip that are in a pending state based on at least one of the following: product type, service time, and current system time in the current trip information. The application determines which nodes in the user's current trip are in a pending state based on the product type (e.g., airfare, train ticket, hotel, attraction ticket), service time (e.g., flight departure time, train departure time, hotel check-in time, scenic spot entry time), and current system time in the current trip information.

[0081] For example, if the application detects that a user has booked a flight from Shanghai to Sanya tomorrow morning, and the current time is more than 2 hours but less than 24 hours before the flight's departure time, and the user has not yet checked in, then the target itinerary node is determined to be "Pending Check-in." If the application detects that a user has booked a hotel in Sanya for the day after tomorrow, and the current time is more than 24 hours before the check-in time, then the target itinerary node is determined to be "Pending Check-in" or still in a normal waiting state. If the application detects that the user has already checked into a hotel and there are still 12 hours before the check-out time, then nodes such as "Pending Extension of Stay" or "Pending Review" can be displayed. If the user currently has no pending itinerary nodes (e.g., all itineraries have been completed), then the guidance prompt area can display "No pending itineraries" or recommend other travel inspiration and product information. Target itinerary nodes include at least one of the following: Pending Check-in, Pending Seat Selection, Pending Check-in, Pending Ticket Collection, Pending Ticket Inspection, Pending Boarding, Pending Transportation, Pending Airport Pickup, Pending Airport Drop-off, and Pending Review.

[0082] Furthermore, the application can obtain the node type of the target itinerary node, extract the itinerary card template corresponding to the node type from the pre-set itinerary card template library, and populate the itinerary card template with the product information from the current itinerary information to generate the itinerary card corresponding to the target itinerary node. The application has a pre-built itinerary card template library, in which a corresponding itinerary card template is defined for each node type. The template defines the node information fields to be displayed in the itinerary card (such as flight number, hotel name, time, location, etc.) and the type of node entry (such as button text, jump link, service call interface, etc.). After determining the target itinerary node, the application extracts the template matching the node type from the template library, and populates the specific product information from the current itinerary information into the corresponding fields of the template to generate a complete itinerary card. The itinerary card includes at least the node information of the target itinerary node (e.g., "You are about to take flight MU1234 from Shanghai to Sanya, departure time 14:30, you have not yet checked in") and the node entry (e.g., "Online Check-in" button, which allows users to jump to the check-in page or call the check-in service after clicking).

[0083] In practical applications, the specific content and access points of the itinerary card vary depending on the type of node. For example, when the target itinerary node is the check-in node, the itinerary card may include an online check-in portal and / or check-in deadline information, allowing users to directly check in online without manually navigating to the check-in page within the application; when the target itinerary node is the check-in node, the itinerary card may include an online check-in portal and / or hotel contact information, allowing users to directly check in online; when the target itinerary node is the ticket collection node, the itinerary card may include an e-ticket QR code and / or ticket collection location guidance information, allowing users to directly use the e-ticket QR code to enter the park or follow the guidance information to collect their tickets; when the target itinerary node is the review node, the itinerary card may include a review portal, allowing users to directly review completed itineraries or products. In this way, the itinerary card can centrally present the tasks that need to be handled in the user's current itinerary in the form of cards. Users do not need to repeatedly jump between different pages to find to-do items. They can obtain all the task information that needs to be handled at once in the guidance prompt area of ​​the AI ​​component and perform the operation. This realizes the centralized and visual management of itinerary node information and improves the efficiency of users in handling itinerary to-do items.

[0084] It should be noted that in practical applications, the pre-processing of the above-mentioned scenario labels—including the generation of itinerary tasks, the generation of the first guidance prompt information, the generation of the second guidance prompt information, and the generation of itinerary cards—can all be implemented using various technical paths, and this application does not impose any restrictions on them.

[0085] Specifically, in some implementations, the application can perform the aforementioned data processing locally on the electronic device after obtaining raw data such as the user's current itinerary information, dialogue content in the AI ​​dialogue session, user behavior data, and page context information. For example, the application can perform semantic parsing of the dialogue content locally using a pre-deployed lightweight natural language processing model to extract to-do items and decision-making items; locally, based on a pre-built rule engine and matching algorithm, it can determine the target scenario type based on page type identifiers and page content keywords, and filter matching guidance prompts from a locally cached guidance information library; locally, based on the product type, service time, and current system time in the itinerary information, it can filter the target itinerary nodes to be processed through pre-built itinerary node judgment logic, and extract the corresponding templates from a locally stored itinerary card template library and populate them to generate an itinerary card. The local processing method has advantages such as fast response speed, no network requests required, and protection of user data privacy, and is suitable for scenarios with relatively fixed processing logic and high real-time requirements.

[0086] In other implementations, after obtaining the aforementioned raw data, the application can send a data processing request to the server via the network, remotely invoking a Large Language Model (LLM) or other AI model deployed on the server to execute the data processing. For example, the application can concatenate the user's current itinerary information, the dialogue content in the AI ​​conversation, and page context information into a context-rich prompt, and send it to the LLM server. The LLM then performs deep semantic understanding and analysis of the dialogue content, extracting to-do items and decision-making information, and generating a structured itinerary task list. Alternatively, the application can use user behavior data such as search behavior, browsing behavior, adding items to cart, and favorites behavior, along with the current page context information, as input. Through remote invocation of a recommendation model or LLM, it can perform user intent analysis and interest inference to determine the user's current target product or destination and generate personalized secondary guidance prompts. For example, the application can send the user's current itinerary information and current system time to the server. The server then uses an AI model to comprehensively determine the urgency and processing priority of each itinerary node, selects the most suitable target itinerary node for current presentation, and extracts a template from the server's itinerary card template library to generate an itinerary card, which is then returned to the client for presentation. Remotely calling large language models or AI models has stronger semantic understanding capabilities, richer knowledge reserves, and stronger generalization capabilities, enabling it to handle more complex dialogue scenarios and more personalized generation needs. It is especially suitable for scenarios with complex and varied dialogue content that require a deep understanding of user intent.

[0087] In other implementations, applications can combine local processing with remote calls to achieve the aforementioned data processing. For example, the application can pre-deploy a lightweight model and rule engine locally to handle itinerary task generation and guidance prompt information filtering in routine scenarios, ensuring response speed and offline availability. Simultaneously, when the local model has low confidence or encounters complex scenarios, the application can initiate remote calls to the server's large language model via the network to obtain higher-quality processing results. The results returned remotely are then used to update the local model and cached data, achieving a dynamic balance between local processing efficiency and remote processing capabilities. Through the flexible selection and combination of these various technical implementation paths, the application can fully leverage the semantic understanding and generation capabilities of advanced AI technologies such as large language models while ensuring response speed and user experience. This allows it to provide users with accurate and personalized itinerary tasks, guidance prompts, and itinerary cards, further enhancing the intelligence level and user experience of AI-assisted functions in the travel service platform.

[0088] Furthermore, the application can create corresponding AI dialogue sessions in response to triggered actions related to a target task, a first guidance prompt, a second guidance prompt, or a travel itinerary card. Specifically, when a user clicks on a target task, a first guidance prompt, a second guidance prompt, or a travel itinerary card displayed in the guidance prompt area, the application creates an AI dialogue session based on preset input information or operation types associated with the clicked content. For example, after a user clicks on the first guidance prompt "Help me plan a 7-day self-driving tour around Hainan Island," the application automatically sends a request to the AI ​​with this guidance prompt as input information, creating a corresponding AI dialogue session. The AI ​​system can generate personalized travel planning suggestions based on information such as the user's current city and travel time preferences. After a user clicks on the second guidance prompt "Book the Sheraton Sanya Haitang Bay Hotel," the application automatically creates an AI dialogue session with this booking request as input information. The AI ​​system can query the hotel's real-time room availability and pricing information and guide the user through the booking process. After a user clicks the "Online Check-in" button in the "Pending Check-in" itinerary card, the application can create a corresponding AI dialogue session. The AI ​​system can guide the user through the check-in process, including selecting a seat and confirming baggage allowance. After a user clicks the "Confirm Sanya Return Flight Time" pending decision task, the application creates a corresponding AI dialogue session. The AI ​​system can query the user's currently booked return flight information and provide alternative options for the user's reference.

[0089] Based on the aforementioned process, the application can display an AI dialogue window corresponding to the AI ​​conversation within the current traditional function page. Specifically, after creating an AI conversation, the application can display the AI ​​dialogue window within the current traditional function page. This AI dialogue window can be presented as a floating layer, a half-screen panel, or a full-screen panel, and its display position and size can be adapted to the layout of the current traditional function page and the user's operating habits. Furthermore, the AI ​​dialogue window displays the dialogue content associated with the AI ​​conversation, namely the AI ​​system's response to the user's triggered content and the user's subsequent input. Through this process, users can quickly initiate various types of AI conversations with minimal cognitive burden using the functional scenario tags and guidance prompts in the third component form, without manually entering lengthy descriptions, further enhancing the usability and user experience of AI functions in complex travel scenarios.

[0090] In some examples, the interaction method of the travel service application provided in this application embodiment is illustrated through a complete application scenario: taking the scenario where a user enters the product details page from the homepage, initiates an AI dialogue, and then switches pages as an example. Specifically: After a user opens the homepage of a travel service app, the app obtains the page context information of the homepage, determines the page type as "Homepage," and determines the third component form as the target component form matching the homepage according to the matching rules. It then renders an AI component (complete panel form) in the third component form on the homepage, containing a functional category area (including three functional scenario tags: "Help Me Think," "Help Me Book," and "Help Me Apply") and a guidance prompt area (including guidance prompts associated with each tag). When the user clicks the "Help Me Think" tag in the AI ​​component on the homepage, the app responds to this selection by obtaining the user's current itinerary information (at this time, the user has no booked itineraries), extracting to-do items and decision-making items from the existing dialogue content in the AI ​​dialogue session (at this time, there is no dialogue content), and using the homepage's page context information to determine the target scenario type of the current traditional function page as "Itinerary Planning Scenario." It then extracts candidate guidance prompts matching "Itinerary Planning Scenario" from a preset guidance information library and generates first guidance prompts such as "May Day Family Trip Recommendation" which are displayed in the guidance prompt area. When a user clicks on the "May Day Family Trip Recommendations" prompt, the application creates a corresponding AI dialogue session and displays the AI ​​dialogue window. The AI ​​system generates recommended itineraries based on the user's location and travel preferences. After closing the AI ​​dialogue window, the user continues browsing the homepage and clicks on a recommended product to enter the product details page.

[0091] Furthermore, after detecting a page switching event, the application obtains the page context information of the product details page, determines that the page type is a product details page, and calculates the page information density value of the current product details page (e.g., 0.82). If the first preset threshold is 0.7, then the current page information density value is greater than the first preset threshold, and the application determines the first component form as the target component form matching the product details page according to the matching rules. At this time, the component form of the AI ​​component switches from the third component form to the first component form (i.e., a floating bubble form containing only a global floating trigger icon). During the switching process, the application passes the session identifier and dialogue context of the "May Day Family Trip Recommendation" AI dialogue session previously initiated by the user on the homepage to the AI ​​component in the switched first component form, so that when the user clicks the Q icon on the current product details page to re-revive the AI ​​dialogue window, they can continue the previous dialogue without having to re-describe their needs.

[0092] On the product details page, if the user clicks the Q icon again, the application obtains the time difference between the last time the AI ​​component in the first component form was collapsed and the current time. Assuming the user closed the AI ​​dialogue window after completing the conversation on the homepage, and 8 minutes have passed since then (with a preset time threshold of 5 minutes), if 8 minutes is greater than 5 minutes, the application switches the AI ​​component from the first component form to the third component form (full panel form). After switching to the third component form, the application obtains the user's browsing behavior data on the current product details page (such as the user spending a long time on the page, viewing room images multiple times, etc.), determines that the user's current target product is the hotel, and filters out relevant guidance prompts from a preset guidance information library, generating second guidance prompts such as "Does this hotel have family rooms?" and "Are there any family-friendly attractions near the hotel?" which are displayed in the guidance prompt area. When the user clicks on one of the guidance prompts, the application creates a new AI dialogue session or continues the dialogue based on an existing one, displaying the AI ​​dialogue window on the product details page. The user can receive personalized recommendations and answers from the AI ​​without leaving the product details page. Through the above process, the AI ​​component achieves adaptive form switching and session persistence between different functional pages, ensuring that the AI-assisted service remains continuously available throughout the user's browsing and decision-making process from the homepage to the product details page.

[0093] Furthermore, in some feasible implementations, after the AI ​​dialogue session has been created and the AI ​​dialogue window has been displayed on the current traditional function page, the method may also include session storage and restoration operations, optionally: The application can respond to closing actions on the AI ​​dialogue window (e.g., clicking the close button in the upper right corner of the dialogue window or performing a downward swipe gesture) by collapsing the AI ​​dialogue window on the current traditional function page and displaying a conversation shard control associated with the AI ​​dialogue window on the same page. This shard control also displays dialogue information associated with the AI ​​dialogue session (e.g., a summary of the latest message, the session creation time, key information from the dialogue, etc.). The conversation shard control is a floating area on the current traditional function page that presents the dialogue content of the AI ​​dialogue window in summary form. Next, if the user wants to restore the AI ​​dialogue window, they can enter a trigger action on the conversation shard control. The application, responding to this trigger action (e.g., clicking the control), reopens the AI ​​dialogue window on the current traditional function page and continues displaying the dialogue content of the AI ​​dialogue session.

[0094] In practical applications, the conversation pause control can be a small floating card or bubble displayed at the edge or corner of the current traditional function page, briefly displaying a summary of the current AI dialogue session, such as "Inquiries about Sanya hotels" or a portion of the last message. This conversation pause control and the global floating trigger icon of the AI ​​component can coexist independently or be displayed collaboratively. Users can quickly restore previously closed dialogue windows by clicking the control, without needing to re-initiate the dialogue through the AI ​​component.

[0095] In the above process, the session storing control allows users to retain visual cues of the session context even after closing the AI ​​dialogue window. When needed, users can quickly restore the dialogue window by triggering the control, without having to re-initiate AI interactions or re-search the dialogue history, further enhancing the continuity of interaction across sessions. Simultaneously, the session storing control can be displayed in conjunction with traditional function pages in the graphical user interface (e.g., embedded, floating, parallel display), without occupying excessive screen space and avoiding the cognitive burden of reopening the full dialogue window. Users can flexibly switch between browsing page content and resuming AI dialogues at any time. Applications can maintain multiple session storing controls simultaneously (for example, users may initiate multiple independent AI dialogues on different product detail pages), with each control corresponding to an independent AI dialogue session. Users can switch between different sessions by clicking different controls.

[0096] In other feasible implementations, when the first AI component is in the form of a third component (a complete panel), the application can also provide a task control as an entry point for initiating another AI dialogue session. Optionally, in response to a trigger operation on the task control, the application can display a task window (which can be an independent window presented as a pop-up, overlay, or half-screen panel). The task window contains at least one schedule task, which includes at least a task parsed from the dialogue content in the AI ​​dialogue session (for example, if the user said "I want to book a hotel in Sanya" in the dialogue, the application can parse the to-do task "Book a hotel in Sanya" from the dialogue content), and a task generated based on the schedule node of the user's current schedule (for example, if the application detects that the user has booked a flight to Sanya but has not yet booked a hotel, it can generate the to-do task "Book a hotel in Sanya"); in response to a trigger operation on the schedule task, a corresponding AI dialogue session is created; and the AI ​​dialogue window corresponding to the AI ​​dialogue session is displayed on the current traditional function page. Through the task control, the application can automatically generate a list of to-do tasks based on the user's conversation content and schedule status, and automatically create a corresponding AI dialogue session when the user clicks on a task. This allows users to quickly obtain targeted AI assistance without having to manually enter a task description, further reducing the user's operational burden.

[0097] In some application scenarios, let's take the conversation staking function as an example. A user initiates an inquiry about pet-free policies at a hotel through an AI component on a product details page. After receiving a response from the AI, the user wants to temporarily close the conversation window to continue browsing other information on the product details page, so they click the close button in the upper right corner of the AI ​​conversation window. Upon detecting this close operation, the application collapses the AI ​​conversation window on the current traditional function page and displays a conversation staking control in the lower right or lower left corner of the current traditional function page. This control displays a summary of the inquiry about pet policies and a summary of the latest message in the conversation (e.g., "Hello, according to our inquiry, this hotel allows pets weighing no more than 10kg, but an additional cleaning fee is required"). After the user continues browsing other content on the product details page, they click the conversation staking control, and the application reopens the AI ​​conversation window on the current traditional function page, displaying the complete history of the AI ​​conversation. The user can then enter new questions in this window (e.g., "How much is the pet cleaning fee?"), and the AI ​​system can respond based on the previous conversation context.

[0098] It should be noted that the embodiments of this application include, but are not limited to, the examples described above. It is understood that, under the guidance of the ideas in the embodiments of this application, those skilled in the art can also make further processing according to actual needs, and this application does not impose any restrictions on this.

[0099] In this embodiment, the travel service application displays a first AI component with a target component form matching the current traditional function page. This allows the AI ​​component to be presented in a component form adapted to the current scenario within the current traditional function page, improving the flexibility and scenario adaptability of the AI ​​entry point presentation. In response to a component form change command or page switching event for the first AI component, the application switches the first AI component to a second AI component with a different component form. This enables flexible switching between different component forms of the AI ​​component. Furthermore, if an AI dialogue session exists, the first and second AI components share this AI dialogue session. This allows users to continue interacting within the same AI dialogue session even after a component form change or page switching, without needing to worry about component changes. By switching forms or jumping between pages to re-initiate AI interactions or redescribe needs, a comprehensive mechanism combining component form switching and session sharing is achieved. This not only enables flexible form switching of AI components under different scenario requirements, reducing the obstruction of page content and interference with user operations by AI entry points, but also ensures the continuity of AI dialogue sessions across forms before and after component form changes or page switching. Furthermore, by sharing the same AI dialogue session before and after switching, users can continue the interaction progress of the same AI dialogue in different component forms. The overall design of the above-mentioned solution in this application improves the adaptability of AI interaction in different scenarios and the service continuity across pages and forms in travel service applications, reducing the operational burden and decision-making costs for users in the process of travel planning and decision-making.

[0100] To enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the following examples are provided for illustrative purposes: As an example, refer to Figure 2The diagram illustrates an AI component provided in an embodiment of the present invention. The first component-type AI component may consist of only a single global floating trigger icon, or, in a specific function page, may consist of a global floating trigger icon and corresponding prompts / guidance information. The second component-type AI component can present different forms depending on the type of function page. Regardless of the form, the second component-type AI component can include corresponding input controls, and in the corresponding function page, it can also display corresponding prompts / guidance information, while providing a corresponding close control so that the user can switch the second component-type AI component back to the first component-type AI component. The third component-type AI component includes a receiver... The AI ​​component has two forms: collapsed and expanded. In collapsed form, the AI ​​component in the third form can contain only the corresponding functional scenario labels (such as "Help me think", "Help me order", "Help me do it" etc.) and the corresponding task controls (which can display the number of tasks to be viewed, etc.). In expanded form, the AI ​​component in the third form can be divided into a function allocation area and a guidance prompt area. The function allocation area displays functional scenario labels and task controls, while the guidance prompt area displays at least one of the following based on the currently selected target functional scenario label: the corresponding trip task, guidance prompt information, and trip card. In addition, the AI ​​component in the third form also includes input controls to provide users with complete component functionality.

[0101] Accordingly, refer to Figure 3 This diagram illustrates an application interface provided in an embodiment of this application. A travel service application can run on an electronic device, and a graphical user interface displays the traditional function page currently being viewed by the user. Figure 3 From left to right, the layout consists of the homepage, message page, trip details page, and personal center page. The AI ​​component can be presented in a way that adapts to the current scenario within the existing traditional function pages. For example, on the homepage, the AI ​​component can be displayed as a third component; on the message page, it can be displayed as a first component; on the trip details page, it can be displayed as a second component; and on the personal center page, it can be displayed as a third component, etc. This application does not impose any restrictions on this.

[0102] Furthermore, in Figure 3 Building upon this, for the globally floating trigger icons that are permanently displayed on each functional page, clicking these icons will switch the component's form. For details, refer to... Figure 4For example, on the homepage, in the third display mode, when the user clicks the icon, the application can switch the collapsed AI component to an expanded AI component; on the messages page, in the first display mode, when the user clicks the icon, the application can switch the component from the first display mode to the third display mode; on the trip details page, in the second display mode, when the user clicks the icon, the application can switch the component from the second display mode to the third display mode; on the personal center page, in the third display mode, when the user clicks the icon, the application can switch the collapsed AI component to an expanded AI component, and so on. It should be noted that... Figure 4 The process of switching from other display modes to the third display mode is shown. It is understood that the third display mode can also be switched to the second or first display mode. For example, by inputting the corresponding collapse swipe operation or clicking the icon again, the relevant process can be referred to the description in the foregoing embodiments, and will not be repeated here.

[0103] Furthermore, referring to Figure 5 This document illustrates a flowchart of the AI ​​conversation processing provided in an embodiment of the present invention. When the AI ​​component is displayed as a third component, it can display corresponding guidance prompts based on the page context, such as: "Departing from Ningbo to Xi'an in early June, there are discounts on airfare and hotels, would you like to check the remaining tickets?", "Chongqing is about to reach 40 degrees Celsius, I suggest going to Wutai Mountain in Shanxi to escape the heat and enjoy the cool 20-degree mountain breeze.", and "I've been looking at hotels and surrounding tours in Guiyang lately? Late May is a perfect time to escape the heat in Guizhou, check out these hotels near the subway...". When a user clicks on the prompt "Departing from Ningbo to Xi'an in early June, there are discounts on airfare and hotels, would you like to check the remaining tickets?", the application can create a corresponding AI conversation session, display the AI ​​conversation window corresponding to that session, and show the corresponding conversation content within that window. Next, if the user wants to exit the AI ​​dialogue window, they can input the corresponding swipe-down operation or click the corresponding collapse control. The application can respond to the user's operation, collapse the AI ​​dialogue window while retaining the AI ​​dialogue session, and display the corresponding session storage control 510, as well as the corresponding dialogue message, such as "Which day is the cheapest flight from Hangzhou to Beijing?", within the session storage control 510. In addition, a corresponding close control 520 can be provided within the session storage control 510. The user can close the session storage control 510 through this close control 520. When the application closes the display of the session storage control 510, it can also close or destroy the corresponding AI dialogue session, and then redisplay the AI ​​component in the appropriate component form according to the current page type.

[0104] Furthermore, when the AI ​​component is displayed as a third component, refer to Figure 6 When a user selects the "Help Me Think" tab, the application expands the AI ​​component and displays guidance information associated with the "Help Me Think" tab, such as "Ningbo to Xi'an departing in early June, there are discounts on airfare and hotels, do you want to check the remaining tickets?", "Chongqing will soon reach 40℃, I suggest going to Wutai Mountain in Shanxi to escape the heat", and "I've been looking at hotels and surrounding tours in Guiyang lately?". In addition, when the "Help Me Think" tab is selected, the AI ​​component can also display the corresponding travel task, such as "Seat reserved: Xiaofei (38A), check-in after confirmation---Pending decision>", indicating that there is a travel task node that needs to be processed.

[0105] Accordingly, refer to Figure 7 When a user selects the "Book for me" tab, the application can expand the AI ​​component and display corresponding guidance prompts, such as "You are about to go to Altay, do you want me to book the high-speed train from Hangzhou for you?", "The Hangzhou to Beijing route is in high demand, let me check for discounted tickets for you", "Disney summer parade, let me book a day trip for you", and so on.

[0106] Accordingly, refer to Figure 8 When a user selects the "Help me with this" tab, the application can expand the AI ​​component and display the corresponding itinerary card if the user's current itinerary information is detected. For example, if the user has a check-in node, the AI ​​component can display an itinerary card containing flight information, which includes at least: flight information, check-in counter, boarding gate, and check-in trigger entry.

[0107] In the above process, the comprehensive mechanism combining component form switching and session sharing not only enables flexible form switching of AI components under different scenario requirements, reducing the obstruction of page content and interference with user operations by AI entry points, but also ensures the continuity of AI dialogue sessions across forms before and after component form changes or page switching. At the same time, by sharing the same AI dialogue session before and after switching, users can continue the interaction progress of the same AI dialogue in different component forms. This improves the adaptability of AI interaction in different scenarios and the service continuity across pages and forms in travel service applications, reducing the operational burden and decision-making costs for users in the process of travel planning and decision-making.

[0108] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of this application are not limited to the described order of actions, because according to the embodiments of this application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of this application.

[0109] In addition, this application also provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the various processes of the above-described interactive method embodiment of the travel service application and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0110] This application also provides a readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described travel service application interaction method embodiment and achieves the same technical effect. To avoid repetition, it will not be described again here. The readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0111] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0112] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, EEPROM, Flash, and eMMC, etc.) containing computer-usable program code.

[0113] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0114] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0115] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0116] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0117] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0118] The interactive method and interactive device for a travel service application provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. An interaction method for a travel service application, characterized in that, The method of providing a graphical user interface via an electronic device, wherein the content presented by the graphical user interface includes at least the currently used traditional function pages of the travel service application that the user is currently browsing, includes: The first AI component displays the target component form that matches the current traditional function page; In response to a component form change instruction or a page switching event for the first AI component, the first AI component is switched to a second AI component with another component form; If an AI dialogue session exists, the first AI component and the second AI component share the AI ​​dialogue session.

2. The method according to claim 1, characterized in that, The first AI component, which displays a target component form that matches the current traditional function page, includes: Obtain the first page context information corresponding to the current traditional function page; Extract at least one page parameter from the first page context information, wherein the first page parameter includes at least one of the first page type and the first page information density value; Based on at least one of the first page type and the first page information density value, a first target component form matching the current traditional function page is determined from a preset component form library; Render the first AI component corresponding to the target component form in the current traditional function page.

3. The method according to claim 2, characterized in that, The component form library includes at least one of a first component form, a second component form, and a third component form; wherein: The first component form corresponds to the first size parameter and includes a global floating trigger icon, or the global floating trigger icon and guidance prompt information; The second component shape corresponds to the second size parameter and includes an input control, which includes at least one of a voice input control, a text input box, an image input control, and an extended control. The third component form corresponds to the third size parameter and includes a functional classification area and a guidance prompt area. The functional classification area includes at least one functional scene label and a task control. The guidance prompt area displays one of the following: task information, guidance prompt information, and itinerary card associated with the currently selected functional scene label. Wherein, the first dimension parameter is smaller than the second dimension parameter, and the second dimension parameter is smaller than the third dimension parameter.

4. The method according to claim 3, characterized in that, The step of determining a first target component form matching the current traditional functional page from a preset component form library based on at least one of the first page type and the first page information density value includes: If the page type is a message page, then the first component form will be used as the first target component form; If the page type is one of a product details page, a product details page, or an order details page, and the page information density value is greater than a first preset threshold, then the first component form will be used as the first target component form. If the page type is one of a list page, a search results page, or a trip planning page, then the second component form will be used as the first target component form. If the page type is a homepage or a personal center page, then the third component form will be used as the first target component form; If the page type does not belong to the preset page type set, the first target component form is determined based on the page information density value, including: if the page information density value is greater than the first preset threshold, the first component form is taken as the first target component form; if the page information density value is less than or equal to the first preset threshold and greater than the second preset threshold, the second component form is taken as the first target component form; if the page information density value is less than or equal to the second preset threshold, the third component form is taken as the first target component form, where the second preset threshold is less than the first preset threshold.

5. The method according to claim 3, characterized in that, The second AI component, which switches the first AI component to another component form in response to a component form change instruction for the first AI component, includes: When the first AI component is an AI component of the first component form, in response to the trigger operation for the global floating trigger icon, the display of the global floating trigger icon in the graphical user interface is maintained, and the input control of the AI ​​component of the second component form, or the functional classification area of ​​the AI ​​component of the third component form, or the functional classification area and guidance prompt area of ​​the AI ​​component of the third component form are displayed in the preset adjacent area of ​​the global floating trigger icon. When the first AI component is an AI component in the form of the second component, in response to the form compression instruction for the AI ​​component in the form of the second component, the input control is hidden; or, in response to the form expansion instruction for the AI ​​component in the form of the second component, the input control is hidden, and the functional classification area of ​​the AI ​​component in the form of the third component, or the functional classification area and guidance prompt area of ​​the AI ​​component in the form of the third component are displayed in a preset adjacent area of ​​the global floating trigger icon. When the first AI component is an AI component in the form of the third component, in response to the form compression instruction for the AI ​​component in the form of the third component, the function classification area is hidden, or the function classification area and the guidance prompt area are hidden, and the input control is displayed in a preset adjacent area of ​​the global floating trigger icon, or only the global floating trigger icon is displayed. The form expansion command is any one of the expansion operation or the upward swipe operation applied to the currently displayed AI component; the form compression command is any one of the collapse operation, the downward swipe operation, or the click operation on the blank area outside the component applied to the currently displayed AI component.

6. The method according to claim 3, characterized in that, The step of switching the first AI component to a second AI component of another component form in response to a page switching event for the first AI component includes: In response to switching the current traditional function page to the target traditional function page, the second page context information of the target traditional function page is obtained; Extract the second page parameters from the second page context information, wherein the second page parameters include at least the second page type and / or the second page information density value; Based on the second page type and / or the second page information density value, determine a second target component form that matches the target traditional functional page from the component form library; If the form of the first target component is the same as that of the second target component, then the form of the first AI component is maintained in the target traditional function page; If the first target component form is inconsistent with the second target component form, then in the target traditional function page, the first AI component is switched to the second AI component corresponding to the second target component form.

7. The method according to claim 3, characterized in that, When the first AI component is in the first component form, before the step of sharing the AI ​​dialogue session between the first AI component and the second AI component if an AI dialogue session currently exists, the method further includes: In response to a trigger operation on the global floating trigger icon in the first AI component, the first AI component is switched to an AI component of the second component form in the current traditional function page, or the first AI component is switched to an AI component of the third component form. If the AI ​​component in the second component form contains at least one guidance prompt message, then in response to a trigger operation for the guidance prompt message, an AI dialogue session corresponding to the guidance prompt message is created; or, in response to a trigger operation for the input control in the AI ​​component in the second component form, the target input information corresponding to the trigger operation is obtained, and an AI dialogue session corresponding to the target input information is created. If the AI ​​component in the second component form does not contain guidance prompt information, then in response to the trigger operation of the input control in the AI ​​component in the second component form, the target input information corresponding to the trigger operation is obtained, and an AI dialogue session corresponding to the target input information is created; Alternatively, in response to a triggering operation on the task information, the guidance prompt information, or the itinerary card in the AI ​​component of the third component form, an AI dialogue session corresponding to the task information, the guidance prompt information, or the itinerary card is created; In addition, an AI dialogue window corresponding to the AI ​​dialogue session is displayed on the current traditional function page, and the dialogue content associated with the AI ​​dialogue session is displayed in the AI ​​dialogue window.

8. The method according to claim 3, characterized in that, When the first AI component is in the form of the third component, before the step of sharing the AI ​​dialogue session between the first AI component and the second AI component if an AI dialogue session currently exists, the method further includes: Display the first scene label, the second scene label, and the third scene label in the first AI component; In response to selecting the first scene tag, at least one trip task to be decided or executed, and / or at least one first guidance prompt related to travel inspiration or trip planning is displayed in the guidance prompt area. The trip task includes at least the task obtained by parsing the dialogue content in the AI ​​dialogue session, and the task generated according to the trip node of the user's current trip. Alternatively, in response to selecting the second scene label, at least one second guidance prompt message is displayed in the guidance prompt area, wherein the second guidance prompt message is a prompt message generated based on user behavior; Alternatively, in response to selecting the third scenario label, a travel card corresponding to the target travel node in the user's current itinerary is displayed in the guidance prompt area. The travel card includes at least the node information and node entry of the target travel node. The target travel node includes at least one of the following: check-in node, seat selection node, check-in node, ticket collection node, ticket inspection node, boarding node, boarding node, airport pick-up node, airport drop-off node, and evaluation node. In response to a triggering operation on the target task, the first guidance prompt, the second guidance prompt, or the itinerary card, a corresponding AI dialogue session is created; The AI ​​dialogue window corresponding to the AI ​​dialogue session is displayed on the current traditional function page.

9. The method according to claim 8, characterized in that, Before displaying at least one trip task to be decided or executed, and / or at least one first guidance prompt related to travel inspiration or trip planning, in the guidance prompt area, the method further includes: Obtain the user's current travel information; Extract the information on to-do items and / or pending decision items contained in the dialogue content of the AI ​​dialogue session, and / or identify the associated trip nodes in the current trip information; Generate the to-do list information, and / or, the decision-making information, and / or, at least one decision-making or execution task corresponding to the trip node; And / or, using the first page context to determine the target scene type to which the current traditional function page belongs, and extracting at least one candidate guidance prompt information that matches the target scene type from a preset guidance information library, generating at least one first guidance prompt information related to travel inspiration or itinerary planning.

10. The method according to claim 8, characterized in that, Before displaying at least one second guidance prompt message in the guidance prompt area, the method further includes: Acquire user behavior data in the travel service application, the user behavior data including at least one of search behavior, browsing behavior, add-to-cart behavior, and favorites behavior; Based on the user behavior data, determine the target product or target destination that the user is currently interested in; At least one guidance prompt associated with the target product or the target destination is selected from a preset guidance information database and used as the second guidance prompt information.

11. The method according to claim 8, characterized in that, Before displaying the itinerary card corresponding to the target itinerary node in the user's current itinerary in the guidance prompt area, the method further includes: Obtain the user's current travel information, which includes at least one of the following: booked transportation products, hotel accommodation products, and / or scenic spot tour products; Based on at least one of the product type, service time, and current system time in the current trip information, filter out the target trip nodes in the user's current trip that are in a pending state; Obtain the node type of the target trip node, extract the trip card template corresponding to the node type from the preset trip card template library, and fill the product information in the current trip information into the trip card template to generate the trip card corresponding to the target trip node.

12. The method according to claim 3, characterized in that, When the first AI component is in the form of the third component, before the step of sharing the AI ​​dialogue session between the first AI component and the second AI component if an AI dialogue session currently exists, the method further includes: In response to a trigger operation on the task control, a task window is displayed, the task window containing at least one schedule task, the schedule task including at least a task parsed from the dialogue content in the AI ​​dialogue session, and a task generated based on the schedule node of the user's current schedule; In response to a triggered operation for the aforementioned task, a corresponding AI dialogue session is created; The AI ​​dialogue window corresponding to the AI ​​dialogue session is displayed on the current traditional function page.

13. The method according to any one of claims 7-12, characterized in that, After displaying the AI ​​dialogue window corresponding to the AI ​​dialogue session on the current traditional function page, the method further includes: In response to the closing operation of the AI ​​dialogue window, the AI ​​dialogue window is collapsed on the current traditional function page, and a session storage control related to the AI ​​dialogue window is displayed on the current traditional function page, and dialogue information associated with the AI ​​dialogue session is displayed in the session storage control. In response to a trigger operation on the session storage control, the AI ​​dialogue window is reopened on the current traditional function page, and the dialogue content of the AI ​​dialogue session continues to be displayed in the AI ​​dialogue window.

14. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory is used to store computer programs; When the processor executes a program stored in the memory, it implements the method as described in any one of claims 1-13.

15. A readable storage medium having instructions stored thereon that, when executed by one or more processors, cause the processors to perform the method as described in any one of claims 1-13.

Images