User interface generation methods, computing systems, and storage media

Abstract

This specification provides a method, computing system, and storage medium for generating a user interface. In the method, the computing system receives an interface generation request from a client. This request includes target text described by an operator in natural language. An intelligent agent group then performs an interface generation task based on the target text and a semantic knowledge base to obtain at least one user interface. The semantic knowledge base includes predefined information for multiple interface elements, each comprising multiple interface components and multiple visual attributes. The predefined information for each interface element characterizes its style and usage description.

Description

Technical Field

[0001] This specification relates to the field of Internet technology, and in particular to a method for generating a user interface, a computing system, and a storage medium. Background Technology

[0002] With the development of internet technology, applications (APPs) have become important tools for people to obtain information, enjoy services, and conduct related activities. APPs provide users with an interactive environment through interactive pages to meet their interactive needs. These interactive pages are not only windows for APPs to showcase their platform's products or services, but also present the features and advantages of these products or services to users, achieving marketing and promotional goals.

[0003] Typically, the interactive pages displayed in an app, as well as the display cards for specific booths within those pages, rely primarily on manual design. Designers use their professional skills and experience to conceive the overall layout and visual style of the interactive pages or display cards from scratch, which is time-consuming, energy-intensive, and inefficient.

[0004] The information in the background section is merely information known only to the inventor and does not imply that such information had entered the public domain before the date of this application, nor does it imply that it can be considered prior art in this disclosure. Summary of the Invention

[0005] This specification provides a method for generating user interfaces, a computing system, and a storage medium. It can achieve the systematization and automation of interface generation through intelligent agent groups, thereby improving the efficiency of interface generation. In addition, when generating interfaces, the intelligent agent groups refer to the preset interface components and visual attributes in the knowledge base, so that the generated interfaces can be personalized within a specific range, avoiding unlimited randomness and improving the stability and standardization of interface generation.

[0006] In a first aspect, this specification provides a method for generating a user interface, the method comprising: receiving an interface generation request from a client, the interface generation request including target text described by an operator in natural language; performing an interface generation task by an intelligent agent group based on the target text and a semantic knowledge base to obtain at least one user interface, the semantic knowledge base including predefined information of multiple interface elements, the multiple interface elements including multiple interface components and multiple visual attributes, the predefined information of each interface element representing the style and usage description of the interface element, and the intelligent agent group including at least one intelligent agent.

[0007] In some embodiments, the interface generation task includes multiple sub-tasks, and the agent group includes multiple dedicated agents that perform different sub-tasks.

[0008] In some embodiments, the plurality of subtasks include: understanding the target text and converting it into interface requirements; generating at least one interface template to be filled based on the interface requirements; and filling each of the interface templates to obtain the at least one user interface.

[0009] In some embodiments, the plurality of dedicated intelligent agents include a first intelligent agent, a second intelligent agent, and a third intelligent agent. The step of performing an interface generation task based on the target text and a semantic knowledge base by the group of intelligent agents to obtain at least one user interface includes: understanding the target text and determining at least one target component from the plurality of interface components by the first intelligent agent; generating at least one interface template to be filled based on the at least one target component by the second intelligent agent, each interface template including a layout structure and a visual theme; and filling each interface template based on the target text by the third intelligent agent to obtain the at least one user interface.

[0010] In some embodiments, the step of understanding the target text and determining at least one target component from the plurality of interface components by the first agent includes: executing the following by the first agent: generating first guidance information based on the target text, wherein the first guidance information is used to guide a first large language model to extract information to be displayed from the target text and recommend interface components that can carry the information to be displayed based on the usage descriptions of each interface component in the semantic knowledge base; and inputting the first guidance information into the first large language model and determining the at least one target component based on the output of the first large language model.

[0011] In some embodiments, generating at least one interface template to be filled based on the at least one target component by the second agent includes: executing the following by the second agent: obtaining a preset layout rule for the interface component; generating at least one layout structure based on the preset layout rule and the at least one target component; selecting styles of at least some visual attributes to obtain the visual theme; and generating the at least one interface template based on the at least one layout structure and the visual theme.

[0012] In some embodiments, generating at least one layout structure based on the preset layout rules and the at least one target component includes: generating second guidance information based on the preset layout rules and the at least one target component, wherein the second guidance information is used to guide a second language model to arrange and combine the at least one target component according to the preset layout rules to generate a preset number of layout structures; and inputting the second guidance information into the second language model, and determining the at least one layout structure based on the output of the second language model.

[0013] In some embodiments, selecting at least some visual attributes to obtain the visual theme includes: generating third guidance information based on the target text, the third guidance information being used to guide a third language model to recommend visual attributes and their styles that are adapted to the target text based on the style and usage description of each visual attribute in the semantic knowledge base; and inputting the third guidance information into the third language model, and determining the visual theme based on the output of the third language model.

[0014] In some embodiments, selecting styles of at least some visual attributes to obtain the visual theme includes: obtaining visual requirements specified by the operator; and selecting styles of the at least some visual attributes based on the visual requirements to obtain the visual theme.

[0015] In some embodiments, the usage descriptions corresponding to at least some interface components include usage specifications. The step of filling each interface template with the target text by the third agent to obtain the at least one user interface includes: the third agent performing the following: for each interface template, obtaining the target usage specifications corresponding to each target component in the interface template from the semantic knowledge base; generating the content to be filled for each target component in the interface template based on the target text and the target usage specifications; and filling each target component in the interface template with the content to be filled for each target component to obtain the user interface.

[0016] In some embodiments, the interface template includes a text component, and the content to be filled in the text component is obtained in the following manner: generating fourth guidance information based on the text component and the target text, the fourth guidance information being used to guide a fourth language model to generate content to be filled that conforms to the usage specifications of the text component based on the target text; and inputting the fourth guidance information into the fourth language model, and determining the content to be filled in the text component based on the output of the fourth language model.

[0017] In some embodiments, the interface template includes an image component, the content to be filled in the image component is obtained by: generating an image generation instruction based on the target text, the visual theme of the interface template and the usage specifications of the image component; and calling an image generation model based on the image generation instruction to generate the content to be filled in the image component.

[0018] In some embodiments, the plurality of interface components include a plurality of atomic components and at least one molecular component, each molecular component including at least two atomic components.

[0019] In some embodiments, the usage description corresponding to each interface element includes at least one of the following dimensions: the characteristics of the interface element; the applicable scenarios of the interface element; or the usage specifications of the interface element.

[0020] In some embodiments, the method further includes: sending the at least one user interface to the client, wherein the at least one user interface is in the format of at least one of the following: an interface image; executable code; or editable data.

[0021] Secondly, this specification also provides a computing system, comprising: at least one storage medium storing at least one instruction set; and at least one processor communicatively connected to the at least one storage medium, wherein the at least one processor reads the at least one instruction set during operation and executes the method described in the first aspect according to the instructions of the at least one instruction set.

[0022] Thirdly, this specification provides a computer-readable non-transitory storage medium, wherein the computer-readable non-transitory storage medium stores at least one instruction set, which, when executed by at least one processor, implements the method as described in the first aspect.

[0023] The methods for generating the user interface, the computing system, and the storage medium provided in this specification are partially listed in the following description. The inventive aspects of the methods for generating the user interface, the computing system, and the storage medium provided in this specification can be fully explained through practice or by using the methods, systems, and combinations described in the detailed examples below. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A schematic diagram of an application scenario generated by an interface provided according to an embodiment of this specification is shown; Figure 2 A hardware structure diagram of a computing system provided according to an embodiment of this specification is shown; Figure 3 A flowchart illustrating a method for generating a user interface according to an embodiment of this specification is shown; Figure 4 A schematic diagram of a molecular component provided according to an embodiment of this specification is shown; Figure 5A flowchart illustrating the execution of subtasks by various agents according to embodiments provided in this specification is shown; Figure 6 A schematic diagram of a user interface provided according to embodiments of this specification is shown; and Figure 7 A schematic diagram illustrating a user interface presented to a user according to an embodiment of this specification is shown. Detailed Implementation

[0026] The following description provides specific application scenarios and requirements for this specification, intended to enable those skilled in the art to make and use the contents of this specification. Various partial modifications to the disclosed embodiments will be apparent to those skilled in the art, and the general principles defined herein can be applied to other embodiments and applications without departing from the spirit and scope of this specification. Therefore, this specification is not limited to the embodiments shown, but rather to the widest scope consistent with the claims.

[0027] The terminology used herein is for the purpose of describing particular exemplary embodiments only and is not restrictive. For example, unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “the” used herein may also include the plural forms. Unless otherwise stated, the term “a plurality” means two or more, and “at least one” means one or more. Terms such as “first,” “second,” etc., may be used in this specification to describe various information, but such information should not be limited to these terms. These terms are used to distinguish information of the same type from one another and do not necessarily imply a specific order or sequence. For example, “first” may also be referred to as “second” without departing from the scope of embodiments described herein, and similarly, “second” may also be referred to as “first.”

[0028] The term "at least one of A, B, or C" includes seven cases: A only, B only, C only, both A and B, both A and C, both B and C, and both A, B, and C. Similarly, the statement "at least one of multiple items" refers to all possible combinations based on these items. The term "and / or" refers to any or all possible combinations of one or more related listed items. For example, "A and / or B" includes three cases: A only, B only, and both A and B. "A, B, and / or C" is equivalent to "at least one of A, B, or C". The character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0029] The term "comprising" is an open-ended description and should be understood as "including but not limited to," potentially including other content beyond what has been described. When used in this specification, the terms "comprising," "including," and / or "containing" mean the presence of the associated integers, steps, operations, elements, and / or components, but do not exclude the presence of one or more other features, integers, steps, operations, elements, components, and / or groups, or the possibility of adding other features, integers, steps, operations, elements, components, and / or groups to the system / method.

[0030] Considering the following description, these and other features of this specification, as well as the operation and function of the related components of the structure, and the economy of assembly and manufacture of the parts, can be significantly improved. All of these form part of this specification with reference to the accompanying drawings. However, it should be clearly understood that the drawings are for illustrative and descriptive purposes only and are not intended to limit the scope of this specification. It should also be understood that the drawings are not drawn to scale.

[0031] The flowcharts used in this specification illustrate operations implemented according to some embodiments of this specification. It should be clearly understood that the operations in the flowcharts may not be implemented in a sequential order. Instead, the operations may be implemented in reverse order or simultaneously. Furthermore, one or more additional operations may be added to the flowcharts. One or more operations may be removed from the flowcharts.

[0032] For ease of description, the terms that will appear later in this manual will be explained first.

[0033] Intelligent Agent: An intelligent agent is a computing entity that perceives the environment and takes actions to achieve a specific goal. The computing entity corresponding to an intelligent agent can be a computing device that runs computer programs or deploys a large language model (or can call a large language model), or a chip, intelligent module, etc. integrated in the computing device.

[0034] Tools: Tools for an intelligent agent refer to the external resources, functional modules, or auxiliary capabilities that the agent invokes to achieve a specific goal. For example, in a virtual environment, tools may include application programming interfaces (APIs), databases, memory, simulators, knowledge graphs, and algorithm libraries, while in a physical environment, tools may include sensors, actuators, and communication modules.

[0035] Figure 1 A schematic diagram of an application scenario generated by an interface provided according to embodiments of this specification is shown. For example... Figure 1 As shown, scenario 100 includes terminal device 10, server 20 and database 30.

[0036] Terminal device 10 is equipped with a client that can display an interactive page to the operator. The operator can input target text described in natural language on the interactive page. Upon completion of the operation in response to the operator's input, the client can generate an interface generation request carrying the target text and send the request to server 20. The operator can be an interface designer, a user requester, or a reviewer; this specification does not impose any restrictions.

[0037] Terminal device 10 may include mobile devices, tablets, laptops, built-in devices in motor vehicles, or similar content, or any combination thereof. In some embodiments, the mobile device may include smart home devices, smart mobile devices, virtual reality devices, augmented reality devices, or similar devices, or any combination thereof. In some embodiments, the smart home device may include smart TVs, desktop computers, etc., or any combination thereof. In some embodiments, the smart mobile device may include smartphones, personal digital assistants, gaming devices, navigation devices, etc., or any combination thereof. In some embodiments, the virtual reality device or augmented reality device may include virtual reality headsets, virtual reality glasses, virtual reality patches, augmented reality headsets, augmented reality glasses, augmented reality patches, or similar content, or any combination thereof. For example, the virtual reality device or the augmented reality device may include AR glasses, head-mounted displays, VR, etc. In some embodiments, the built-in device in the motor vehicle may include an in-vehicle computer, an in-vehicle television, etc.

[0038] The client has capabilities such as input acquisition, command parsing, request generation, and page rendering. It can be a standalone application (APP) of the interface generation service provider (i.e., the service provider offering interface generation services), or a web client (such as a browser-based web application, online office platform, etc.) that can access the website of the interface generation service provider. This specification does not impose any restrictions on this. The client's responsibility is to receive the operator's input operations, convert them into standardized interface generation requests, and pass them to the corresponding processing unit (such as server 20, or the server-side).

[0039] Server 20 is a system that executes interface generation tasks in response to interface generation requests to obtain a user interface, wherein a group of intelligent agents (including at least one intelligent agent) is deployed. Server 20 can serve as a server-side device for the interface generation service provider. Server 20 can correspond to a single computing device or a cluster of devices containing multiple computing devices. For example, server 20 can correspond to a server, a server cluster, a cloud server, etc.

[0040] For example, after receiving a user interface generation request, server 20 can invoke an intelligent agent group to generate a user interface with reference to relevant data in one or more databases 30, and then send the user interface to terminal device 10. After receiving the user interface, terminal device 10 displays the user interface to the operator through a client. The user interface mentioned in this specification can be the entire interactive page displayed in the APP window, or it can be a display card of a specific booth in the interactive page; this specification does not limit this.

[0041] One or more databases 30 include, but are not limited to, a semantic knowledge base and a layout rule base. The semantic knowledge base and layout rule base serve as reference data sources for the agent group to generate the user interface, providing data or rule support for the decision-making of the agents within the group. During the execution of the interface generation task, the agents within the agent group, or the large language model invoked by the agents, can autonomously obtain the necessary information from the semantic knowledge base or layout rule base.

[0042] It should be noted that one or more databases 30 can be databases deployed inside server 20 or databases deployed outside server 20; this manual does not limit this.

[0043] It should be understood that Figure 1 The number of terminal devices 10, servers 20, and databases 30 shown is merely illustrative. Depending on implementation needs, there can be any number of terminal devices 10, servers 20, and databases 30.

[0044] In some embodiments, server 20 may store data or instructions for implementing the user interface generation method provided in this specification, and may execute or be used to execute said data or instructions. In some embodiments, server 20 may include hardware devices with data processing capabilities and the necessary programs required to drive said hardware devices to operate.

[0045] In this specification, the Large Language Model (LLM) may also be referred to simply as the Large Model. A Large Language Model is a natural language processing model based on deep learning techniques, typically with billions to hundreds of billions or even more parameters, possessing powerful language understanding and generation capabilities. Large Language Models can employ the Transformer architecture or its variants (such as GPT, BERT, etc.), which utilizes an attention mechanism to globally model sequential data, efficiently handling long-distance dependencies and thus performing exceptionally well in natural language tasks. Large Language Models learn the statistical features and semantic relationships of language through pre-training on large-scale corpora, giving them good generalization capabilities. The core capabilities of Large Language Models include, but are not limited to: understanding contextual semantics, generating coherent and grammatically correct text, performing logical reasoning, and handling multi-task scenarios. Its usage typically includes two modes: direct inference and fine-tuning. In direct inference mode, the user guides the Large Language Model to generate specific outputs by designing prompts. Cue words can be task descriptions or instructions in text form, used to stimulate the semantic understanding and generation capabilities of large language models. In fine-tuning mode, large language models are further trained on small-scale datasets in specific domains to optimize their performance on specific tasks. The powerful generalization ability and flexibility of large language models make them an important tool in the field of artificial intelligence, providing efficient and accurate solutions for automated text generation and understanding.

[0046] In some embodiments, large language models can also understand and generate data from other modalities (such as visual and audio data). In this case, large language models can also be called multimodal large language models (MLLMs). MLLMs provide a richer and more natural interactive experience by integrating multiple types of input and output, such as text, images, and sound. The core advantage of MLLMs lies in their ability to process and understand information from different modalities and fuse this information to complete complex tasks. For example, MLLMs can analyze an image and generate descriptive text, or generate a corresponding image based on a text description. This cross-modal understanding and generation capability makes MLLMs widely applicable across multiple fields.

[0047] It should be noted that the key technologies of large language models can be found in the detailed description in the paper "A Survey of Large Language Models" (paper number: arXiv:2303.18223v16, published on March 11, 2025, public link: https: / / doi.org / 10.48550 / arXiv.2303.18223), and will not be repeated here.

[0048] Figure 2 A hardware structure diagram of a computing system 200 provided according to an embodiment of this specification is shown. The computing system 200 may be... Figure 1 Server 20 in the middle.

[0049] like Figure 2 As shown, the computing system 200 may include at least one storage medium 230 and at least one processor 220. In some embodiments, the computing system 200 may also include a communication port 250 and an internal communication bus 210. Furthermore, the computing system 200 may also include I / O components 260.

[0050] The internal communication bus 210 can connect to different system components. For example, the internal communication bus 210 can connect to storage medium 230, processor 220, communication port 250, and I / O component 260.

[0051] I / O component 260 supports input / output between computing system 200 and other components.

[0052] Communication port 250 is used for data communication between computing system 200 and the outside world. For example, communication port 250 can be used for data communication between computing system 200 and a network. Communication port 250 can be a wired communication port or a wireless communication port.

[0053] In some embodiments, the network can be any type of wired or wireless network, or a combination thereof. For example, the network may include a cable network, a wired network, a fiber optic network, a telecommunications network, an intranet, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), a public switched telephone network (PSTN), a Bluetooth network™, a ZigBee™ short-range wireless network, a near field communication (NFC) network, or a similar network.

[0054] In some embodiments, the network may include one or more network access points. For example, the network may include wired or wireless network access points, such as base stations or internet switching points. Through these access points, one or more components of various devices corresponding to the computing system 200 can connect to the network to exchange data or information.

[0055] Storage medium 230 may include a data storage device. The data storage device may be a non-transitory storage medium or a temporary storage medium. For example, the data storage device may include one or more of a disk 232, a read-only storage medium (ROM) 234, or a random access storage medium (RAM) 236. Storage medium 230 also includes at least one instruction set stored in the data storage device. The instruction set may include computer program code, which may include programs, routines, objects, components, data structures, procedures, modules, etc., that execute the methods for generating the user interface provided in this specification.

[0056] Processor 220 can be communicatively connected to storage medium 230. Processor 220 is used to execute at least one of the above-described instruction sets. When computing system 200 is running, processor 220 reads the at least one instruction set and executes the user interface generation method provided in this specification according to the instructions of the at least one instruction set.

[0057] Processor 220 may be in the form of one or more processors. In some embodiments, processor 220 may include one or more hardware processors, such as microcontrollers, microprocessors, reduced instruction set computers (RISC), application-specific integrated circuits (ASICs), application-specific instruction set processors (ASIPs), central processing units (CPUs), graphics processing units (GPUs), physical processing units (PPUs), microcontroller units, digital signal processors (DSPs), field-programmable gate arrays (FPGAs), advanced RISC machines (ARMs), programmable logic devices (PLDs), any circuit or processor capable of performing one or more functions, or any combination thereof.

[0058] For the purpose of illustrating the point only, in the appendix Figure 2 Only one processor 220 is shown in the computing system 200. However, it should be noted that the computing system 200 may also include multiple processors. Therefore, the operation and / or method steps disclosed in this specification may be executed by one processor as described in this specification, or they may be executed jointly by multiple processors. For example, if processor 220 of computing system 200 in this specification executes steps A and B, it should be understood that steps A and B may also be executed jointly or separately by two different processors 220 (e.g., the first processor executes step A, the second processor executes step B, or the first and second processors jointly execute steps A and B).

[0059] Figure 3 A flowchart of a user interface generation method P300 according to an embodiment of this specification is shown. The server and client can collaboratively execute the user interface generation method P300.

[0060] like Figure 3 As shown, the user interface generation method P300 includes the following steps.

[0061] S310: In response to the operator's input to complete the operation, the client generates a UI generation request. The UI generation request includes the target text described by the operator in natural language.

[0062] The client on the terminal device displays an interactive page, including an input area, to the operator. The operator can input text in the input area using various methods, such as character input or voice input. After the operator has finished inputting the target text, they can perform an input completion action, such as clicking the "Finish" or "Confirm" button. In response to the operator's input completion action, the client generates a request for an interface based on the target text entered by the operator.

[0063] The client in the terminal device can be an APP client (for the operator) provided by the interface generation service provider, or a web client that can access the website of the interface generation service provider; this specification does not impose any restrictions on this. The interface generation service provider and the provider of the target APP (for ordinary users) that displays the generated user interface can be the same party, so that the generated user interface can better conform to the relevant standards of the target APP that displays the user interface.

[0064] The target text in natural language description refers to the information content written or expressed by the operator using the language used in everyday human communication. For example, "We predict that Company A's new energy battery technology will dominate the market within the next three years, with its energy density increasing by 30%." This target text is the seed for interface generation, that is, the basis or initial data that drives interface generation, influencing the structure and content of the generated interface.

[0065] S320: The client sends a UI generation request to the server.

[0066] S330: After receiving the interface generation request, the server uses an intelligent agent group to perform the interface generation task based on the target text and semantic knowledge base carried in the interface generation request, so as to obtain at least one user interface.

[0067] After receiving a UI generation request, the server can parse the request to obtain the target text and then provide it to the intelligent agent group. Alternatively, the server can provide the UI generation request to the intelligent agent group, which can then parse the target text from it. This specification does not restrict the method by which the intelligent agent group obtains the target text.

[0068] After obtaining the target text, the agent group can perform an interface generation task with reference to a semantic knowledge base, thereby generating a user interface that matches the target text, or a user interface that can present the effective information in the target text. An agent group includes at least one agent. When an agent group includes multiple agents, the multiple agents are configured to perform different subtasks in the interface generation task. This specification does not limit the number of agents included in an agent group.

[0069] Semantic knowledge bases store structured data, which is inherently meaningful or can embody its content, and can be understood by intelligent agents or large language models. Semantic knowledge bases can be generated manually by relevant personnel or pre-generated using artificial intelligence technology.

[0070] For example, during the generation process, relevant personnel can collect a large number of existing user interfaces and use artificial intelligence tools to analyze their structure and framework, deconstructing these user interfaces into reusable basic units, i.e., interface elements. Personnel can also perform semantic engineering on each interface element based on their own experience or with the help of artificial intelligence tools. Semantic engineering can be understood as standardizing the definitions of each deconstructed interface element to obtain predefined information for each element; that is, "translating" the interface elements from a visual perspective into readable and understandable standardized structured information such as intelligent agents and large language models. After semantic engineering, personnel or artificial intelligence tools can store the predefined information of each interface element in a corresponding database, thus the database storing the predefined information is called a semantic knowledge base.

[0071] The semantic knowledge base includes predefined information for multiple interface elements. These interface elements include multiple interface components and multiple visual attributes. The predefined information for each interface element represents its style and usage description. Interface components can be understood as interface elements that carry specific content (such as text or images). Visual attributes can be understood as interface elements that present specific content, also known as visual variables. In other words, interface components focus on "what is used to present it," while visual attributes focus on "how it is presented."

[0072] Interface components include, but are not limited to: text components, interactive components, and image components. Text components include titles at different levels (such as main titles, subtitles, etc.) and body text. Interactive components include buttons, input boxes, and switches. Image components include charts (such as bar charts, comparison charts, trend charts, etc.), icons, and images.

[0073] Visual attributes define the overall or partial visual presentation of the user interface (such as the content carried by interface components), including but not limited to: typography attributes, color attributes, and effect attributes. Typography attributes include font and font size. Color attributes include color, transparency, and gradients. Effect attributes include shadows, blur, and borders.

[0074] For example, the style of interface elements can be a Cascading Style Sheets (CSS) style, and this specification does not limit this. Taking a button in an interface component as an example, its style can include: the button's shape, size, etc. Taking font in visual attributes as an example, its style can include: font weight (such as normal, bold, medium), whether it is italic, etc.

[0075] It should be noted that for interface components such as main titles, subtitles, and buttons, each component can have a corresponding usage description. For visual attributes, if a visual attribute includes multiple styles, each style can have a corresponding usage description, or multiple styles of a visual attribute can have a corresponding usage description; this specification does not impose such restrictions. For example, if the visual attribute is font, and the multiple styles of this visual attribute are different types of fonts, each type of font can have a corresponding usage description. As another example, if the visual attribute is color, and the multiple styles of this visual attribute are different colors, different colors within the same color family can have a corresponding usage description, and colors from different color families can have different usage descriptions.

[0076] The usage description, also known as a user manual or purpose description, is expressed in natural language in semantic knowledge bases, and this specification does not impose any restrictions on it. Large language models or agents can use the usage description of a particular interface element to determine whether to use it.

[0077] For example, in addition to style and usage description, the predefined information for each interface element can also include: name, effect, usage restrictions, usage suggestions, etc. In short, the richer the content of the predefined information for each interface element, the more reliable and detailed the data that the intelligent agent group can refer to when performing the interface generation task, and the higher the quality of the generated user interface.

[0078] In some embodiments, the multiple interface components in the semantic knowledge base include multiple atomic components and at least one molecular component, and each molecular component includes at least two atomic components.

[0079] An atomic component can be understood as the smallest granular component obtained after the interface has been deconstructed. Molecular components are composed of atomic components. For example, Figure 4 As shown, molecular component 1 consists of a title and a diagram. Molecular component 2 consists of a title and body text. The semantic knowledge base can store the following information for each molecular component: the name of the molecular component, the names or serial numbers of the atomic components included in the molecular component, the relative positions and size ratios of the atomic components in the molecular component, and a description of the use of the molecular component. Among these, the names or serial numbers of the atomic components included in the molecular component, the relative positions and size ratios of the atomic components can be understood as the style of the molecular component.

[0080] In this embodiment, the semantic knowledge base divides components into atomic components and molecular components composed of atomic components. This allows the intelligent agent group to flexibly select either basic atomic components or pre-combined molecular components when generating the interface. Atomic components ensure high flexibility in interface construction, while molecular components improve generation efficiency and avoid layout complexity issues that may arise from combining multiple atomic components. By directly reusing pre-built, mature component combinations, not only can repetitive design work be reduced, but the efficiency and consistency of interface generation can also be further improved.

[0081] In some embodiments, the usage description corresponding to each interface element includes at least one of the following dimensions: the characteristics of the interface element; the applicable scenarios of the interface element; or the usage specifications of the interface element. Therefore, when the intelligent agent group performs the interface generation task, it can refer to the usage description of each interface element in the semantic database, automatically select suitable interface elements based on the scenario features of the target text, and constrain the interface generation process with the usage specifications of the interface elements, so that the generated user interface can ensure personalization while avoiding excessive randomness, and make it conform to the relevant standards of APP (an APP that displays a user interface to ordinary users) or the needs of operators.

[0082] Taking a button in a UI component as an example, its usage is described as follows: This UI component is a basic button, one of the core interactive elements in the user interface, used to trigger a specified action or event.

[0083] **Applicable Scenarios** Call to Action (CTA): A key entry point on a webpage or app that guides users to complete a specific task, such as "Buy Now" or "Register".

[0084] Navigation and Confirmation: Used for page navigation (such as "Next") and selections in dialog boxes (such as "OK" and "Cancel").

[0085] **Component Features** Appearance: Usually rounded rectangle or capsule shape, with a clickable visual cues.

[0086] Size: Available in two styles: long buttons and short buttons. Long buttons are more prominent and attract users to click, while short buttons save more space.

[0087] Text label: Contains brief text that directly describes the action to be performed after clicking.

[0088] **Usage Guidelines** The text displayed on the button cannot exceed four characters. In some embodiments, the interface generation task includes multiple subtasks. The agent group includes multiple dedicated agents that perform different subtasks.

[0089] Specialized intelligent agents are artificial intelligence systems focused on specific tasks, enabling them to perform those tasks accurately and efficiently. Specialized intelligent agents can be created by fine-tuning general models, injecting knowledge, and defining roles.

[0090] This embodiment decomposes the complex interface generation task into multiple specialized sub-tasks. By having multiple specialized intelligent agents work together to execute each sub-task, the professional advantages of each specialized intelligent agent for specific tasks can be fully utilized, achieving a fine division of labor. This reduces problems such as generation deviation or inefficiency caused by overly complex interface generation tasks, thereby improving the quality of the generated user interface.

[0091] For example, the interface generation task includes multiple sub-tasks as follows: the first sub-task is to understand the target text and transform it into interface requirements (such as interface components needed to generate a user interface); the second sub-task is to generate at least one interface template to be filled based on the interface requirements (such as an interface skeleton or architecture, without specific content); the third sub-task is to fill each interface template to obtain at least one user interface. Thus, the intelligent agent group may include a dedicated intelligent agent for performing the first sub-task, a dedicated intelligent agent for performing the second sub-task, and a dedicated intelligent agent for performing the third sub-task.

[0092] In this example, each specialized agent in the agent group collaborates to improve the professionalism and efficiency of each agent in performing sub-tasks, reduce the complexity of a small number of agents performing interface generation tasks to reduce deviations, and achieve high-quality and high-efficiency interface generation.

[0093] In some embodiments, the multiple dedicated intelligent agents include a first intelligent agent, a second intelligent agent, and a third intelligent agent. The server uses the first intelligent agent to understand the target text and determine at least one target component from multiple interface components. The second intelligent agent then generates at least one interface template to be populated based on the at least one target component. Each interface template includes a layout structure and a visual theme. The server can also use the third intelligent agent to populate each interface template based on the target text to obtain at least one user interface.

[0094] The server can store the calling rules and timings for multiple dedicated agents, and then call each dedicated agent based on the calling rules and timings to complete a specific sub-task.

[0095] For example, see Figure 5The server inputs the target text into the first agent. The first agent, based on the target text, calls a semantic knowledge base to select at least one target component and outputs it. The second agent, having obtained at least one target component (which can be input from the server or the first agent), can call a layout rule base. Referring to the preset layout rules for interface components in the layout rule base, it arranges and combines the at least one target component to generate N layout structures and determines the visual theme (e.g., by referring to the content of the target text or through other methods). Then, it combines the N layout structures and the visual theme to obtain N interface templates and outputs them, where N is an integer greater than or equal to 1. The third agent, having obtained N interface templates (which can be input from the server or the second agent), determines each target component in each interface template and calls the semantic knowledge base to determine the target usage specifications for each target component. Based on the target text and the target usage specifications, it fills each interface template to obtain N user interfaces. The third agent, having obtained N user interfaces, outputs the N user interfaces, thus providing the server with N user interfaces.

[0096] The following mainly describes the specific process by which each dedicated intelligent agent executes a particular subtask.

[0097] In some embodiments, the server executes the following process through or by invoking a first intelligent agent: generating first guidance information based on the target text, inputting the first guidance information into a first large language model, and determining at least one target component based on the output of the first large language model. The first guidance information is used to guide the first large language model to extract information to be displayed from the target text and to recommend interface components capable of carrying the information to be displayed based on the usage descriptions of each interface component in the semantic knowledge base.

[0098] The first language model can utilize its natural language processing capabilities to analyze, extract, and summarize target text to obtain the information to be displayed. This information includes the user interface's theme and key benefits for attracting users. For example, a theme like "Company A's Battery Technology Revolution." For fund-related user interfaces displayed within the target app, key benefits for attracting users include "30% increase in energy density," "Market leader for the next three years," and "Disruptive technology." After extracting the information, the first language model may or may not output it as a basis for subsequent selection of target components; this specification does not impose any restrictions on this.

[0099] For example, the first intelligent agent can obtain a preset guidance template and assemble the target text and the preset guidance template to obtain first guidance information. The preset guidance template may include the following information: the role of the first large language model, a description of the task performed by the first large language model, aspects to be noted when performing the task (such as the usage description of each interface component in the semantic knowledge base), the output content and its format requirements, and the storage location of the semantic knowledge base. Thus, the first large language model can output at least one target component based on the first guidance information. For example, the first large language model selects a data display component or an indicator highlighting component as the target component based on the extracted "energy density increase of 30%", selects a trend chart component as the target component based on the extracted "market leader in the next three years", and selects a button component as the target component.

[0100] In the above embodiments, the first intelligent agent uses a first large language model to deeply analyze the target text, quickly filter redundant content, and accurately extract the key information that can be presented, improving processing efficiency, especially suitable for long texts or complex scenarios. Furthermore, the target component selected by the first large language model based on the extracted information focuses on displaying key information, reducing visual interference on the interface, enabling users to efficiently obtain core content, and improving the quality of the user interface.

[0101] It should be noted that if the first language model outputs the information to be displayed extracted from the target text, the first agent can select at least one target component suitable for carrying the information from multiple interface components based on the information to be displayed. Specifically, the first agent can call the first language model again to select the target component, or it can call other tools to select the target component; this specification does not impose any restrictions on this.

[0102] In some embodiments, the server executes the following process through or by invoking a second intelligent agent: obtaining a preset layout rule for the interface component, generating N layout structures based on the preset layout rule and the at least one target component, selecting styles of at least some visual attributes to obtain a visual theme, and then generating N interface templates based on the N layout structures and the visual theme.

[0103] Preset layout rules can be defined in advance and stored in the layout rule library, including multiple rules or principles that conform to visual balance or design aesthetics. Preset layout rules include, but are not limited to: placing title components at the top of the user interface, placing indicator highlight components in secondary positions, placing trend chart components in the main area, and placing button components at the bottom, etc.

[0104] The target components included in each of the N layout structures can be the same (e.g., each layout structure includes at least one target component) or different (e.g., some layout structures include at least one target component, and some layout structures include some of their target components). This specification does not impose any restrictions on this. When each layout structure includes the same target component, the relative positions of the different target components can be different.

[0105] The second agent can select different visual themes for each layout structure, or it can select the same visual theme for N layout structures; this specification does not impose any restrictions on this. A visual theme can be understood as a style combination of at least some visual attributes selected by the second agent. For example, if the second agent selects font and color as visual attributes, and selects the style as SimSun and blue, then the visual theme obtained by the second agent is that the text in the user interface is presented in SimSun font, and the areas in the user interface that need to be colored are rendered in blue.

[0106] In this embodiment, the second intelligent agent arranges and combines at least one target component selected by the first intelligent agent according to a preset layout rule, taking into account both flexibility and personalization. This can generate multiple layout structures that conform to visual balance or design aesthetics, achieving standardization and controllability of layout generation, improving the generation efficiency of interface templates, and ensuring that the user interface generated accordingly reaches a professional level in terms of visual appeal, ease of operation, and rationality.

[0107] For example, the second agent generates second guidance information based on preset layout rules and the at least one target component, inputs the second guidance information into the second large language model, and determines N layout structures based on the output of the second large language model. The second guidance information is used to guide the second large language model to arrange and combine the at least one target component according to the preset layout rules to generate a preset number of layout structures. The second guidance information may include: relevant information of the preset guidance template (the role of the second large language model, the task to be performed, output examples of layout structures, and the number of outputs, etc.), the preset layout rules, and the at least one target component.

[0108] The second agent can also assemble the preset guidance template and the at least one target component to obtain the second guidance information. The second guidance information includes relevant information about the preset guidance template, the storage location of the layout rule base, and the at least one target component, etc. This specification does not limit the specific content of the second guidance information.

[0109] It should be noted that the second agent can also call other tools besides the large language model to obtain N layout structures, and this specification does not impose any restrictions on this.

[0110] In this embodiment, the second agent explicitly specifies the pre-defined layout rules and the required number of layout structures to the second language model, providing clear design standards and requirements. This achieves a high degree of controllability in the layout structure generation process, avoiding deviations or conflicts caused by the second language model's free exploration. Furthermore, specifying the number of layout structures guides the second language model to generate diverse layout structures in a targeted manner, enabling batch generation of layout structures, improving generation efficiency, and shortening the design cycle. Additionally, this embodiment supports dynamic adjustment of pre-defined layout rules in the layout rule base, enhancing development flexibility and reducing maintenance costs.

[0111] The second agent can either choose a style for visual attributes based on the text content to obtain a visual theme, or it can choose a style for visual attributes based on the operator's instructions. The visual theme determines the overall visual style of the user interface, such as influencing the color, font, and effects of components (e.g., glass texture, metallic texture, gradient, cartoon, etc.).

[0112] For example, the second agent generates third guidance information based on the target text, inputs the third guidance information into the third language model, and determines the visual theme based on the output of the third language model. The third guidance information is used to guide the third language model to recommend visual attributes and styles that match the target text based on the style and usage description of each visual attribute in the semantic knowledge base. The third guidance information may include: the target text, the storage location of the semantic knowledge base, information on the preset guidance template, such as the role of the third language model, the task to be performed (e.g., recommending visual attributes and styles that match the target text based on the style and usage description of each visual attribute in the semantic knowledge base), and logical examples of selecting visual attributes and styles (e.g., for text containing content related to new energy, such as batteries, blue tones and simple fonts can be selected), etc.

[0113] In this example, the third language model deeply analyzes the text content provided by the operator and the usage description corresponding to the visual attributes in the knowledge base, intelligently selects the interface visual attributes that match the target text, realizes cross-modal alignment and fusion of text content and visual form, ensures that the visual attributes are highly consistent with the text connotation, and enables the subsequently generated user interface to accurately convey the core content of the text in terms of visuals.

[0114] For example, the second intelligent agent obtains the visual requirements specified by the operator, and selects styles of at least some visual attributes from the semantic knowledge base based on the visual requirements to obtain a visual theme. The visual requirements specified by the operator can be explicit, such as specific styles of visual attributes like font, style, and color, or they can be abstract, such as textual descriptions that are similar to but not exactly the same as specific styles of visual attributes. The visual requirements can be contained in the target text, or they can be manually selected by the operator on the interactive page while the target text is input on the client. For example, after the operator inputs the target text, the client presents a selection window on the interactive page, displaying style options for each visual attribute, so that the operator can choose the style they want. This specification does not impose any restrictions on this.

[0115] In this example, the operator explicitly specifies the visual requirements, providing a clear direction for the interface design of the third-party intelligent agent, ensuring that the generated interface visual theme is highly consistent with the user's expectations, returning the choice of visual theme to the operator, and supporting personalized customization.

[0116] After obtaining N layout structures and visual themes, the second agent can combine the layout with the selected visual theme for each layout structure to solidify it into a structured interface template. Each interface template is a "wireframe" or "skeleton" with clear area divisions and style definitions but empty content (or placeholders set in corresponding positions), serving as a "container" for subsequent content generation.

[0117] In some embodiments, the semantic knowledge base includes usage descriptions for at least some interface components, including usage specifications. The server executes the following process through a third agent: for each interface template, it obtains the target usage specifications corresponding to each target component in the interface template from the semantic knowledge base; it generates the content to be filled for each target component in the interface template based on the target text and the target usage specifications; and then it fills each target component in the interface template with the content to be filled for each target component to obtain the user interface.

[0118] The server can execute the above process N times through a third-party agent to obtain N user interfaces; the server can also execute the above process once in parallel through a third-party agent to obtain N user interfaces, and this specification does not impose any restrictions on this. Furthermore, if some target components in the interface template do not have usage specifications stored in the semantic knowledge base, the third-party agent (or the large language model called by the third-party agent) can determine the content to be filled for the corresponding components based on the general specifications of the corresponding components learned during pre-training or fine-tuning.

[0119] In this embodiment, the third agent strictly follows the usage specifications of the target component (such as text length limits, image ratio constraints, etc.) to generate content to be filled that is adapted to the target component, so as to avoid layout disorder and visual imbalance caused by too much or too little content to be filled, and improve the visual quality of the subsequently generated interface.

[0120] The third agent can invoke different types of tools to generate content to be populated for different types of target components in each interface template. These tools include, but are not limited to: large language models and image generation models.

[0121] For example, the interface template includes a text component. The content to be filled in the text component is obtained as follows: a third agent generates fourth guiding information based on the text component and the target text, inputs the fourth guiding information into a fourth language model, and determines the content to be filled in the text component based on the output of the fourth language model. The fourth guiding information is used to guide the fourth language model to generate content to be filled that conforms to the usage specifications of the text component based on the target text.

[0122] For each interface template, the third agent can parse the target components contained in the template, identify the text components (e.g., title, body, button), and generate fourth guidance information based on the text components and target text in the interface template. The fourth guidance information may include: the target text, the text components contained in each of the N interface templates, and relevant information about the preset guidance templates, such as the role of the fourth language model, its execution task (e.g., extracting content from the target text and generating content to be filled for the text components in each interface template based on the usage specifications of the corresponding text components in the semantic knowledge base), and the storage location of the semantic knowledge base. For example, the usage specification for the title component is "no more than 20 characters." The content to be filled generated by the third agent for the title component might be: "Company A's battery revolution: energy density soars by 30%", for the body component: "Explore disruptive technologies guiding the next three years, redefining industry standards", and for the button component: "Go and see".

[0123] In this embodiment, the fourth language model understands and extracts key information from the target text, ensuring that the generated text is highly consistent with the operator's intention. Furthermore, it is constrained by usage guidelines, so that the content to be filled can reflect key information while complying with the usage guidelines of the component, avoiding problems such as layout disorder and poor visual effects caused by exceeding the word limit.

[0124] For example, the interface template includes an image component, and the content to be filled in the image component is obtained in the following way: a third agent generates an image generation instruction based on the target text, the visual theme of the interface template and the usage specifications of the image component, and calls the image generation model based on the image generation instruction to generate the content to be filled in the image component.

[0125] The third agent can invoke a large language model, enabling it to generate image generation instructions based on the target text, the visual theme of the interface template, and the usage specifications of image components.

[0126] For example, taking the main image component as an example, the usage specification for the image component is "image aspect ratio of 16:9". Image generation instructions, for example, can generate a technologically advanced schematic diagram showing the internal structure of a battery, or an abstract background depicting a growth trend, with the image color scheme being blue and the image aspect ratio being 16:9.

[0127] In this example, the third agent integrates the target text and visual theme, calls the raw image model to generate an image that conforms to the image component usage specifications and is adapted to the target text, ensuring that the image and the overall style of the interface are harmoniously integrated, improving the efficiency of information transmission and the consistency of user experience.

[0128] After obtaining the content to be filled for each text component and / or image component in each interface template, the third agent can fill or assemble the content to be filled for each component into the corresponding position in the interface template, resulting in N user interfaces. One of the user interfaces is as follows: Figure 6 As shown.

[0129] S340: The server sends the at least one user interface to the client. Each user interface is in the format of at least one of the following: an interface image; executable code; or editable data.

[0130] After receiving at least one user interface, the client can display the above-described format of each user interface to the operator. For example, the client receives two user interfaces, User Interface 1 and User Interface 2. The client can display the interface image, executable code, and editable data of User Interface 1, as well as the interface image, executable code, and editable data of User Interface 2 to the operator.

[0131] The server sends user interfaces in various formats to the client, allowing the user to quickly perform subsequent operations. These include interface images (such as...). Figure 6(As shown) This allows operators to quickly preview the interface, improving the efficiency of deciding whether to select it. When an operator deems a user interface ready to use, they can directly integrate the executable code (e.g., webpage code) into the target app, allowing the app to display the interface to the user in the corresponding location, avoiding the time-consuming and error-prone manual coding. When an operator deems a user interface needs modification, they can directly modify its editable data (e.g., JSON format), avoiding the time-consuming process of regenerating the interface. Multiple formats are provided to operators, eliminating the need for additional coding for the user interface, significantly reducing labor costs and time, and improving interface design efficiency.

[0132] For example, the user interface generated by this specification can be as follows: Figure 7 The interactive page shown (such as the interactive page of the target APP) can also be Figure 7 This manual does not limit the display cards (such as the landing page or booth information of the target APP and partner organizations) shown on the interactive page.

[0133] The multiple large language models mentioned in this specification can be the same large language model or different large language models; this specification does not impose any restrictions on this. Large language models can be pre-injected with relevant content from a semantic knowledge base, or they can be linked to the semantic knowledge base via code, allowing the large language model to read the required data from the semantic knowledge base when performing a corresponding task.

[0134] It is understood that the various guidance messages mentioned in this specification can be prompts, and this specification does not impose any restrictions on them. Furthermore, different guidance messages mentioned in this specification, provided they conform to the execution logic, can be merged into a single guidance message and input into the large language model for execution, or a single guidance message can be split into multiple guidance messages and input into the large language model for execution. While the description of the subtasks executed by each dedicated agent above primarily uses a large language model as an example, this specification is not limited to this. Provided the execution logic conforms, multiple large language models can also be used to implement the subtasks executed by each dedicated agent.

[0135] It should be noted that the embodiments in this specification mainly use three intelligent agents as an example to introduce the specific process of each sub-task in the interface generation task executed by each intelligent agent. However, this specification is not limited to this, and the intelligent agent group may also include more or fewer intelligent agents.

[0136] For example, an agent group may consist of fewer agents, in which case one agent can execute multiple subtasks from the first, second, and third subtasks, and the guidance information involved in the above process needs to be adaptively adjusted. Alternatively, the server can divide the interface generation task in other ways, and the guidance information involved in the agent's execution of specific tasks can be adaptively adjusted based on the task division; this specification does not impose any restrictions on this.

[0137] For example, an agent group can include a larger number of agents, allowing the server to further subdivide the interface generation task. For instance, the first subtask can be divided into multiple tasks (e.g., extracting information to be displayed and matching interface components could be two tasks), the second subtask into multiple tasks (e.g., generating layout structure, visual theme, and assembling into an interface template could be three tasks), and the third subtask into multiple tasks (e.g., generating content to be filled for image components and generating content to be filled for text components could be two tasks). This allows a larger number of agents to execute the further subdivided tasks, adaptively adjusting the guidance information invoked by each agent from the large language model during execution. These variations are not listed here.

[0138] In summary, the user interface generation method, computing system, and storage medium provided in this specification enable intelligent agent groups to generate user interfaces using relevant data from a semantic knowledge base. This systematizes and automates the interface generation task, significantly improving generation efficiency and diversity, reducing reliance on manual design and copywriting, saving labor costs, and lowering the technical barrier to manual design. It is suitable for large-scale interface production and iteration in concurrent scenarios. Furthermore, while providing a reference, the semantic knowledge base also constrains the divergent thinking of the intelligent agent groups, preventing overly random user interfaces and improving the stability and standardization of interface generation.

[0139] This specification, in another aspect, provides a computer-readable non-transitory storage medium storing at least one set of executable instructions for generating an interface. When the executable instructions are executed by a processor, they instruct the processor to perform the steps of method P300 described herein. In some possible embodiments, various aspects of this specification may also be implemented as a program product comprising program code. When the program product is run on computing system 200, the program code causes computing system 200 to perform the steps of method P300 described herein. The program product for implementing the above method may employ a portable compact disk read-only memory (CD-ROM) containing program code and may run on computing system 200. However, the program product of this specification is not limited thereto. In this specification, the readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system. The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. The computer-readable storage medium may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium may also be any readable medium other than a readable storage medium that can send, propagate, or transmit programs for use by or in connection with an instruction execution system, apparatus, or device. Program code contained on a readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof. Program code for performing the operations described herein can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on computing system 200, partially on computing system 200, as a standalone software package, partially on computing system 200 and partially on a remote electronic device, or entirely on a remote electronic device.

[0140] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0141] In summary, after reading this detailed disclosure, those skilled in the art will understand that the foregoing detailed disclosure is presented by way of example only and is not restrictive. Although not explicitly stated herein, those skilled in the art will understand that this specification requires various reasonable changes, improvements, and modifications to the embodiments. These changes, improvements, and modifications are intended to be made by this specification and are within the spirit and scope of the exemplary embodiments described herein.

[0142] Furthermore, certain terms in this specification have been used to describe embodiments of this specification. For example, "an embodiment," "an embodiment," and / or "some embodiments" mean that a particular feature, structure, or characteristic described in connection with that embodiment may be included in at least one embodiment of this specification. Therefore, it is to be emphasized and understood that two or more references to "an embodiment" or "an embodiment" or "alternative embodiment" in various parts of this specification do not necessarily refer to the same embodiment. Moreover, specific features, structures, or characteristics may be suitably combined in one or more embodiments of this specification.

[0143] It should be understood that in the foregoing description of the embodiments in this specification, various features are combined in a single embodiment, drawing, or description for the purpose of simplifying the description and aiding in the understanding of a feature. However, this does not mean that the combination of these features is necessary, and those skilled in the art will readily identify some of them as separate embodiments when reading this specification. That is, the embodiments in this specification can also be understood as an integration of multiple secondary embodiments. It is also valid when each secondary embodiment contains fewer than all the features of a single foregoing disclosed embodiment.

[0144] Every patent, patent application, publication of a patent application, and other material cited herein, such as articles, books, specifications, publications, documents, and literature (excluding any related historical examination documents), is referenced for all purposes relevant to this document, including in the specification and claims herein. However, in the event of any inconsistency or conflict between the descriptions, definitions, and / or terms used in the foregoing and those used herein, the descriptions, definitions, and / or terms used herein shall prevail.

[0145] Finally, it should be understood that the embodiments disclosed herein are illustrative of the principles of the embodiments described in this specification. Other modified embodiments are also within the scope of this specification. Therefore, the embodiments disclosed in this specification are merely examples and not limitations. Those skilled in the art can implement the applications described in this specification using alternative configurations based on the embodiments in this specification. Therefore, the embodiments in this specification are not limited to the embodiments precisely described in the applications.

Claims

1. A method for generating a user interface, comprising: Receive an interface generation request from the client, the interface generation request including target text described by the operator in natural language; as well as An intelligent agent group performs an interface generation task based on the target text and a semantic knowledge base to obtain at least one user interface. The semantic knowledge base includes predefined information of multiple interface elements, each of which includes multiple interface components and multiple visual attributes. The predefined information of each interface element represents the style and usage description of the interface element. The intelligent agent group includes at least one intelligent agent.

2. The method according to claim 1, wherein, The interface generation task includes multiple sub-tasks, and the intelligent agent group includes multiple dedicated intelligent agents that execute different sub-tasks.

3. The method according to claim 2, wherein, The multiple sub-tasks include: Understand the target text and translate it into interface requirements; At least one interface template to be populated is generated based on the aforementioned interface requirements; and Fill each of the interface templates to obtain the at least one user interface.

4. The method according to claim 2 or 3, wherein, The plurality of dedicated intelligent agents include a first intelligent agent, a second intelligent agent, and a third intelligent agent. The process of the intelligent agent group performing an interface generation task based on the target text and a semantic knowledge base to obtain at least one user interface includes: The first agent understands the target text and determines at least one target component from the plurality of interface components; The second agent generates at least one interface template to be populated based on the at least one target component, each interface template including a layout structure and a visual theme; and The at least one user interface is obtained by the third agent filling each interface template with the target text.

5. The method according to claim 4, wherein, The step of understanding the target text and determining at least one target component from the plurality of interface components through the first intelligent agent includes: Executed by the first intelligent agent: First guidance information is generated based on the target text. This first guidance information guides a first language model to extract information to be displayed from the target text and recommends interface components capable of carrying the information to be displayed based on the usage descriptions of each interface component in the semantic knowledge base. The first guidance information is input into the first large language model, and the at least one target component is determined based on the output of the first large language model.

6. The method according to claim 4, wherein, The step of generating at least one interface template to be filled based on the at least one target component by the second intelligent agent includes: Executed by the second intelligent agent: Obtain the preset layout rules of the interface components, and generate at least one layout structure based on the preset layout rules and the at least one target component; By selecting styles for at least some of the visual attributes, the visual theme is obtained; and The at least one interface template is generated based on the at least one layout structure and the at least one visual theme.

7. The method according to claim 6, wherein, The step of generating at least one layout structure based on the preset layout rules and the at least one target component includes: Based on the preset layout rules and the at least one target component, second guidance information is generated. This second guidance information guides the second language model to arrange and combine the at least one target component according to the preset layout rules to generate a preset number of layout structures. The second guidance information is input into the second large language model, and the at least one layout structure is determined based on the output of the second large language model.

8. The method according to claim 6, wherein, The step of selecting styles for at least some visual attributes to obtain the visual theme includes: Based on the target text, third guidance information is generated. This third guidance information guides a third language model to recommend visual attributes and their styles that are compatible with the target text, based on the style and usage description of each visual attribute in the semantic knowledge base. The third guiding information is input into the third major language model, and the visual theme is determined based on the output of the third major language model.

9. The method according to claim 6, wherein, The step of selecting styles for at least some visual attributes to obtain the visual theme includes: Obtain the visual requirements specified by the operator; and Based on the stated visual requirements, styles of at least some of the visual attributes are selected to obtain the visual theme.

10. The method according to claim 4, wherein, At least some of the interface components have usage descriptions including usage guidelines. The process of obtaining the at least one user interface by filling each interface template with the target text through the third agent includes: Executed by the third intelligent agent: For each interface template, the target usage specifications corresponding to each target component in the interface template are obtained from the semantic knowledge base; Based on the target text and the target usage specifications, generate the content to be filled for each target component in the interface template; and The user interface is obtained by filling each target component in the interface template with the content to be filled for each target component.

11. The method of claim 10, wherein, The interface template includes a text component, and the content to be filled in the text component is obtained based on the following method: Based on the text component and the target text, a fourth guiding information is generated. The fourth guiding information is used to guide the fourth language model to generate content to be filled that conforms to the usage specifications of the text component based on the target text. as well as The fourth guiding information is input into the fourth major language model, and the content to be filled in the text component is determined based on the output of the fourth major language model.

12. The method of claim 10, wherein, The interface template includes an image component, and the content to be filled in the image component is obtained based on the following method: Image generation instructions are generated based on the target text, the visual theme of the interface template, and the usage specifications of the image components; as well as Based on the image generation instruction, the image generation model is invoked to generate the content to be filled in the image component.

13. The method according to claim 1, wherein, The plurality of interface components include a plurality of atomic components and at least one molecular component, and each molecular component includes at least two atomic components.

14. The method according to claim 1, wherein, The usage description for each UI element includes at least one of the following dimensions: The characteristics of the interface elements; The applicable scenarios for the interface elements; or The usage guidelines for interface elements.

15. The method according to claim 1, wherein, The method further includes: The at least one user interface is sent to the client, wherein the at least one user interface is in the format of at least one of the following: an interface image; executable code; or editable data.

16. A computing system, comprising: At least one storage medium storing at least one instruction set; as well as At least one processor is communicatively connected to the at least one storage medium, wherein the at least one processor reads the at least one instruction set during operation and performs the method as described in any one of claims 1-15 according to the instructions of the at least one instruction set.

17. A computer-readable non-transitory storage medium, wherein, The computer-readable non-transitory storage medium stores at least one set of instructions, which, when executed by at least one processor, implement the method as described in any one of claims 1-15.

Images