CN118154726B

CN118154726B - Resource processing design method, device and computer equipment based on large language model

Info

Publication number: CN118154726B
Application number: CN202410579100.8A
Authority: CN
Inventors: 黄惠; 黄期瑞
Original assignee: Shenzhen University
Current assignee: Shenzhen University
Priority date: 2024-05-11
Filing date: 2024-05-11
Publication date: 2024-10-15
Anticipated expiration: 2044-05-11
Also published as: CN118154726A

Abstract

The application relates to a resource processing design method, a device and computer equipment based on a large language model. The method comprises the following steps: the method comprises the steps of receiving drawing resource inquiry information sent by a terminal, wherein the drawing resource inquiry information is input based on an intelligent dialogue interface in a drawing design interface, and the drawing design interface comprises the intelligent dialogue interface and a canvas operation interface; determining a drawing resource type and a resource demand parameter corresponding to drawing resource inquiry information based on a large language model, calling a target resource generation application matched with the drawing resource type, generating drawing resources based on the resource demand parameter through the target resource generation application, and sending the drawing resources to a terminal; and displaying drawing resources on the intelligent dialogue interface by the terminal, wherein the drawing resources are configured to target positions corresponding to the canvas operation interface according to the selection operation, and obtaining a corresponding design drawing. The drawing efficiency can be improved by adopting the method.

Description

Resource processing design method and device based on large language model and computer equipment

Technical Field

The present application relates to the field of computer technology, and in particular, to a method, an apparatus, a computer device, a computer readable storage medium, and a computer program product for designing a resource process based on a large language model.

Background

With the invention of computer technology, the technology of graph design is correspondingly developed. Graph design techniques, i.e., graph design based on graph authoring tools, such as informational graphs or other types of art graphs.

However, in the prior art, the graph is designed mainly according to the conventional graph creation tool, and in the process of designing the graph, the graph creation tool needs to be separated from the graph creation tool to perform more complicated searching work of design resources, so that the manufacturing efficiency of the design graph is lower.

Disclosure of Invention

In view of the foregoing, it is desirable to provide a resource processing design method, apparatus, computer device, computer-readable storage medium, and computer program product based on a large language model, which can improve drawing efficiency.

In a first aspect, the present application provides a method for designing a resource process based on a large language model, which is applied to a server, and includes:

The method comprises the steps of receiving drawing resource inquiry information sent by a terminal, wherein the drawing resource inquiry information is input based on an intelligent dialogue interface in a drawing design interface, and the drawing design interface comprises the intelligent dialogue interface and a canvas operation interface;

determining a drawing resource type and a resource demand parameter corresponding to drawing resource inquiry information based on a large language model, calling a target resource generation application matched with the drawing resource type, generating drawing resources based on the resource demand parameter through the target resource generation application, and sending the drawing resources to a terminal;

And displaying drawing resources on the intelligent dialogue interface by the terminal, wherein the drawing resources are configured to target positions corresponding to the canvas operation interface according to the selection operation, and obtaining a corresponding design drawing.

The application provides a resource processing design method based on a large language model, which is applied to a terminal and comprises the following steps:

Responding to drawing resource inquiry information input by an intelligent dialogue interface in a drawing design interface, and sending the drawing resource inquiry information to a server corresponding to the drawing design interface; the server determines drawing resource types and resource demand parameters corresponding to the drawing resource inquiry information based on the large language model, invokes a target resource generation application matched with the drawing resource types, and generates drawing resources based on the resource demand parameters through the target resource generation application; the diagram design interface comprises an intelligent dialogue interface and a canvas operation interface;

And displaying drawing resources returned by the server on the intelligent dialogue interface, and responding to the selection operation in the drawing design interface, and setting the drawing resources corresponding to the selection operation to the target positions corresponding to the canvas operation interface to obtain the corresponding design drawing.

In a second aspect, the present application further provides a resource processing design device based on a large language model, which is applied to a server, and includes:

The drawing resource query information is input based on an intelligent dialogue interface in a drawing design interface, and the drawing design interface comprises the intelligent dialogue interface and a canvas operation interface;

The calling module is used for determining drawing resource types and resource demand parameters corresponding to drawing resource inquiry information based on the large language model, calling a target resource generation application matched with the drawing resource types, generating drawing resources based on the resource demand parameters through the target resource generation application, and sending the drawing resources to the terminal;

And the design module is used for enabling the terminal to display drawing resources on the intelligent dialogue interface, wherein the drawing resources are used for being configured to target positions corresponding to the canvas operation interface according to the selection operation, and a corresponding design diagram is obtained.

The application also provides a resource design processing device based on the large language model, which is applied to the terminal and comprises:

The input module is used for responding to the drawing resource inquiry information input by the intelligent dialogue interface in the drawing design interface and sending the drawing resource inquiry information to a server corresponding to the drawing design interface; the server determines drawing resource types and resource demand parameters corresponding to the drawing resource inquiry information based on the large language model, invokes a target resource generation application matched with the drawing resource types, and generates drawing resources based on the resource demand parameters through the target resource generation application; the diagram design interface comprises an intelligent dialogue interface and a canvas operation interface;

And the setting module is used for displaying drawing resources returned by the server on the intelligent dialogue interface, responding to the selection operation in the drawing design interface, and setting the drawing resources corresponding to the selection operation to the target positions corresponding to the canvas operation interface to obtain the corresponding design drawing.

In a third aspect, the present application also provides a computer device comprising a memory storing a computer program and a processor implementing the steps of the method described above when the processor executes the computer program.

In a fourth aspect, the present application also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of the method described above.

In a fifth aspect, the application also provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of the method described above.

The resource processing design method, the device, the computer equipment, the computer readable storage medium and the computer program product based on the large language model respond to the drawing resource inquiry information input by the intelligent dialogue interface in the drawing design interface through the terminal and send the drawing resource inquiry information to the server corresponding to the drawing design interface; the method comprises the steps that a server receives drawing resource inquiry information sent by a terminal, determines drawing resource types and resource demand parameters corresponding to the drawing resource inquiry information based on a large language model, calls a target resource generation application matched with the drawing resource types, generates drawing resources based on the resource demand parameters through the target resource generation application, and sends the drawing resources to the terminal; the terminal displays drawing resources returned by the server on the intelligent dialogue interface, responds to selection operation in the drawing design interface, sets the drawing resources corresponding to the selection operation to a target position corresponding to a canvas operation interface also in the drawing design interface, obtains a corresponding design drawing, is beneficial to directly obtaining the drawing resources on the intelligent dialogue interface in the same interface, sets the drawing resources in the intelligent dialogue interface to the canvas operation interface according to interactive operation between the intelligent dialogue interface and the canvas operation interface in the same interface, and in addition, when the drawing resources are obtained, identifies the drawing resource types through a large language model and determines corresponding resource generation application tools according to the drawing resource types, so that the accuracy of obtaining drawing resources meeting drawing requirements can be improved, the interaction between the large language model and various resource generation application tools and the interaction between the intelligent dialogue interface and the canvas operation interface are improved, and the manufacturing efficiency of the drawing design is improved well.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the related art, the drawings that are needed in the description of the embodiments of the present application or the related technologies will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other related drawings may be obtained according to these drawings without inventive effort to those of ordinary skill in the art.

FIG. 1 is an application environment diagram of a large language model-based resource process design methodology in one embodiment;

FIG. 2 is a flow diagram of a method of designing a resource process based on a large language model in one embodiment;

FIG. 3 is a flow chart of a resource processing design method based on a large language model in another embodiment;

FIG. 4 is a schematic diagram of a design interface of the diagram in one embodiment;

FIG. 5 is a schematic diagram of a drawing resource in one embodiment;

FIG. 6 is a flow diagram of an information graph in one embodiment;

FIG. 7 is a flow diagram of layout style generation in one embodiment;

FIG. 8 is a schematic illustration of a questionnaire in one embodiment;

FIG. 9 is a schematic diagram of a graphical time comparison of different systems in one embodiment;

FIG. 10 is a schematic diagram of an interactive sequence visualization of different systems in one embodiment;

FIG. 11 is a schematic diagram of a design quality assessment of the present application in one embodiment;

FIG. 12 is a diagram of graph quality assessment contrast for different systems in one embodiment;

FIG. 13 is a schematic diagram of a matting function in one embodiment;

FIG. 14 is a block diagram of a resource processing design apparatus based on a large language model in one embodiment;

FIG. 15 is a block diagram of a resource processing design apparatus based on a large language model in another embodiment;

FIG. 16 is an internal block diagram of a computer device in one embodiment;

Fig. 17 is an internal structural view of a computer device in one embodiment.

Detailed Description

The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.

The resource processing design method based on the large language model provided by the embodiment of the application can be applied to an application environment shown in fig. 1, wherein resources refer to drawing resources, and the resource processing design method comprises the steps of but not limited to design elements such as texts, pictures and icons. Wherein the terminal 102 communicates with the server 104 via a network. The data storage system may store data that the server 104 needs to process. The server 104 is configured to receive drawing resource query information sent by a terminal, where the drawing resource query information is input based on an intelligent dialogue interface in a drawing design interface, and the drawing design interface includes the intelligent dialogue interface and a canvas operation interface; and determining a drawing resource type and a resource demand parameter corresponding to the drawing resource query information based on the large language model, calling a target resource generation application matched with the drawing resource type, generating drawing resources based on the resource demand parameter through the target resource generation application, and sending the drawing resources to the terminal. The terminal 102 is configured to display drawing resources on the intelligent dialogue interface, where the drawing resources are configured to target positions corresponding to the canvas operation interface according to the selection operation, so as to obtain a corresponding design drawing. The terminal 102 is configured to respond to drawing resource query information input by the intelligent dialogue interface in the drawing design interface, and send the drawing resource query information to a server corresponding to the drawing design interface. The terminal 102 is configured to display drawing resources returned by the server on the intelligent dialogue interface, and respond to a selection operation in the drawing design interface, and set the drawing resources corresponding to the selection operation to a target position corresponding to the canvas operation interface, so as to obtain a corresponding design drawing. The data storage system may be integrated on the server 104 or may be located on a cloud or other network server. The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, internet of things devices and portable wearable devices, where the internet of things devices may be smart speakers, smart televisions, smart air conditioners, smart vehicle devices, projection devices, and the like. The portable wearable device may be a smart watch, smart bracelet, headset, or the like. The head-mounted device may be a Virtual Reality (VR) device, an augmented Reality (Augmented Reality, AR) device, smart glasses, or the like. The server 104 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing cloud computing services.

In an exemplary embodiment, as shown in fig. 2, a resource processing design method based on a large language model is provided, and an example of application of the method to the server in fig. 1 is described, which includes the following steps S200 to S204. Wherein:

Step S200, drawing resource inquiry information sent by the terminal is received, wherein the drawing resource inquiry information is input based on an intelligent dialogue interface in a drawing design interface, and the drawing design interface comprises the intelligent dialogue interface and a canvas operation interface.

The drawing resource inquiry information refers to the input of session content about drawing resource acquisition in the intelligent dialogue interface; the drawing resource inquiry information comprises inquiry information of a certain type of resource, such as first inquiring about the session content of a theme picture, and also comprises inquiry information for repeatedly inquiring about the same type of resource, such as when drawing resources about the theme picture are acquired, the theme picture returned in the intelligent dialogue interface is considered to be insufficient, the theme picture needs to be re-acquired, but the theme picture is not described in the dialogue content of the re-acquired resource, and only're-acquisition' is input, so that the drawing resource inquiry information comprises the session content about the acquired theme picture before're-acquisition', and the large language model identifies the user intention according to the dialogue content of the context.

The diagram design interface is an interface for interaction according to the intelligent dialogue interface and the canvas operation interface; the intelligent dialogue interface and the canvas operation interface are both in the diagram design interface, namely, the intelligent dialogue interface and the canvas operation interface are part of the diagram design interface, as shown in fig. 4, fig. 4 is the diagram design interface, the left side in fig. 4 is the intelligent dialogue interface, and the right side is the canvas operation interface; the intelligent dialogue interface and the canvas operation interface can be mutually independent, interaction can exist, namely the intelligent dialogue interface can only conduct intelligent dialogue, the canvas operation interface only conducts picture design adjustment, the design adjustment is similar to canvas operation tools in the prior art, visual attributes such as positions, sizes, colors, font styles, font sizes and the like of design elements in the canvas operation interface are changed, the design elements comprise texts, pictures, icons and the like, the design elements are drawing resources, and the drawing resources can be dragged from the intelligent dialogue interface to the canvas operation interface to conduct design adjustment, so that interaction between the intelligent dialogue interface and the canvas operation interface is achieved.

The intelligent dialogue interface refers to an interface for carrying out intelligent dialogue with the virtual robot; the user intention can be input in the intelligent dialogue interface, so that the terminal responds to the user intention input in the intelligent dialogue interface, the user intention is sent to the server, the server recognizes the user intention based on the large language model, feedback information is generated, and the feedback information is sent to the terminal, so that the terminal displays the feedback information on the intelligent dialogue interface; user intent includes, but is not limited to, drawing resource query information.

The canvas operation interface refers to an operation interface for carrying out design adjustment on drawing resources; visual properties such as position, size, color, font style, font size and the like of each drawing resource can be modified in the canvas operation interface, and intelligent tools for designing and adjusting different visual properties, such as but not limited to an image clipping tool, are included in the canvas operation interface; canvas dimensions of the canvas operator interface may be adjusted, wherein the adjusted dimension scale includes, but is not limited to, 16: 9. 9: 16. 3: 2. 2: 3. 4: 3. 3:4, etc. Drawing resources refer to resources required for making a design drawing, including but not limited to text, pictures, icons, layout styles; text includes, but is not limited to, title text and paragraph content text of different levels of design, pictures include, but are not limited to, background pictures and theme pictures of design, and layout style refers to the layout type, such as a wavy layout, of which the design drawings are made.

The intelligent dialogue interface and the canvas operation interface can jointly construct a graph design interface, intelligent dialogue can be conducted in the same graph design interface, graph design can be conducted, a terminal can respond to drawing resource inquiry information input in the intelligent dialogue interface and send the drawing resource inquiry information to a server, and the server receives the drawing resource inquiry information sent by the terminal and provides data for subsequent large language model-based user intention identification.

For example, other dialog contents may be input in the intelligent dialog interface, not limited to sending drawing resource query information, and whether the specific dialog contents are intents for querying drawing resources may be identified by the server based on a large language model. When other dialogue contents are input in the intelligent dialogue interface, the intelligent dialogue can be equivalent to the intelligent dialogue with the virtual robot, the server can recognize the dialogue contents input in the intelligent dialogue interface based on a large language model and generate corresponding feedback information, so that the terminal returns the feedback information in the intelligent dialogue interface, and the intelligent dialogue between the virtual robot and a user is realized.

Step S202, based on the large language model, determining a drawing resource type and a resource demand parameter corresponding to drawing resource inquiry information, calling a target resource generation application matched with the drawing resource type, generating drawing resources based on the resource demand parameter through the target resource generation application, and sending the drawing resources to the terminal.

The large language model is an artificial intelligent model, aims to understand and generate human language, is also called a large language model, is trained on a large amount of text data, and can perform a wide range of tasks, including text summarization, translation, emotion analysis and the like; large language models include, but are not limited to ChatGPT (CHAT GENERATIVE PRE-trained Transformer). Drawing resource types refer to resource types for making a design drawing, including, but not limited to, main title text, theme pictures, background pictures, paragraph title text, paragraph content text, icons, and the like.

The resource demand parameter refers to the demand parameter of the resource required by the production design drawing, such as "an lovely cat" converted into the character string data type; the resource requirement parameter may include a resource production requirement parameter and a resource single return number, where the resource production requirement parameter refers to an attribute requirement of the resource, such as "an lovely cat", and the resource single return number refers to a number of resources each time the resource is returned to the terminal, such as "4". The target resource generating application refers to an application tool for generating drawing resources; the number of the resource generating applications corresponding to the various drawing resource types can be one or two or more, and the number of the resource generating applications corresponding to the various drawing resource types is not limited.

The server may identify an intent of the drawing resource query information based on the large language model, determine a drawing resource type and a resource requirement parameter according to the identified intent, and call a target resource generation application matched with the drawing resource type according to a configuration relation between the drawing resource type and the resource generation application, wherein for the same drawing resource type, a plurality of different resource generation applications may be configured, each of the configured plurality of resource generation applications may generate drawing resources corresponding to the drawing resource type, and for different resource generation applications, there is a difference in functional characteristics, the large language model may determine a resource generation application most matched with the drawing resource type corresponding to the drawing resource query information according to the intent of the drawing resource query information, and call the target resource generation application as the target resource generation application, convert the resource requirement parameter into a resource requirement command in a matching format through the target resource generation application, generate drawing resources corresponding to the drawing resource type based on the resource requirement command, and send the drawing resources to the terminal for display, thereby realizing interaction between the large language model and application tools of the drawing resources.

In one embodiment, if the large language model identifies multiple drawing resource types and resource demand parameters corresponding to the drawing resource types from the intent of the drawing resource query information, the target resource generating application matched with the drawing resource types can be respectively called, drawing resources corresponding to the drawing resource types are generated based on the resource demand parameters corresponding to the drawing resource types through the target resource generating application corresponding to the drawing resource types, and the drawing resources corresponding to the drawing resource types are sent to the terminal, so that the terminal can display the drawing resources of the drawing resource types in the intelligent dialogue interface.

Step S204, the terminal displays drawing resources on the intelligent dialogue interface, wherein the drawing resources are configured to target positions corresponding to the canvas operation interface according to the selection operation, and a corresponding design diagram is obtained.

The selection operation refers to operations performed in the graph design interface, including a click operation and a drag-and-drop operation. The target position refers to a position corresponding to when the selection operation is terminated. The design diagram refers to a diagram which is correspondingly designed by arranging drawing resources under the current selection operation; the diagram content in the design diagram includes, but is not limited to, pictures, texts, icons and the like; the types of designs include, but are not limited to, informational designs, which may be a variety of other types of artistic designs. An information graph is a visual representation that conveys information in a visually attractive manner by combining text and various graphical elements, the purpose of which is to convey information in a clear and concise manner.

The server returns the drawing resources to the terminal, so that the drawing resources are displayed in the intelligent dialogue interface by the terminal, the intelligent dialogue interface and the canvas operation interface are positioned in the same interface, drawing resources required by drawing design can be directly obtained in the intelligent dialogue interface when the canvas operation interface performs drawing design, drawing elements meeting requirements can be quickly obtained without searching the drawing resources from other places, and drawing efficiency is effectively improved.

According to the resource processing design method based on the large language model, the terminal responds to drawing resource inquiry information input by an intelligent dialogue interface in a drawing design interface, and the drawing resource inquiry information is sent to a server corresponding to the drawing design interface; the method comprises the steps that a server receives drawing resource inquiry information sent by a terminal, determines drawing resource types and resource demand parameters corresponding to the drawing resource inquiry information based on a large language model, calls a target resource generation application matched with the drawing resource types, generates drawing resources based on the resource demand parameters through the target resource generation application, and sends the drawing resources to the terminal; the terminal displays drawing resources returned by the server on the intelligent dialogue interface, responds to selection operation in the drawing design interface, sets the drawing resources corresponding to the selection operation to a target position corresponding to a canvas operation interface also in the drawing design interface, obtains a corresponding design drawing, is beneficial to directly obtaining the drawing resources on the intelligent dialogue interface in the same interface, sets the drawing resources in the intelligent dialogue interface to the canvas operation interface according to interactive operation between the intelligent dialogue interface and the canvas operation interface in the same interface, and in addition, when the drawing resources are obtained, identifies the drawing resource types through a large language model and determines corresponding resource generation application tools according to the drawing resource types, so that the accuracy of obtaining drawing resources meeting drawing requirements can be improved, the interaction between the large language model and various resource generation application tools and the interaction between the intelligent dialogue interface and the canvas operation interface are improved, and the manufacturing efficiency of the drawing design is improved well.

In some embodiments, generating drawing resources based on the resource demand parameters by the target resource generation application in step S202 includes:

In step S300, if the target resource generating application is a picture generating application, the picture generating application converts the resource requirement parameters into a resource requirement command in a matching format, and generates a picture based on the resource requirement command.

Wherein, the picture generation application refers to an application tool for generating pictures meeting specified requirements, including but not limited to Stable Diffusion XL 1.0.0; the picture generation application comprises, but is not limited to, a theme picture generation application and a background picture generation application, wherein the theme picture generation application refers to an application tool for generating pictures meeting the requirements of a specified theme, the background picture generation application refers to an application tool for generating pictures meeting the requirements of the background, the theme picture generation application and the background picture generation application can be the same or different, and the number of the theme picture generation application and the background picture generation application can be one, two or more, and the number of the application tools is not limited; different picture generation applications may be configured for different picture types in the design drawing. The resource demand command refers to a command that can be directly executed by the target resource generating application; for different resource generating applications, when executing the resource corresponding to the generated resource demand parameter, the resource demand parameter can be converted into a resource demand command matched with the format of the resource demand command.

For example, if the drawing resource type is a picture type such as a theme picture, a background picture, etc., the target resource generating application is a picture generating application, and the resource demand parameter can be converted into a resource demand command in a matching format by the picture generating application, and then a picture conforming to the resource demand command requirement is generated based on the resource demand command. For pictures of different semantic types, different picture generation applications may also be configured, for example:

If the drawing resource type is a theme picture, a theme picture generation application matched with the theme picture is used as a target resource generation application, the theme picture generation application is called, the resource demand parameters are converted into a resource demand command in a theme picture generation application matching format through the theme picture generation application, and the theme picture is generated based on the resource demand command; and if the drawing resource type is the background picture, using a background picture generating application matched with the background picture as a target resource generating application, calling the background picture generating application, converting the resource demand parameter into a resource demand command in a background picture generating application matching format through the background picture generating application, and generating the background picture based on the resource demand command.

For example, if the large language model may identify that the large language model is suitable for the background picture if the large language model is used for generating the melted iceberg, and determine that the drawing resource type is the background picture, call the background picture generation application, and use Stable Diffusion XL 1.0.0 as the background picture generation application, convert the resource requirement parameter of the "generate melted iceberg" into a resource requirement command in a matching format based on the background picture generation application, that is, "{ generate the background picture of the melted iceberg }, and the background picture generation application generates the background picture conforming to the semantic meaning of the" melted iceberg }, based on the resource requirement command of "{ generate the background picture of the melted iceberg }.

Step S302, if the target resource generating application is a text generating application, converting the resource demand parameters into resource demand commands in a matching format through the text generating application, and generating texts based on the resource demand commands, wherein the texts comprise icon keywords; and acquiring the icon matched with the icon keyword from the database, or calling a picture generation application, and generating the icon matched with the icon keyword based on the icon keyword through the picture generation application.

Where a text generation application refers to an application tool that generates text meeting specified requirements, including, but not limited to ChatGPT. Text includes, but is not limited to, topic title text, paragraph content text, and icon keywords; the texts can be integrated in the form of an array or in other self-defined forms, and for the form of array integration, each item in the array can comprise a sub-key such as a theme title, an icon keyword, a paragraph title, a paragraph content and the like, the text (such as the theme title) related to the theme can be separated from the text (such as the icon keyword, the paragraph title and the paragraph content) related to the paragraph, the text can be packaged into a text resource package, and the text resource package is converted into a JSON format or other self-defined formats. The icon keyword refers to text information representing icon attribute information; an icon keyword may be a word or phrase that matches the semantics of a subject title or paragraph title that may be used to retrieve an icon that matches the semantics of a subject title or paragraph title. Icons include, but are not limited to, icons related to paragraph titles.

For example, if the drawing resource type is a text type such as a title text, a paragraph content text, etc., the target resource generating application is a text generating application, the text generating process may be implemented by an application tool formed by sequentially connecting different components, and the application tool formed by sequentially connecting different components may include a text generating component and an icon acquiring component; the text generation component can convert the resource demand parameters into resource demand commands in a matching format by calling the text generation application, generate texts conforming to the resource demand command requirements based on the resource demand commands, integrate the generated texts in an array form to generate text resource packages, and format the text resource packages into JSON objects as structured text resource packages so as to simplify data analysis and realize seamless integration into a canvas operation interface when the texts are displayed on the intelligent dialogue interface and set on the canvas operation interface.

In addition, the generated text comprises icon keywords, icons matched with the icon keywords semantically can be acquired in a database based on the icon keywords through an icon acquisition component, the database comprises but is not limited to Iconify the database, the format of the acquired icons can be SVG icons, the process of searching the icons matched with the icon keywords semantically in the database is expandable, and the icons can be synchronized or added into an additional online icon database along with updating of the database; or calling a picture generation application through an icon acquisition component, converting the icon keywords into icon generation commands in a matching format based on the picture generation application, and generating icons which are matched with the icon keywords in a semantic mode based on the icon generation commands.

For example, if "generate text related to iceberg", the large language model may recognize that text related to iceberg content is generated, may generate a JSON format text resource package related to iceberg according to ChatGPT application itself, where the text resource package includes but is not limited to a theme text, an icon keyword, a paragraph title text, and a paragraph content text related to iceberg, and call Iconify database to search for an icon that matches with the icon keyword semantic, so as to obtain an icon in SVG format.

In step S304, if the target resource generating application is a layout style generating application, the resource demand parameters are converted into resource demand commands in a matching format by the layout style generating application, and the layout style is generated based on the resource demand commands.

Wherein, the layout style generation application refers to an application tool for generating a layout style meeting specified requirements, including but not limited to GPT-4; different placement styles may correspond to different DSL (domain-specific language) document packages. The generated layout style is a DSL script written according to the DSL document package.

For example, if the drawing resource type is a layout style, the target resource generating application is a layout style generating application, the layout style generating application may generate a DSL document package with a hierarchical layout, and write the DSL document package into a corresponding DSL script based on the DSL document package, where the DSL document package includes content for guiding writing of the DSL script, and the hierarchical layout refers to a visual layout of drawing resources such as a theme title, a paragraph content, an icon, a theme picture, a background picture, etc., where the DSL adopts a tuple-based hierarchical structure, including, but not limited to, an interface coordinate node characterizing a hierarchical layout frame and a container node for packaging the drawing resources, and for different resource demand commands, the generated interface coordinate node and container node are different, and the layout style generating application may generate a plurality of different DSL document packages based on the same resource demand command, and the generated plurality of DSL document packages are all layout styles that are in accordance with the semantics of the resource demand command. In more detail, the DSL document package includes layout constraint rules corresponding to each interface coordinate node and each container node, for example, "minimizing overlap between different drawing resources (also being design elements)", and "only allowing one icon, one paragraph title and one paragraph content element in one container group (corresponding to a container node)", etc., the server may write a corresponding DSL script based on the DSL document package, the terminal may parse the DSL script into a tree structure, and render the tree structure into a visual layout displayed in the interface, and if there is no drawing resource in the canvas operation interface at this time, each layout area in the visual layout (that is, an area set by a corresponding container node) is also free of content.

For example, if "generate a waveform layout", the large language model may identify that the layout style generation application is invoked, and may use GPT-4 as the layout style generation application, and convert the resource requirement parameter of "generate a waveform layout" into a resource requirement command in a matching format through the layout style generation application, and generate a DSL document package according to the resource requirement command semantically, and write a corresponding DSL script, that is, a layout style, based on the DSL document package.

In the embodiment, the matched target resource generation application is correspondingly called through different drawing resource types, so that drawing resources meeting the requirements of users are generated in a targeted manner, the collection time of the drawing resources in the drawing design process is greatly shortened, the drawing resources such as pictures and layout patterns meeting the intention of the users are generated rapidly and accurately, the creativity of the users is promoted, and the efficiency of drawing design of the users is greatly improved.

In some embodiments, the method for processing resource design based on large language model is applied to a server, and further comprises:

Step S400, receiving a resource modification request sent by a terminal, wherein the resource modification request is input through an intelligent dialogue interface.

Step S402, based on the large language model, the modified resource identification information and the modified standard parameters corresponding to the resource modification request are determined, the resource modification application is called, updated resources are generated through the resource modification application based on the modified standard parameters, and the modified resource identification information and the updated resources are sent to the terminal.

Step S404, the terminal replaces the resource corresponding to the modified resource identification information with the updated resource.

Wherein the resource modification request refers to a request for performing design adjustment on drawing resources in the design drawing, and the request comprises request modification dialogue information; the request to modify dialog information refers to user modification intention dialog information entered at the intelligent dialog interface. Modifying the information of the resource identification information to identify the information of the modified resource; if the style of the theme picture is modified, the modification resource identification information is information which can identify the picture type of the theme picture in the canvas operation interface (namely, the theme picture), the position of the canvas operation interface and the like.

Modifying the standard parameters refers to modifying the demand parameters; for example, "change the theme picture style into the XXX style", the "XXX style of the theme picture" is a modification standard parameter, in addition, each picture returned in the intelligent dialogue interface has a corresponding number, and the number information is also displayed together in the intelligent interface, when a specified picture needs to be modified (the modification is not limited to the modification of the style), the "change picture 1234 into the XXX style", the picture 1234 is the picture number can be directly input. The resource modification application refers to an application tool matched when modifying the resource, and may be a text generation application, a picture generation application, a layout generation application or other editing tool, specifically determined by the modification intention of the resource modification request, that is, the resource modification application is an application tool matched with the modification intention of the resource modification request. Updated resources refer to drawing resources that meet the requirements of modifying standard parameters.

The modification of drawing resources of the design drawing in the canvas operation interface can be realized through the interaction between the large language model and the application tool. The method comprises the steps of receiving a resource modification request sent by a terminal, identifying modification resource identification information and modification standard parameters for the resource modification request based on a large language model, determining a modification resource type (namely the type of modified drawing resource) by the resource modification identification information, determining a resource modification application based on the modification resource type, calling the resource modification application, converting the modification standard parameters into modification standard commands in a matching format by the resource modification application based on the resource modification application, generating update resources conforming to the modification standard commands based on the modification standard commands, sending the update resources and the modification resource identification information to the terminal, enabling the terminal to display the update resources on an intelligent dialogue interface, and replacing the resources for which the modification resource identification information is in a canvas operation interface with the update resources.

For example, the modification resource request may be "change the theme picture style into the watercolor style", the large language model may identify that the style of the theme picture is modified, and determine the picture number corresponding to the modification resource identification information, the identification information of the position in the canvas operation interface, and the like, and may take the "generate the theme picture of the watercolor style" as the modification standard parameter, then call the picture modification application that is the best match with the intention of the "generate the theme picture of the watercolor style", and may take the image operation tool based on the InstructPix Pix deep learning model as the picture modification application, convert the modification standard parameter into the modification standard command that is matched with the format through the picture modification application, generate the update theme picture (i.e. update resource) that conforms to the modification standard command based on the modification standard command, and may send the update theme picture and the modification resource identification information to the terminal so that the terminal displays the update theme picture on the intelligent dialogue interface, and replace the original theme picture with the update theme picture based on the modification resource identification information.

In one embodiment, resource modification tools, such as an image clipping tool and a text attribute (including but not limited to color, size and the like) modification tool, are also configured in the canvas operation interface, and the drawing resources in the design drawing can be directly adjusted by clicking and selecting the resource modification tool in the canvas operation interface through a mouse, wherein the adjustment operation includes, but is not limited to, dragging a layout frame, adjusting the font size, adjusting the font color, changing the background color and transparency of a mask of the text frame, changing the color and thickness of an icon and the like, that is, the canvas operation interface also has the operation function of a drawing design canvas, and besides the design processing of the design drawing according to the interaction between the intelligent dialogue interface and the canvas operation interface, the design processing of the design drawing can be realized by operating the canvas operation interface, so that the operation function of the drawing design is enriched, and the convenience and the flexibility of the drawing design are improved.

In the embodiment, through the interaction between the large language model and the resource modification application, update operations such as modification, regeneration and the like of the modification resource are realized, thereby being beneficial to quickly and efficiently acquiring the drawing resource required in the current drawing design process, being beneficial to improving the modification operation in the drawing design processing process, enriching the operation function of the drawing design and simultaneously improving the drawing efficiency better.

In an exemplary embodiment, as shown in fig. 3, a resource design processing method based on a large language model is provided, and an example of application of the method to the terminal in fig. 1 is described, including the following steps S500 to S502. Wherein:

Step S500, drawing resource inquiry information input by an intelligent dialogue interface in a drawing design interface is responded, and the drawing resource inquiry information is sent to a server corresponding to the drawing design interface; the server determines drawing resource types and resource demand parameters corresponding to the drawing resource inquiry information based on the large language model, invokes a target resource generation application matched with the drawing resource types, and generates drawing resources based on the resource demand parameters through the target resource generation application; the diagram design interface includes an intelligent dialog interface and a canvas operation interface.

The intelligent dialogue interface can carry out intelligent dialogue with the virtual robot, and can also directly input information to be searched in the intelligent dialogue interface; the canvas operation interface can be used for carrying out diagram design, drawing resources can be directly obtained in the intelligent dialogue interface through intelligent dialogue in the process of carrying out diagram design, and the adjustment of the drawing resources is realized according to the interaction between the intelligent dialogue interface and the canvas operation interface; in order to obtain drawing resources meeting the design requirements of the drawing more quickly, drawing resource inquiry information can be input into the intelligent dialogue interface, and the terminal responds to the drawing resource inquiry information input by the intelligent dialogue interface and sends the drawing resource inquiry information to the terminal; the terminal can identify the intention of drawing resource inquiry information based on a large language model, and determine the drawing resource type and resource demand parameters based on the identified intention, and because the same drawing resource type can be configured with a plurality of resource generating applications for generating drawing resource type resources, the large language model can determine the resource generating application which is most matched with the drawing resource type based on the intention of the drawing resource inquiry information, and take the most matched resource generating application as a target resource generating application, call the target resource generating application, convert the resource demand parameters into resource demand commands in a matching format through the target resource generating application, and generate the drawing resources which are matched semantically based on the resource demand commands.

In one embodiment, if the drawing design is not performed and drawing resources required by the drawing design do not need to be searched, dialog content can be input into the intelligent dialog interface only, intelligent dialog is performed with the virtual robot, and the terminal can respond to the dialog content input into the intelligent dialog interface and send the dialog content to the server; the server recognizes the dialogue content based on the large language model, generates feedback information and sends the feedback information to the terminal; and the terminal displays the received feedback information in the intelligent dialogue interface, thereby realizing intelligent dialogue.

Step S502, drawing resources returned by the server are displayed on the intelligent dialogue interface, and in response to selection operation in the drawing design, drawing resources corresponding to the selection operation are set to target positions corresponding to the canvas operation interface, so that a corresponding design drawing is obtained.

By way of example, drawing resources returned by the server can be displayed on the intelligent dialogue interface, target selection resources corresponding to clicking operations are determined in response to clicking operations in the intelligent dialogue interface, the target selection resources are dragged and dropped to target positions corresponding to the canvas operation interface in response to drag-and-drop operations of the drawing design interface, and interaction between the intelligent dialogue interface and the canvas operation interface is achieved. For example, if the drawing resource is a theme picture, one of the theme pictures returned in the intelligent dialogue interface can be clicked and selected by the mouse, the selected theme picture is dragged to a position which is desired to be placed in the canvas operation interface based on the mouse, and after the selected theme picture is placed in the canvas operation interface, if the theme picture is desired to be adjusted, the position, the size and the like of the theme picture can be adjusted on the canvas operation interface directly.

Illustratively, in one embodiment, the resource selection information may also be directly input in the intelligent dialogue interface, and the terminal sends the resource selection information input in response to the intelligent dialogue interface to the server; the server identifies the intention of the resource selection information based on the large language model, determines the target selection resource identification information based on the identified intention, and returns the target selection resource identification information to the terminal; the terminal sets the resource corresponding to the target selection resource identification information to a preset position in the canvas operation interface to obtain a corresponding design drawing, so that the operation of designing the design drawing in the canvas operation interface without operating a mouse is realized, namely, the automatic modification operation on drawing resources in the canvas operation interface can be realized. For example, the returned theme pictures are displayed in the intelligent dialog interface at present, and the dialog content with the intention of approximately "selecting XXX (number) pictures" may be input at the terminal, and the terminal may send the dialog content to the server, so that the server identifies the target selection theme picture identification information corresponding to the dialog content based on the large language model, and sends the identified target selection theme picture identification information to the terminal, so that the terminal may set the theme picture corresponding to the target selection theme picture identification information at a preset position.

According to the resource design processing method based on the large language model, the terminal responds to drawing resource inquiry information input by an intelligent dialogue interface in a drawing design interface, and the drawing resource inquiry information is sent to a server corresponding to the drawing design interface; the method comprises the steps that a server receives drawing resource inquiry information sent by a terminal, determines drawing resource types and resource demand parameters corresponding to the drawing resource inquiry information based on a large language model, calls a target resource generation application matched with the drawing resource types, generates drawing resources based on the resource demand parameters through the target resource generation application, and sends the drawing resources to the terminal; the terminal displays drawing resources returned by the server on the intelligent dialogue interface, responds to selection operation in the drawing design interface, sets the drawing resources corresponding to the selection operation to a target position corresponding to a canvas operation interface also in the drawing design interface, obtains a corresponding design drawing, is beneficial to directly obtaining the drawing resources on the intelligent dialogue interface in the same interface, sets the drawing resources in the intelligent dialogue interface to the canvas operation interface according to interactive operation between the intelligent dialogue interface and the canvas operation interface in the same interface, and in addition, when the drawing resources are obtained, identifies the drawing resource types through a large language model and determines corresponding resource generation application tools according to the drawing resource types, so that the accuracy of obtaining drawing resources meeting drawing requirements can be improved, the interaction between the large language model and various resource generation application tools and the interaction between the intelligent dialogue interface and the canvas operation interface are improved, and the manufacturing efficiency of the drawing design is improved well.

In some embodiments, the selection operation includes a click operation and a drag-and-drop operation; in step S502, in response to a selection operation in the graphic design interface, a drawing resource corresponding to the selection operation is set to a target position corresponding to the canvas operation interface, so as to obtain a corresponding design drawing, which includes:

And step S600, determining a target selection resource from the drawing resource in response to clicking operation in the intelligent dialogue interface, and dragging and dropping the target selection resource to a corresponding target position in the canvas operation interface in response to dragging and dropping operation of the target selection resource in the drawing design interface to obtain a corresponding design drawing.

The clicking operation refers to clicking operation performed by a mouse. The drag-and-drop operation refers to a drag operation performed by a mouse. The target selection resource refers to the drawing resource selected by the clicking operation. The target position refers to a position where the target resource is placed when dragged and dropped by the mouse.

The terminal can determine the target selection resource based on the click operation of the intelligent dialogue interface and drag and drop the target selection resource to the corresponding position in the canvas operation interface, and after the target selection resource is dragged and dropped to the canvas operation interface, the drawing resource of the same type can be obtained again in the intelligent dialogue interface, or drawing resources of other types can be obtained, a new round of selection operation is performed, drawing resources corresponding to the new round of selection operation are adjusted in the canvas operation interface, and the like until a design drawing meeting the requirement is obtained. In addition, in the process of manufacturing the design drawing, drawing resources in the design drawing can be flexibly adjusted based on operation functions configured by the canvas operation interface until the requirements are met, and finally, the design drawing meeting the requirements can be stored.

In one embodiment, the pictures in the intelligent dialog interface can be further scratched, and the images obtained by the scratching can be dragged to a canvas operation interface, as shown in fig. 13, when the terminal responds to a pause operation on a target picture displayed in the intelligent dialog interface, a scratching function button (as shown in (a) in fig. 13) is displayed above the target picture, wherein the scratching function button comprises, but is not limited to, a 'point' button (which can be called a first type scratching button) for determining a scratching object in a point form, a 'line' button (which can be called a second type scratching button) for determining a scratching object in a line form, a 'frame' button (which can be called a third type scratching button) for determining a scratching object in a rectangular frame form, a 'delete' button (which can be called a delete button) for deleting the determining the scratching object, and a 'determine' button (which can be called a determine) for the scratching object; the terminal responds to clicking operation of selecting the matting function button, if the matting function button is a first form matting button, then responds to an object specified point corresponding to clicking operation of the target picture, and determines the target matting object based on the object specified point, specifically as shown in (b) in fig. 13, wherein the number of the object specified points is not limited; if the matting function button is a second form matting button, responding to an object specifying line corresponding to the target picture, and determining a target matting object based on the object specifying line, wherein the method can be specifically shown as (c) in fig. 13; if the matting function button is a matting button in the third form, responding to a corresponding target block diagram on the target picture, and determining a target matting object based on the target block diagram, wherein the target matting object can be specifically shown as (d) in fig. 13; after the terminal responds to the click operation of the determining button, the terminal responds to the drag-and-drop operation of the target image-matting object, drags and drops the target image-matting object from the intelligent dialogue interface to the target position corresponding to the canvas operation interface to obtain a corresponding design image, and the terminal can not set the determining button, directly responds to the drag-and-drop operation of the target image-matting object after determining the target image-matting object, drags and drops the target image-matting object to the target position corresponding to the canvas operation interface to obtain a corresponding design image, determines the setting of the button, and can improve the stability of the image-matting object.

In the embodiment, through the interaction between the intelligent dialogue interface and the canvas operation interface, the efficient manufacturing of the design drawing is realized on the basis that the intelligent dialogue interface rapidly acquires drawing resources, so that the manufacturing efficiency of the design drawing is improved better.

In some embodiments, the method for processing resource design based on a large language model is applied to a terminal, and further includes:

And step S700, the resource selection information is sent to the server in response to the resource selection information input in the intelligent dialogue interface, so that the server identifies target selection resource identification information corresponding to the resource selection information based on the large language model.

Step S702, receiving target selection resource identification information sent by a server, and configuring target selection resources corresponding to the target selection resource identification information at preset positions in a canvas operation interface based on the target selection resource identification information to obtain a corresponding design drawing.

Wherein the resource selection information refers to the session content of the selection designated drawing resource, such as "selection XXX (number) picture", which is input at the intelligent dialog interface. The target selection resource identification information refers to information for identifying the drawing resource selected. The preset position refers to a preset arrangement position; if the design diagram in the canvas operation interface selects a layout style, and the layout style has an arrangement area which accords with the type of the target selection resource, the preset position is the position corresponding to the arrangement area, and if the design diagram in the canvas operation interface does not use the arrangement style, the preset position is a preset fixed default position.

For example, in order to further improve drawing efficiency and flexibility of drawing production, the drawing design function may be enriched, resource selection information may be directly input in the intelligent dialogue interface, and the terminal may send resource selection information in response to the input of the intelligent dialogue interface to the server; the server identifies the target selection resource identification information corresponding to the resource selection information based on the large language model, and returns the target selection resource identification information to the terminal; the terminal can set the target selection resource corresponding to the target selection resource identification information to a preset fixed default position when the layout style is not used in the canvas operation interface, and further can adjust the target selection resource in the canvas operation interface through a mouse; the terminal can also adaptively adjust the target selection resource to the matched arrangement area if the arrangement area matched with the target selection resource type in the arrangement pattern is empty when the arrangement pattern is used in the canvas operation interface.

In the embodiment, the resource selection information is input into the intelligent dialogue interface, so that the automatic operation of setting the target selection resource into the canvas operation interface is realized according to the resource selection information, the functions of making the design diagram are enriched, and the flexibility of making the diagram is improved.

step S800, if drawing resources exist in the canvas operation interface, layout resource inquiry information is input in the intelligent dialogue interface, and the layout resource inquiry information is sent to a server; the server determines the layout resource demand parameters corresponding to the layout resource inquiry information based on the large language model, invokes the layout pattern generation application, and generates the layout pattern based on the layout resource demand parameters through the layout pattern generation application.

Step S802, displaying the layout style returned by the server on the intelligent dialogue interface, responding to the selection operation in the graphic design interface, and carrying out self-adaptive layout adjustment on the drawing resources in the canvas operation interface based on the layout style corresponding to the selection operation to obtain the design graphic corresponding to the layout style.

Wherein, the layout resource inquiry information refers to inputting dialog content for acquiring a layout style in the intelligent dialog interface, such as but not limited to "generating a wave-shaped layout", or "generating a layout comprising a main title, a theme picture and three paragraphs". The layout resource demand parameter refers to demand parameters of content layout style of the production design drawing, such as 'a wave-shaped layout' converted into character string data type, and the self-adaptive layout setting refers to the matching layout area in the layout style, which is adaptively adjusted by drawing resources in the drawing operation interface; for example, the design diagram in the canvas operation interface comprises theme pictures, the layout pattern comprises layout areas of the theme pictures, and after responding to the layout pattern corresponding to the selection operation, the theme pictures in the canvas operation interface are adaptively adjusted to the layout areas matched with the theme pictures.

For enriching the graph design function, the reference direction of more graph designs is provided for a user, the drawing efficiency is improved, and when drawing resources exist in the canvas operation interface, the terminal can respond to dialogue content which belongs to layout resource inquiry information and is input in the intelligent dialogue interface when the existing drawing resources comprise at least one of pictures, texts and icons, and the dialogue content is sent to a server; the server identifies the intention of the layout resource inquiry information based on the large language model, determines the layout resource demand parameters based on the identified intention, invokes a layout pattern generation application which is matched with the identification intention, converts the layout resource demand parameters into layout resource demand commands in a matching format through the layout pattern generation application, generates semantically matched layout patterns based on the layout demand commands, and the generated layout patterns can be characterized by DSL scripts written in a predefined DSL document package.

Further, the server returns the layout style to the terminal, the terminal can parse the layout style belonging to the DSL script into a corresponding tree structure based on the parser, render the tree structure into a corresponding visual layout style, display the visual layout style to the intelligent dialogue interface, respond to the click operation of the intelligent dialogue interface, determine a target layout style corresponding to the click operation, parse the target layout style to obtain a corresponding tree structure, render the tree structure to the canvas operation interface, and adaptively set various drawing materials in the canvas operation interface to matching arrangement areas in the rendered layout style to obtain a corresponding design drawing.

In one embodiment, if the drawing resources existing in the canvas operation interface include the local style, the terminal responds to the selection operation of selecting the layout style in the intelligent dialogue interface again, then the layout style currently existing in the canvas operation interface is replaced by the layout style corresponding to the reselection operation, and the drawing resources existing in the canvas operation interface are adaptively adjusted according to the layout style corresponding to the reselection operation, so that the adjustment time of the drawing resources is shortened well, and the efficiency of drawing design is improved.

For example, as shown in fig. 7, the user may input "may design a waveform layout" layout resource query information on the intelligent dialog interface, the server may identify a layout resource requirement parameter corresponding to the layout resource query information based on the large language model, call a matched layout style generation application (i.e. GPT-4 in fig. 7), generate a layout style (i.e. DSL script in fig. 7) based on the layout resource requirement parameter through the layout style generation application, and return the layout style to the terminal, so that the terminal parses the layout style belonging to the DSL script through the parser to obtain a corresponding tree structure, and render the tree structure into a corresponding visualized layout style (i.e. the layout style of the last step in the flowchart in fig. 7) based on the cartograph.

In the above embodiment, when the drawing resources exist in the canvas operation interface, the intelligent dialog interface inquires the layout style resources, selects the layout style returned by the server from the intelligent dialog interface, applies the selected layout style to the canvas operation interface, realizes the self-adaptive layout adjustment of the drawing resources existing in the canvas operation interface in the selected layout style, provides rich layout styles for the drawing design, enriches the function of the drawing design, and well shortens the adjustment time of the drawing resources based on the self-adaptive layout adjustment of the layout style, thereby improving the efficiency of the drawing design.

In some embodiments, the drawing resource type can be added according to actual requirements, and the resource generating application corresponding to the configuration drawing resource type is added, so that more refined resource collection or generation is realized, and the accuracy of drawing resource searching is improved.

In some embodiments, the large language model is not limited specifically, that is, the large language model is not limited to belong to a specific large language model, and the large language model can be replaced according to actual requirements; the functions in the canvas operation interface are not limited, and can be expanded according to actual requirements; the position layout of the intelligent dialogue interface and the canvas operation interface in the diagram design interface is not limited, and the intelligent dialogue interface and the canvas operation interface are in the same diagram design interface.

In some embodiments, describing a design information diagram as an example, a system for implementing interaction between an intelligent dialogue interface and a canvas operation interface by taking a large language model as a center may be referred to as a GRAPHIMIND system, where the information diagram design interface by taking the large language model as a center may be as shown in fig. 4, fig. 4 is a diagram design interface corresponding to the GRAPHIMIND system, the upper left corner of the diagram design interface in fig. 4 may be the information diagram currently being manufactured, and if a new information diagram is to be manufactured, a new information diagram may be added by clicking "New Canva"; the left dialogue interface is the intelligent dialogue interface; the right part is the canvas operation interface; It should be noted that, the system application for implementing the interaction between the intelligent dialogue interface and the canvas operation interface by using the large language model as the center of the application is not limited in dialogue language type, the Chinese dialogue in fig. 4 is only one specific embodiment, and the English or other languages can be adopted for dialogue. The process of designing the information graph can be described with reference to fig. 4, drawing resource inquiry information can be input in the intelligent dialogue interface shown in fig. 4, and the terminal can respond to the drawing resource inquiry information input by the intelligent dialogue interface and send the drawing resource inquiry information to the server; the server can identify the intention of drawing resource inquiry information based on a large language model, determine drawing resource types and resource demand parameters based on the identified intention, call a target resource generation application with the best matching drawing resource types, generate drawing resources which are matched with the resource demand parameters semantically based on the resource demand parameters through the target resource generation application, and return the drawing resources to the terminal; The terminal displays the drawing resources into an intelligent dialogue interface, responds to the selection operation of a drawing design interface, adjusts the drawing resources corresponding to the selection operation into a canvas operation interface, responds to the adjustment operation in the canvas operation interface, and adjusts the drawing resources corresponding to the adjustment operation to obtain a design drawing corresponding to the current adjustment; Specifically, as shown in fig. 4, when a dialog content of 'recommending some information about polar bear' is input in the intelligent dialog interface in fig. 4, the server may recognize the intention of the dialog content and return the generated information about polar bear to the terminal, so that the terminal may display in the intelligent dialog interface, where the displayed content includes a main title text, a paragraph content text, an icon suitable for polar bear theme, and the like suitable for polar bear information, and the terminal may drag and drop a design element (i.e. drawing resource) corresponding to the clicking operation into the canvas operation interface in response to the clicking operation on the intelligent dialog interface and the drag and drop operation on the drawing design interface, And respond to the adjustment operation in the canvas operation interface, get the correspondent design diagram, in addition, if want other drawing resources, can also continue to input the dialogue content of the relative intention in the intelligent dialogue interface, for example continue to input in the intelligent dialogue interface in figure 4' too good-! Creating dialogue contents of some theme pictures', the server can identify dialogue contents sent by the terminal based on a large language model, identify the intention of acquiring the theme pictures conforming to the polar bear theme, call the application related to the theme pictures to generate corresponding theme pictures, and return the theme pictures to the terminal so that the terminal can be displayed in an intelligent dialogue interface, specifically, the dialogue reply contents in fig. 4 can be recovered, likewise, the theme pictures in the intelligent dialogue interface can be selected through mouse clicking, and the selected theme pictures are dragged into a canvas operation interface for graphic design, as shown in fig. 4, the lower part of the canvas operation interface is provided with tools of a traditional canvas operation interface, the corresponding tool can be selected, the current design drawing in the canvas operation interface can be adjusted, and so on until a satisfactory information drawing is obtained.

In more detail, the process of implementing interaction between the intelligent dialog interface and the canvas operation interface with the large language model as the center can be described with reference to fig. 6, dialog content can be input in the intelligent dialog interface, the terminal can respond to the dialog content input by the intelligent dialog interface, the dialog content can include drawing resource query information, the terminal can send the dialog content including the drawing resource query information to the server, the server identifies the intention of the dialog content through the large language model, the drawing resource type (corresponding to the design task of fig. 6) and the resource requirement parameter are determined based on the identified intention, wherein the drawing resource type includes a plurality of types, which can be but not limited to information collection (can correspond to text collection) in fig. 6, visual element design (can correspond to icon retrieval), background graph design (can correspond to background graph generation), topic graph design (can correspond to topic graph generation), and layout design (can correspond to layout style generation), and different drawing resource types can be configured with a plurality of application tools capable of generating the type of resources, the same drawing resource type can be configured with the application tools capable of generating the type of resources, and the application tool types also include but not limited to GPT-table-search, visual search, and visual search, background map, and theme map display, and graphic display are shown in fig. 6, and MidJourney; the server can call a target resource generation application which is most matched with the type of the drawing resource determined under the recognition intention, and generates drawing resources (corresponding to the design resources in fig. 6) based on the resource demand parameters through the target resource generation application, wherein some types of the drawing resources can be shown as fig. 5, and when the drawing resources are acquired through the intelligent dialogue interface, the drawing resources of the types such as texts, icons, pictures (including theme pictures and background pictures), layout patterns and the like returned in a dialogue form are shown in fig. 5; further, the server may send the drawing resource to the terminal, the terminal may display the drawing resource in the intelligent dialogue interface, and in response to a selection operation of the drawing design interface, set the drawing resource corresponding to the selection operation to the canvas operation interface, and in response to an adjustment operation of the drawing resource by the canvas operation interface, obtain a corresponding design drawing (corresponding to the information diagram in fig. 6). The intelligent dialogue interface provides a rapid window for drawing resource acquisition by taking the large language model as a center, the interaction between the large language model and various application tools provides kernel support for drawing resource acquisition, rapid and accurate acquisition of drawing resources required by users is facilitated, the diversity of drawing resource patterns is facilitated, creativity of non-artistic profession and artistic profession users is facilitated to be stimulated, and therefore efficiency of making design drawings satisfactory to users is facilitated to be improved.

In some embodiments, to assess the effectiveness of the present invention, user studies may also be performed, specifically as follows:

The experiment includes two parts, in the first part, the participant is required to create two information graphs using the invention (the system which can realize the interaction of the intelligent dialogue interface and the canvas operation interface by taking the large language model as the center is called GRAPHIMIND) or PowerPoint; in the second part, the participants are required to be free to use the invention, after which we have investigated the participants.

1. Study design. Each participant uses only one system, either the system of the present invention or the baseline system. To be able to directly compare the differences in the exact same design task (corresponding drawing resource types) using the present invention and using the baseline system, an inter-group design can be used. In view of the wide popularity of PowerPoint in information and graphical editing and intuitive and easy-to-use interfaces, powerPoint may be selected as the baseline system. It is contemplated that professional graphic tools such as Adobe Illustrator may not be selected because these products have a high learning curve and may be overwhelming to participants of the novice design. To align with the design functions outlined in the present invention, the baseline system group is allowed to use Internet search tools, including Google pictures and any other online design resources, and the "PowerPoint+Internet" baseline system may be hereafter referred to simply as PowerPoint.

2. A participant. 16 participants can be recruited by online advertising published at local universities, particularly looking for individuals without professional design expertise. Of these 16 participants, 11 were considered men, 5 were women, 6 were grasses, and 10 were study groups. 12. The age of one participant (75%) was between 18 and 24 years and the age of 4 people was between 25 and 35 years. All participants were informed of little to no experience in graphic design (6 persons were titled none, 10 persons had little experience). In familiarity with PowerPoint, three participants were told to be low in proficiency, five medium, and eight relatively high.

3. Setup and procedure. 16. The participants were randomly assigned to two groups, each group comprising eight participants. Based on previous study designs, one group used only the system of the present invention, while the other group used only the PowerPoint baseline system. Experiments were performed on an individual basis, one participant at a time. Each participant was assigned a desktop, mouse, and keyboard equipped with 27 inches of available computer equipment, including a display screen. In the first part of the experiment, the experiment coordinator (one of the authors of the present invention) first introduced the information graph, including showing several examples and explaining the experimental system, one of which is the present invention, and the other of which is PowerPoint. All functions related to the information graph design task are interpreted and demonstrated (i.e., the functions of the resource generation application corresponding to each drawing resource type), with the order being random to avoid implications for any design workflow. Next, the participants are required to create two information graphs. The first topic is "you prefer roles" and the second topic is "ancient civilization". These broad topics are chosen to allow participants flexibility in their creativity without being limited by highly specific or technical topics. For each of these two topics, the experiment coordinator first introduced the scope of the topic and then instructed the participants to design their favorite information graph based on the topic at hand. The participants are explicitly notified that their screen is to be video recorded and licensed by the participants.

In the second research section, the invention was presented to participants in the PowerPoint group and the use of the system of the invention was explained in detail. All participants then conducted a freeform investigation of the invention in that they were asked to free the experimental system, try the function of the system, and create the graphical design (for the PowerPoint group) they chose. This portion is limited to a maximum of 30 minutes. At the end of this section, the participant is asked to fill out a questionnaire asking the participant for the overall impression and creative process of the invention. The entire session lasts approximately one and a half hours.

4. And (5) measuring. For the questionnaire in the second section, the inventive support index (CSI, customer Satisfaction Index) may be employed to measure the inventive support of the present invention. The original CSI has six creative support dimensions including exploration, expressive, immersive, enjoyment, and resulting worth of effort and collaboration. Since the present invention is not a cooperative software, four dimensions of CSI can be used: enjoyment, exploration, expressive and results are worth of effort. Furthermore, we add one question about ease of use and six more questions, investigating the degree of creativity of specific function support, including layout (corresponding to layout style), resources (corresponding to drawing resources of pictures or icons) and information (corresponding to drawing resources of texts), each having two questions. Finally, we add three open questions aimed at introducing comprehensive user feedback regarding the advantages and disadvantages of the present invention, focusing on the user's opinion of canvas and dialog interactions. In general, the questionnaire includes 11 Likett scale questions and 3 open questions, as shown in FIG. 8.

5. And (5) analyzing. The inventors of the present invention analyzed and encoded the video recorded in the first study, categorizing each interaction related to creating an information graph into six predefined design tasks (i.e., six predefined cartographic resource types). Interactions that do not fall within these six categories are categorized as "other" categories. During encoding, the two authors sit together and divide the video into paragraphs. Each paragraph refers to a series of interactions associated with the same category. It may be noted that in the PowerPoint group, there are several design categories, such as "Critical persona design" and "visual element design", which may cover a range of interactions from online search resources, uploading to PowerPoint and design. For each paragraph, its start time, end time and category are identified and saved. A total of 32 interaction logs were analyzed (2 systems x 8 participants x2 information graphs). A visual analysis method based on a timeline can then be employed to examine the time cost and time pattern in the interaction logs of the present group and PowerPoint group.

6. As a result. Firstly introducing interaction analysis, and comparing the modes of using the method and the PowerPoint group design information diagram by participants; the questionnaire results and analysis of open questions are then presented to provide more details about the participants' intent of use.

1. And (5) interaction analysis. (1) interaction time. Figure 9 shows the average time taken for the different operations of the two systems. Overall, the time required for participants to complete a satisfactory graphic design using the present invention (i.e., GRAPHIMIND in fig. 9) is shorter (m=18.26 minutes, sd=8.86) (m=33.40 minutes, sd=12.24) than PowerPoint. This difference is significant, t (14) =3.28, p < 0.01. Among the six interaction categories, the invention saves the most time in information collection. On average, the time taken for participants to collect information related to a topic using the present invention is 2.03 minutes, while the PowerPoint group is 10.76 minutes (sd=9.47), which is also the most time-consuming task of the six tasks of the PowerPoint group. This difference is significant, t (14) =3.37, p < 0.01. A reduction in time costs is also observed over the design tasks of background design, layout customization, and visual element design, although these differences are not significant. (2) design a mode of workflow. Fig. 10 visualizes a sequence of 32 interaction logs based on the use of two systems, PXX referring to participants in fig. 10. For simplicity, the interaction categories of "key character design", "background design" and "visual element design" are combined into one meta-category "resource design". Upon examining the interaction sequence in FIG. 10, several insights can be seen: first, it is apparent that the overall time significantly shortened using the present invention provides a shorter design process for the participants, in addition to participant P6; secondly, compared to the PowerPoint group, the workflow of the group of the present invention appears to be smoother and more sequential, i.e. starting from the resource or layout design, followed by information collection and finally local adjustment, in contrast to the PowerPoint group which shows a more fragmented process interspersed with frequent local adjustment and information collection; finally, it can be seen that there is a long local adjustment use at the end of the session for almost all participants in the group of the present invention, whereas in the PowerPoint group the local adjustment is distributed throughout the process. This can be explained by the need to make adjustments after design resources are created or imported in PowerPoint, rather in the present invention, these resources are seamlessly integrated and adapted to the selected layout on the canvas, potentially reducing time and effort.

2. Questionnaires. As shown in fig. 11, analysis of the 11 rickettsial problem showed an overall high score for most of the items, indicating overall positive acceptance. Notably, the "information collection efficiency" achieved the highest score (m=4.75±0.58), consistent with the first part of research findings, where the present invention demonstrated significant advantages in information collection. In contrast, the scores related to the layout, i.e. "layout adjustment" (m=3.38±0.89) and "layout adaptability" (m=3.94±1.06), are relatively low. This is also consistent with the first part of research observations, where the inventive user spends more time on local tuning than the PowerPoint group.

3. Open-ended problems. In general, participants consider the present invention to simplify and aid the information graph design process. The following is from the open problem: "what functions in the system do you think help create an information graph? "the main topic extracted in the above. The focus here is to identify and analyze the specific functions that the user finds most beneficial and why they are considered helpful in the information graph creation process.

(1) Smooth integration between text and canvas. The participant's responses emphasize that the present invention significantly simplifies and simplifies the information diagram creation process in a manner that integrates a text conversation interface (corresponding to a smart conversation interface) and a graphical operations interface (corresponding to a canvas operations interface). P2 emphasizes natural interactions, indicating that "the interaction mode is very convenient and the switching between left and right views is perceived as natural. "P6 favors the convenience of the layout design and drag-and-drop functionality, emphasizes that" direct drag-and-drop functionality and overall adaptation of left-to-right views make it possible to quickly create an information map. "P11 discusses the user friendly operation of the element, say" I can easily drag and drop text and images from the chat frame, which is quite cool-! "demonstrates the ability of software to facilitate a smooth and intuitive design workflow. Wherein P2, P6 and P11 are all participants.

(2) The design process is easily started. The invention provides a clear and easy-to-start for the information graph design process. The user can easily experiment or obtain new inspiration by simply inputting a text request in the present invention. P13 expresses the convenience of using the present invention to launch design ideas, "i need only enter text and view the returned content, by which decisions become easier to make. "P15 from the PowerPoint group, which is also the longest of all participants, discusses the difficulties faced by the traditional approach, and states that" it (i.e., the present invention) greatly improves the efficiency … of searching for information, and it is difficult to determine the next specific idea when I browse the Internet. Wherein P13 and P15 are both participants.

(3) And (5) generating effective design resources. The present invention is excellent in accurately generating key visual design resources such as images, icons, and layouts. For example, P5 emphasizes the ability of the system to automatically generate a layout that effectively meets design requirements, "automatically generating a layout and corresponding information is very helpful. "P16 emphasizes the speed with which the present invention generates visual elements," it (i.e., the present invention) rapidly generates images, icons, and provides relevant visual resources. "furthermore, participants appreciate the accuracy of the present invention in designing resource suggestions. This is particularly evident in the context of topic image generation, as mentioned in P14, "the system may provide accurate advice of topic images that are very consistent with i's expectations. Wherein P5, P16 and P14 are all participants.

(4) And (5) easy text information collection. Participants reported that the present invention significantly simplified the process of collecting text information for an information graph. The high evaluation of participants is achieved by the fact that the ChatGPT-driven information collection function of the present invention can collect topic-related information from all sources at once, as opposed to conventional information searching through Google or other websites. For example, P3 emphasizes the efficiency of the content generation tool, "batch content generation tools help I get reasonable content quickly. In addition, P8 endorses the effectiveness of the tool in summarizing text information, "the integrity of the text information abstract is high, the content is relevant, and the time of a large amount of research materials is saved. Wherein P3 and P8 are both participants.

(5) Natural dialogue interactions. It stands out as an intuitive and natural communication medium, and uses language, which is essentially the main way for human interaction. This attribute significantly enhances its user friendliness, especially for beginners. Users can express their design requirements in a simple dialogue language, which is consistent with the natural tendency of humans to communicate verbally, thereby reducing the learning curve and cognitive burden compared to conventional graphical design interfaces. Participant feedback emphasizes its advantages. P1 notices its ease of use for beginners: "it (i.e., dialog interactions) is friendly to beginners, does not need to know specific tools, is easier to generate, and" highlights its accessibility ". P16 finds it beneficial in creative excitation: "conversational interactions provide more inspiration".

4. And (5) quality evaluation. To evaluate the quality of the design and aesthetics of the results of the present invention, 20 interviewees can be assigned to score the results obtained in the first part of the study. Each interviewee independently evaluates three claims of each design outcome: (1) well designed; (2) aesthetic; (3) graphics and text content are subject matter compatible. These statements are used to evaluate three aspects of the quality of the information graph: design quality, aesthetics, and theme compatibility.

A 7 point lick scale may be used, where 1 indicates strong disagreement and 7 indicates strong agreement. A total of 32 design achievements were evaluated, half of which were from the present invention and half from PowerPoint. Subsequently, an average score for each design effort in terms of design quality, aesthetics, and subject compatibility was calculated.

The scoring results can be shown in fig. 12, with the information graph generated using the present invention obtaining slightly higher average scores than the information graph generated using PowerPoint, in terms of three evaluations of design quality, aesthetics, and subject compatibility. The two systems were found by inspection analysis to have no significant statistical difference in evaluation (design quality: p=0.737; aesthetics: p=0.285; subject compatibility: p=0.492), indicating that the performance of the invention is comparable to PowerPoint. In fig. 12, t-value represents the difference between the scores of the design effects of the two systems, and the larger the v-value is, the more obvious the difference of the design effects of the two systems is; the higher the p-value, which indicates the degree of confidence in the difference in the graph design effect between the two systems characterized by the t-value, the more evidence that there is no statistically significant difference in the scores of the graph design effect for the two systems, and the value of the t-value may be considered to be authentic when p-value < a preset value (the preset value may be 0.05), that is, the difference between the graph design effects for the two systems is considered to be authentic, and otherwise, not authentic.

The research shows that the application obviously shortens the design process of the information graph, simplifies the design flow from content collection to fine granularity manipulation of the object, simultaneously maintains the quality of the design result, ensures the design quality and simultaneously better improves the efficiency of the information graph design.

It should be understood that, although the steps in the flowcharts related to the embodiments described above are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.

Based on the same inventive concept, the embodiment of the application also provides a large language model-based resource processing design device for realizing the large language model-based resource processing design method. The implementation of the solution provided by the apparatus is similar to the implementation described in the above method, so the specific limitation in the embodiments of the resource processing design apparatus based on a large language model provided below may refer to the limitation of the resource processing design method based on a large language model hereinabove, and will not be repeated herein.

In an exemplary embodiment, as shown in fig. 14, there is provided a resource processing design apparatus based on a large language model, applied to a server, including: a receiving module 1400, a calling module 1402, and a designing module 1404, wherein:

the receiving module 1400 is configured to receive drawing resource query information sent by a terminal, where the drawing resource query information is input based on an intelligent dialogue interface in a drawing design interface, and the drawing design interface includes the intelligent dialogue interface and a canvas operation interface.

And a calling module 1402, configured to determine a drawing resource type and a resource requirement parameter corresponding to the drawing resource query information based on the large language model, call a target resource generating application matched with the drawing resource type, generate a drawing resource based on the resource requirement parameter through the target resource generating application, and send the drawing resource to the terminal.

And the design module 1404 is used for enabling the terminal to display drawing resources on the intelligent dialogue interface, wherein the drawing resources are used for being configured to target positions corresponding to the canvas operation interface according to the selection operation, and a corresponding design diagram is obtained.

In some embodiments, in terms of generating drawing resources based on the resource demand parameters by the target resource generating application, the invoking module 1402 is further configured to, if the target resource generating application is a picture generating application, convert the resource demand parameters to resource demand commands in a matching format by the picture generating application, and generate pictures based on the resource demand commands; if the target resource generating application is a text generating application, converting the resource demand parameters into resource demand commands in a matching format through the text generating application, and generating texts based on the resource demand commands, wherein the texts comprise icon keywords; acquiring an icon matched with the icon keyword from a database, or calling a picture generation application, and generating the icon matched with the icon keyword based on the icon keyword through the picture generation application; if the target resource generating application is a layout style generating application, converting the resource demand parameters into resource demand commands in a matched format through the layout style generating application, and generating the layout style based on the resource demand commands.

In some embodiments, the large language model-based resource processing design device is applied to a server and further comprises a modification module 1406, configured to receive a resource modification request sent by a terminal, where the resource modification request is input through an intelligent dialogue interface; determining modification resource identification information and modification standard parameters corresponding to the resource modification request based on the large language model, calling a resource modification application, generating update resources based on the modification standard parameters through the resource modification application, and sending the modification resource identification information and the update resources to the terminal; so that the terminal replaces the resource corresponding to the modified resource identification information with the updated resource.

In an exemplary embodiment, as shown in fig. 15, there is provided a resource processing design apparatus based on a large language model, applied to a terminal, including: an input module 1500 and a setting module 1502, wherein:

the input module 1500 is configured to respond to drawing resource query information input by the intelligent dialogue interface in the drawing design interface, and send the drawing resource query information to a server corresponding to the drawing design interface; the server determines drawing resource types and resource demand parameters corresponding to the drawing resource inquiry information based on the large language model, invokes a target resource generation application matched with the drawing resource types, and generates drawing resources based on the resource demand parameters through the target resource generation application; the diagram design interface includes an intelligent dialog interface and a canvas operation interface.

And the setting module 1502 is configured to display, on the intelligent dialogue interface, drawing resources returned by the server, and respond to a selection operation in the drawing design interface, set the drawing resources corresponding to the selection operation to a target position corresponding to the canvas operation interface, so as to obtain a corresponding design drawing.

In some embodiments, the selection operation includes a click operation and a drag-and-drop operation; in response to a selection operation in the diagram design interface, the drawing resource corresponding to the selection operation is set to a target position corresponding to the canvas operation interface, so as to obtain a corresponding design diagram, and the setting module 1502 is further configured to determine a target selection resource from the drawing resource in response to a click operation in the intelligent dialogue interface, drag and drop the target selection resource to the corresponding target position in the canvas operation interface in response to a drag and drop operation on the target selection resource in the diagram design interface, so as to obtain a corresponding design diagram.

In some embodiments, the setting module 1502 is further configured to send the resource selection information to the server in response to the resource selection information input at the intelligent dialogue interface, so that the server identifies the target selection resource identification information corresponding to the resource selection information based on the large language model; and receiving target selection resource identification information sent by the server, and configuring target selection resources corresponding to the target selection resource identification information at preset positions in the canvas operation interface based on the target selection resource identification information to obtain a corresponding design drawing.

In some embodiments, the setting module 1502 is further configured to send the layout resource query information to the server if the layout resource query information is entered in the intelligent dialog interface when the drawing resource already exists in the canvas operation interface; the server determines layout resource demand parameters corresponding to the layout resource inquiry information based on the large language model, invokes the layout pattern generation application, and generates the layout pattern based on the layout resource demand parameters through the layout pattern generation application; and displaying the layout style returned by the server on the intelligent dialogue interface, responding to the selection operation in the graphic design interface, and carrying out self-adaptive layout setting on the drawing resources in the canvas operation interface based on the layout style corresponding to the selection operation to obtain the design drawing corresponding to the layout style.

The above-described various modules in the large language model-based resource processing design apparatus may be implemented in whole or in part by software, hardware, and combinations thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.

In one exemplary embodiment, a computer device is provided, which may be a server, and the internal structure thereof may be as shown in fig. 16. The computer device includes a processor, a memory, an Input/Output interface (I/O) and a communication interface. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used to store data related to the resource processing design based on the large language model. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for communicating with an external terminal through a network connection. The computer program, when executed by a processor, implements a resource processing design method based on a large language model.

In an exemplary embodiment, a computer device, which may be a terminal, is provided, and an internal structure thereof may be as shown in fig. 17. The computer device includes a processor, a memory, an input/output interface, a communication interface, a display unit, and an input means. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input/output interface of the computer device is used to exchange information between the processor and the external device. The Communication interface of the computer device is used for conducting wired or wireless Communication with an external terminal, and the wireless Communication can be realized through WIFI, a mobile cellular network, near field Communication (NEAR FIELD Communication) or other technologies. The computer program, when executed by a processor, implements a resource processing design method based on a large language model. The display unit of the computer device is used for forming a visual picture, and can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, can also be a key, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.

It will be appreciated by those skilled in the art that the structures shown in fig. 16 and 17 are merely block diagrams of portions of structures associated with the present inventive arrangements and are not limiting of the computer device to which the present inventive arrangements may be implemented, and that a particular computer device may include more or fewer components than shown, or may be combined with certain components, or may have a different arrangement of components.

In an embodiment, there is also provided a computer device comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing the steps of the method embodiments described above when the computer program is executed.

In one embodiment, a computer-readable storage medium is provided, storing a computer program which, when executed by a processor, implements the steps of the method embodiments described above.

In an embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the steps of the method embodiments described above.

It should be noted that, the user information (including but not limited to user equipment information, user personal information, etc.) and the data (including but not limited to data for analysis, stored data, presented data, etc.) related to the present application are both information and data authorized by the user or sufficiently authorized by each party, and the collection, use and processing of the related data are required to meet the related regulations.

Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in embodiments provided herein may include at least one of non-volatile memory and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (RESISTIVE RANDOM ACCESS MEMORY, reRAM), magneto-resistive Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (PHASE CHANGE Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in various forms such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The databases referred to in the embodiments provided herein may include at least one of a relational database and a non-relational database. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processor referred to in the embodiments provided in the present application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computation, an artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) processor, or the like, but is not limited thereto.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the present application.

The foregoing examples illustrate only a few embodiments of the application and are described in detail herein without thereby limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of the application should be assessed as that of the appended claims.

Claims

1. A resource processing design method based on a large language model, characterized in that it is applied to a server and comprises:

Receiving mapping resource inquiry information sent by a terminal, wherein the mapping resource inquiry information is based on the conversation content inputted through the intelligent dialogue interface in the map design interface, wherein the map design interface includes the intelligent dialogue interface and the canvas operation interface;

Based on the large language model, determine the cartographic resource type and resource requirement parameters corresponding to the cartographic resource query information, call the target resource generation application that matches the cartographic resource type, generate the cartographic resource based on the resource requirement parameters through the target resource generation application, and send the cartographic resource to the terminal; the cartographic resource includes pictures, texts and layout styles; the layout style is a domain-specific language script, which is used to perform adaptive layout adjustment on the cartographic resources in the canvas operation interface;

So that the terminal displays the drawing resources on the intelligent dialogue interface, and the drawing resources are used to be configured to the target position corresponding to the canvas operation interface according to the selection operation to obtain the corresponding design drawing.

2. The method according to claim 1, characterized in that the step of generating a mapping resource based on the resource requirement parameter by the target resource generation application comprises:

If the target resource generation application is a picture generation application, converting the resource requirement parameter into a resource requirement command of a matching format through the picture generation application, and generating a picture based on the resource requirement command;

If the target resource generation application is a text generation application, the resource requirement parameter is converted into a resource requirement command in a matching format by the text generation application, and a text is generated based on the resource requirement command, wherein the text includes an icon keyword; an icon matching the icon keyword is obtained from a database, or an image generation application is called, and an icon matching the icon keyword is generated by the image generation application based on the icon keyword;

If the target resource generation application is a layout style generation application, the resource requirement parameters are converted into a resource requirement command of a matching format by the layout style generation application, and a layout style is generated based on the resource requirement command.

3. The method according to claim 1, characterized in that the method further comprises:

receiving a resource modification request sent by the terminal, wherein the resource modification request is input through the intelligent dialogue interface;

Determine, based on the large language model, modified resource identification information and modification standard parameters corresponding to the resource modification request, call a resource modification application, generate an updated resource based on the modification standard parameters by the resource modification application, and send the modified resource identification information and the updated resource to the terminal;

So that the terminal replaces the resource corresponding to the modified resource identification information with the updated resource.

4. A resource processing design method based on a large language model, characterized in that it is applied to a terminal, and the method comprises:

In response to the mapping resource query information input into the intelligent dialogue interface in the map design interface, the mapping resource query information is sent to the server corresponding to the map design interface, and the mapping resource query information is the conversation content; so that the server determines the mapping resource type and resource demand parameters corresponding to the mapping resource query information based on the large language model, calls the target resource generation application matching the mapping resource type, and generates mapping resources based on the resource demand parameters through the target resource generation application; the map design interface includes the intelligent dialogue interface and the canvas operation interface; the mapping resources include pictures, texts and layout styles; the layout style is a domain-specific language script, which is used to perform adaptive layout adjustment on the mapping resources in the canvas operation interface;

The drawing resources returned by the server are displayed on the intelligent dialogue interface, and in response to a selection operation in the drawing design interface, the drawing resources corresponding to the selection operation are adjusted to a target position corresponding to the canvas operation interface to obtain a corresponding design drawing.

5. The method according to claim 4, characterized in that the selection operation includes a click operation and a drag-and-drop operation, and the response to the selection operation in the drawing design interface, adjusting the drawing resource corresponding to the selection operation to the target position corresponding to the canvas operation interface to obtain the corresponding design drawing, comprises:

In response to a click operation in the intelligent dialogue interface, a target selection resource is determined from the drawing resources, and in response to a drag-and-drop operation on the target selection resource in the drawing design interface, the target selection resource is dragged and dropped to a corresponding target position in the canvas operation interface to obtain a corresponding design drawing;

The method further comprises:

In response to resource selection information inputted in the intelligent dialogue interface, the resource selection information is sent to the server, so that the server identifies target selection resource identification information corresponding to the resource selection information based on the large language model;

The target selection resource identification information sent by the server is received, and based on the target selection resource identification information, the target selection resource corresponding to the target selection resource identification information is configured at a preset position in the canvas operation interface to obtain a corresponding design drawing.

6. The method according to claim 4, characterized in that the method further comprises:

If a drawing resource already exists in the canvas operation interface, layout resource query information is input in the intelligent dialogue interface, and the layout resource query information is sent to the server; so that the server determines the layout resource requirement parameters corresponding to the layout resource query information based on the large language model, calls a layout style generation application, and generates a layout style based on the layout resource requirement parameters through the layout style generation application;

The layout style returned by the server is displayed on the intelligent dialogue interface, and in response to a selection operation in the drawing design interface, the drawing resources in the canvas operation interface are adaptively arranged based on the layout style corresponding to the selection operation to obtain a design drawing corresponding to the layout style.

7. A resource processing design device based on a large language model, characterized in that it is applied to a server, and the device comprises:

A receiving module, configured to receive mapping resource inquiry information sent by a terminal, wherein the mapping resource inquiry information is a conversation content input based on an intelligent dialogue interface in a map design interface, wherein the map design interface includes the intelligent dialogue interface and a canvas operation interface;

A calling module, used to determine the cartographic resource type and resource requirement parameters corresponding to the cartographic resource query information based on a large language model, call a target resource generation application that matches the cartographic resource type, generate a cartographic resource based on the resource requirement parameters through the target resource generation application, and send the cartographic resource to the terminal; the cartographic resource includes pictures, texts and layout styles; the layout style is a domain-specific language script, which is used to perform adaptive layout adjustment on the cartographic resources in the canvas operation interface;

The design module is used to enable the terminal to display the drawing resources on the intelligent dialogue interface, and the drawing resources are used to be configured to the target position corresponding to the canvas operation interface according to the selection operation to obtain the corresponding design drawing.

8. A computer device, comprising a memory and a processor, wherein the memory stores a computer program, wherein the processor implements the steps of the method according to any one of claims 1 to 6 when executing the computer program.

9. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.