Network topology image editing method, device and equipment and readable storage medium

Abstract

The invention discloses a network topology image editing method, device and equipment and a readable storage medium, and is applied to the technical field of network topology images, and the method comprises the steps: recognizing a to-be-edited network topology image and network topology editing demand information, and obtaining a target editing intention and editing parameter information; editing to-be-edited network topology information corresponding to the to-be-edited network topology image by using an editing agent based on the target editing intention and the editing parameter information to obtain edited network topology information; and generating a target topological image based on the edited network topological information, and generating a target network topological image corresponding to the target topological image. Compared with the current mode that the network topological image can only be edited manually, the method has the advantages that the target editing intention and the editing parameter information are intelligently recognized, so that the editing agent is utilized to perform intelligent editing according to the target editing intention and the editing parameter information, and the topological image editing efficiency and accuracy are greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of network topology images, and in particular to a network topology image editing method, device, equipment and readable storage medium. Background Art

[0002] Network topology usually exists in the form of images in various technical solutions, user manuals, configuration documents and other carriers, providing users with intuitive and clear topology information. These topology images usually do not have corresponding formatted topology structure information for users to directly edit and use, and often require users to manually obtain relevant information by viewing. If the topology image needs to be edited, it is usually necessary to manually read the topology structure information, and then manually edit and redraw the image. Therefore, with the complexity of the topology structure, manual redrawing and editing of network topology images is inefficient, error-prone, and difficult to maintain consistency with the original image style.

[0003] It can be seen that how to improve the efficiency and accuracy of network topology image editing is a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0004] In view of this, an object of the present invention is to provide a network topology image editing method, device, equipment and readable storage medium, which solves the technical problem of low efficiency of network topology image editing in the prior art.

[0005] In order to solve the above technical problems, the present invention provides a network topology image editing method, comprising:

[0006] Identify the network topology image to be edited and the network topology editing requirement information, and obtain the target editing intention and editing parameter information;

[0007] Based on the target editing intention and the editing parameter information, the network topology information to be edited corresponding to the network topology image to be edited is edited using an editing agent to obtain the edited network topology information; wherein the editing agent is a logic code including an editing function;

[0008] A target topology image is generated based on the edited network topology information, and a target network topology image corresponding to the target topology image is generated.

[0009] Optionally, the identifying of the network topology image to be edited and the network topology editing requirement information to obtain the target editing intent and editing parameter information includes:

[0010] The network topology image to be edited and the network topology editing requirement information are identified using a network topology intention recognition model to obtain the target editing intention and the editing parameter information.

[0011] Optionally, when the network topology intent recognition model is a large language model, the training process of the network topology intent recognition model includes:

[0012] Acquire an intention recognition data set; wherein the intention recognition data set is a dialogue sample including an input of a network topology image and network topology editing requirement information, and an output of editing intention and editing parameter information;

[0013] The large language model is trained according to the intent recognition data set to obtain the network topology intent recognition model.

[0014] Optionally, based on the target editing intent and the editing parameter information, using an editing agent, editing the network topology information to be edited corresponding to the network topology image to be edited to obtain the edited network topology information, including:

[0015] Extracting network topology information from the network topology image to be edited using a large network topology understanding model to obtain the network topology information to be edited;

[0016] The network topology information to be edited is edited using the editing agent based on the target editing intention and the editing parameter information to obtain the edited network topology information.

[0017] Optionally, based on the target editing intent and the editing parameter information, using an editing agent, editing the network topology information to be edited corresponding to the network topology image to be edited to obtain the edited network topology information, including:

[0018] Arrange tasks based on the target editing intention and the editing parameter information to obtain multiple subtasks;

[0019] The network topology information to be edited is edited using the editing agent corresponding to each of the subtasks to obtain the edited network topology information.

[0020] Optionally, the editing agent includes at least one of an intention recognition agent, a topology information extraction agent, a topology calculation agent, a node editing agent, an edge editing agent, an information viewing agent and a dialogue agent.

[0021] Optionally, generating a target topology image based on the edited network topology information, and generating a target network topology image corresponding to the target topology image, includes:

[0022] Determine the node graph and location information corresponding to the target topology image;

[0023] Determine a node icon corresponding to each node graphic based on the node graphic and the position information;

[0024] The node graphic in the target topology image is replaced by the node icon according to the location information to obtain the target network topology image.

[0025] The embodiment of the present invention further provides a network topology image editing device, comprising:

[0026] An intention recognition module is used to recognize the network topology image to be edited and the network topology editing requirement information, and obtain the target editing intention and editing parameter information;

[0027] A first editing module, configured to edit the network topology information to be edited corresponding to the network topology image to be edited using an editing agent based on the target editing intent and the editing parameter information, to obtain the edited network topology information; wherein the editing agent is a logic code including an editing function;

[0028] The second editing module is used to generate a target topology image based on the edited network topology information, and generate a target network topology image corresponding to the target topology image.

[0029] An optional network topology image editing device includes:

[0030] Memory for storing computer programs;

[0031] A processor is used to execute the computer program to implement the steps of the above-mentioned network topology image editing method.

[0032] An embodiment of the present invention further provides a readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned network topology image editing method are implemented.

[0033] An embodiment of the present invention further provides a computer program product, including a computer program / instruction, which implements the steps of the above-mentioned network topology image editing method when executed by a processor.

[0034] It can be seen that the present invention obtains the target editing intention and editing parameter information by identifying the network topology image to be edited and the network topology editing requirement information; based on the target editing intention and the editing parameter information, the network topology information to be edited corresponding to the network topology image to be edited is edited using an editing agent to obtain the edited network topology information; wherein the editing agent is a logic code including an editing function; a target topology image is generated based on the edited network topology information, and a target network topology image corresponding to the target topology image is generated.

[0035] The beneficial effect of the present invention is that compared with the current manual editing of network topology images, the present application intelligently identifies the target editing intent and editing parameter information, and then uses the editing agent to perform intelligent editing according to the target editing intent and editing parameter information, thereby greatly improving the efficiency and accuracy of topology image editing.

[0036] In addition, the present invention also provides a network topology image editing device, equipment and readable storage medium, which also have the above-mentioned beneficial effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0038] Figure 1 A flowchart of a network topology image editing method provided by an embodiment of the present invention;

[0039] Figure 2 A flowchart of a network topology image editing method provided by an embodiment of the present invention;

[0040] Figure 3 A schematic diagram of the structural framework of a network topology image editing method provided by an embodiment of the present invention;

[0041] Figure 4 A schematic diagram of training data generation provided by an embodiment of the present invention;

[0042] Figure 5 A schematic diagram of a target intention provided by an embodiment of the present invention;

[0043] Figure 6 A schematic diagram of the structure of a network topology image editing device provided by an embodiment of the present invention;

[0044] Figure 7 A schematic diagram of the structure of a network topology image editing device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] Please refer to Figure 1 , Figure 1 A flowchart of a network topology image editing method provided by an embodiment of the present invention. The method may include:

[0047] S101, identifying a network topology image to be edited and network topology editing requirement information, and obtaining target editing intent and editing parameter information.

[0048] The execution subject of this embodiment is an electronic device. In this embodiment, the electronic device is a computer, a mobile phone, etc. In this embodiment, the network topology image to be edited is a network topology image with a simple background, and the nodes and connection relationships (edges) of the network topology, as well as the attributes of the nodes or edges that may exist, can be clearly found in the image. This embodiment does not limit the specific method of identification. For example, this embodiment can be identified based on a machine learning model; or this embodiment can also be identified based on a large model. The target editing intention in this embodiment is an expression of the user's request to edit the network topology diagram (for example: add an edge from A to B in the topology diagram). The editing parameter information in this embodiment is the relationship between nodes. This embodiment does not limit the specific target editing intention. For example, the target editing intention in this embodiment can be the intention of topology information extraction, topology information viewing, shortest path calculation, effective path query, etc.; this embodiment does not limit the specific editing parameter information. For example, the editing parameter information in this embodiment can be the starting point, end point, weight, etc. of the edge. Extracting user intent and related parameter information refers to extracting user intent through intent recognition agents (based on the network topology intent model). User intent can include 14 categories, namely: topology information extraction, topology information viewing, shortest path calculation, valid path query, connectivity status calculation, adding nodes, deleting nodes, modifying nodes, viewing nodes, adding edges, deleting edges, modifying edges, viewing edges, and other intents. Related parameter information refers to parameters associated with intent. For example, in the intent to add an edge, the related parameters include the edge's start point, end point, weight, description, direction, and other information.

[0049] It should be further explained that, based on any of the above embodiments, in order to improve the accuracy of recognition, the above-mentioned recognition of the network topology image to be edited and the network topology editing requirement information to obtain the target editing intent and editing parameter information may include: using a network topology intention recognition model to recognize the network topology image to be edited and the network topology editing requirement information to obtain the target editing intent and editing parameter information. This embodiment does not limit the specific network topology intention recognition model to include a machine learning model (for example, a neural network model, topology data analysis, etc.) and a large language model, etc.

[0050] It should be further explained that, based on any of the above embodiments, when the network topology intent recognition model is a large language model, the training process of the network topology intent recognition model may include:

[0051] S1011, obtaining an intent recognition data set; wherein the intent recognition data set is a dialogue sample including an input of a network topology image and network topology editing requirement information, and an output of editing intent and editing parameter information.

[0052] S1012, training the large language model according to the intent recognition data set to obtain a network topology intent recognition model.

[0053] In this embodiment, a network topology intention recognition dataset (intention recognition dataset) is constructed, and a large model specifically used for identifying intentions related to network topology editing is trained and optimized on the basis of an open source large language model. The instruction fine-tuning method used in model training is a general large model fine-tuning method, which can be one of the fine-tuning methods such as full parameter fine-tuning, Lora (parameter efficient fine-tuning method), P-Tuning (technology for adjusting large pre-trained language models to adapt to specific tasks), P-Tuningv2, etc. The network topology intention recognition dataset is a dataset specially designed and constructed for training a large model of network topology intention. The dataset adopts a prompt-guided method, including components such as intent classification, information extraction, and information verification. Each part uses a question-and-answer method to form instruction fine-tuning data. The construction method is to call a large language model to assist in generating a large number of samples according to manually specified samples, and to generate a certain scale of data samples by manual screening and iteration to construct a dataset. This embodiment does not limit a specific large language model. For example, the large language model in this embodiment can be ChatGPT; or the large language model in this embodiment can be a bidirectional encoder. This embodiment provides a specific model training method to improve the accuracy of model training.

[0054] S102, based on the target editing intent and editing parameter information, using the editing agent, edit the network topology information to be edited corresponding to the network topology image to be edited, and obtain the edited network topology information; wherein the editing agent is a logic code including the editing function.

[0055] The editing agent in this embodiment is a logic code for achieving various editing goals. In this embodiment, when editing, the network topology information to be edited corresponding to the network topology image to be edited is edited, so as to obtain the edited network topology information, and the network topology image is generated according to the edited network topology information. Before editing, this embodiment can determine whether the extraction of all network topology information has been completed. Only when all network topology information extraction has been completed can the editing agent be used for editing. This embodiment does not limit the specific editing agent. For example, the editing agent in this embodiment can be a topology information extraction agent; or the agent in this embodiment can be a topology calculation agent; or the agent in this embodiment can be a node editing agent.

[0056] It should be further explained that, based on any of the above embodiments, the above-mentioned using an editing agent based on the target editing intent and the editing parameter information to edit the network topology information to be edited corresponding to the network topology image to be edited to obtain the edited network topology information may include:

[0057] S1021, extracting network topology information from the network topology image to be edited using the network topology understanding large model to obtain the network topology information to be edited;

[0058] S1022, using an editing agent based on the target editing intent and the editing parameter information, edits the network topology information to be edited to obtain the edited network topology information.

[0059] This embodiment needs to extract the network topology information to be edited before editing it. This embodiment can determine whether the extraction of the network topology information to be edited has been completed. If it has not been completed (that is, the user has the first editing requirement for the currently input network topology image), it is necessary to first call the multimodal network topology understanding model to extract the network topology information in the image; if the topology information extraction has been completed, the corresponding intelligent agent is entered in sequence for calculation and processing. The above-mentioned multimodal network topology understanding model (corresponding to the network topology understanding model) is a large model for extracting network topology information from an image formed by training, which is obtained by constructing a multimodal network topology information extraction data set and training based on an open source multimodal large model. The model is used as an intelligent agent specifically for extracting network topology information. The instruction fine-tuning method used in model training is a general large model fine-tuning method, which can be one of the fine-tuning methods such as full parameter fine-tuning, Lora, P-Tuning, P-Tuningv2, etc. The multimodal network topology information extraction data set is a data set specially designed and constructed for training a modal network topology understanding model. The construction method is as follows: 1) set constraints (such as the number of nodes, the number of edges, node names and description rules, etc.) and randomly generate a network topology structure. The main data of the structure includes node information (name, attributes) and edge information (starting point, focus, weight, direction, attributes, etc.). The network topology structure data is structured data; 2) construct topological question and answer data based on the structured topological data, that is, generate questions and answers from the bottom up (that is, starting from the three most basic dimensions of nodes, edges, and attributes, and going up to part and all of the topology graph) for the topological structure; 3) use the topological visualization method to visualize the generated structured data, generate the original topological image, and combine it with the topological question and answer data generated in the second step to form a multimodal network topology information extraction data; 4) in order to strengthen the model's pattern learning of the real network topology graph, add image data based on the real network topology node icon, including images of various types of network node devices, and generate corresponding question and answer data to form a training data based on the real network node; 5) iteratively use the above method to construct a certain scale of data sets to form a multimodal network topology data set. During training, since each training data may contain multiple question-answer combinations, a certain number of question-answer combinations are randomly selected as text information for training, and a large network topology model is formed through appropriate training fine-tuning parameter training. This embodiment provides a specific method for extracting network topology information, which improves the accuracy of extracting network topology information.

[0060] It should be further explained that the above-mentioned use of an editing agent based on the target editing intent and editing parameter information to edit the network topology information to be edited corresponding to the network topology image to be edited to obtain the edited network topology information may include: performing task scheduling based on the target editing intent and editing parameter information to obtain multiple subtasks; using the editing agent corresponding to each subtask to edit the network topology information to be edited to obtain the edited network topology information. The task scheduling of this embodiment is to decompose and schedule tasks according to the recognized intent, and each scheduled subtask corresponds to a dedicated agent to perform the next step of processing. It can be understood that the user's editing requirement is a combined requirement, which can be decomposed into multiple subtasks.

[0061] It should be further explained that, based on any of the above embodiments, the above-mentioned editing agent includes at least one of an intent recognition agent, a topology information extraction agent, a topology calculation agent, a node editing agent, an edge editing agent, an information viewing agent and a dialogue agent. The agent in this embodiment refers to a set of multiple agents constructed to complete the network topology image editing task. It includes seven agents: intent recognition, topology information extraction, topology calculation, node editing, edge editing, information viewing, and general dialogue. Among them, the intent recognition agent based on the network topology intent large model is the core of the multi-agent, which is used to undertake related tasks such as user interaction, intent recognition, task decomposition, etc., call the corresponding agent to complete the task execution and feedback the execution results to the user.

[0062] S103: Generate a target topology image based on the edited network topology information, and generate a target network topology image corresponding to the target topology image.

[0063] The difference between the target topology image and the target network topology image in this embodiment is that the nodes in the target topology image are node graphics, and the nodes in the target network topology image are node icons. In this embodiment, when the intelligent agent completes the task of editing the network topology image, the original topology image is generated according to the edited network topology structure, and the original information in the original network topology image is extracted as visual guidance information, and the target network topology image after editing is generated by fusion. It can be understood that the node graphic is a graphic element used to represent a specific entity, object or concept in a graphical representation, and it is the basic unit that constitutes the entire graphic structure. The node graphic can be a simple geometric shape (such as a circle, a square, etc.). A node icon refers to a special icon, and the type, name and other information of the node can be directly determined based on the icon.

[0064] It should be further explained that, based on any of the above embodiments, the above generation of a target topology image based on the edited network topology information, and the generation of a target network topology image corresponding to the target topology image, may include: S1031, determining the node graph and position information corresponding to the target topology image; S1032, determining the node icon corresponding to each node graph based on the node graph and the position information; S1033, replacing the node graph in the target topology image with the node icon according to the position information, to obtain the target network topology image. The original topology image (target topology image) generated in this embodiment is a topology image generated by a graph data visualization tool (e.g., networkx) based on the network topology structure information, and the image represents the network nodes with abstract nodes (e.g., circles or rectangles); in this embodiment, the node icon corresponding to each node graph is determined based on the node graph and the position information; the node graph in the target topology image is replaced with the node icon according to the position information, and the process of obtaining the target network topology image is to extract the original information in the original network topology image as visual guidance information, and fuse and generate the edited network topology image. Visual guidance information refers to the use of image recognition methods to extract node target information from the original network topology image, including node icons and their location information; the above fusion generates the edited network topology image, and the fusion method is to replace the abstract node graphics in the original topology image with the node icons in the original network topology image, and complete the image fusion and generation based on the visual guidance information such as the size and position of the node, as well as the network topology structure information, to obtain the edited network topology image. This embodiment provides a specific process for determining the target network topology image, which improves the accuracy of determining the target network topology image.

[0065] A network topology image editing method provided by an embodiment of the present invention may include: S101, identifying a network topology image to be edited and network topology editing requirement information to obtain target editing intent and editing parameter information; S102, using an editing agent based on the target editing intent and editing parameter information to edit the network topology information to be edited corresponding to the network topology image to be edited, to obtain the edited network topology information; wherein the editing agent is a logic code including an editing function; S103, generating a target topology image based on the edited network topology information, and generating a target network topology image corresponding to the target topology image. Compared with the current manual editing of network topology images, the present application intelligently identifies the target editing intent and editing parameter information, thereby using an editing agent to perform intelligent editing according to the target editing intent and editing parameter information, thereby greatly improving the efficiency and accuracy of topology image editing.

[0066] In order to make the present invention easier to understand, please refer to Figure 2 , Figure 2A flowchart of a network topology image editing method provided by an embodiment of the present invention may specifically include:

[0067] S201: Acquire a network topology image to be edited and network topology editing requirement information.

[0068] The network topology image to be edited in this embodiment is an image including network nodes, connection relationships (edges), node attributes and edge attributes; the network topology editing requirement information in this embodiment is the editing operation that needs to be performed on the network topology image to be edited. For example, the network topology editing requirement information may be to add an edge; or the network topology editing requirement information in this embodiment may be to delete a node; or to add multiple nodes. For ease of understanding, please refer to Figure 3 , Figure 3 A schematic diagram of the structural framework of a network topology image editing method provided by an embodiment of the present invention, from Figure 3 It can be seen that the input of the entire network topology image editing method includes image input and text input. Task analysis is mainly for intent recognition, so as to decompose tasks based on the intent recognition results and then schedule tasks. The topology editing part mainly includes topology calculation, node editing, edge editing, information viewing and topology information extraction.

[0069] S202. Based on the network topology image to be edited and the network topology editing requirement information, a network topology intention recognition large model is used to perform user intention recognition to obtain target editing intention and editing parameter information.

[0070] The network topology intention recognition big model in this embodiment corresponds to the network topology intention recognition model mentioned above. The network topology intention recognition big model in this embodiment is a model obtained by training a large language model by constructing a network topology intention recognition data set. The target editing intent in this embodiment includes at least one of topology information extraction, topology information viewing, shortest path calculation, valid path query, connectivity status calculation, adding nodes, deleting nodes, modifying nodes, viewing nodes, adding edges, deleting edges, modifying edges, and viewing edges. The editing parameter information in this embodiment refers to parameters associated with intent. For example, in the intent to add an edge, the relevant parameters include information such as the starting point, end point, weight, description, and direction of the edge. The network topology intent recognition data set for training the large model of network topology intent recognition in this embodiment: construct prompt templates for intent classification, information extraction, information verification, user confirmation and other functions; make several dialogue samples for intent classification, information extraction, information verification, user confirmation and other functions; call the large model to assist in generating a large number of samples based on manually specified samples; manually screen the generated samples, and select typical samples to call the large model again for sample generation and expansion; repeat sample screening and sample generation and expansion based on the large model until the number meets the training requirements; finally, associate the samples with the corresponding prompt to form a network topology intent recognition data set. For ease of understanding, please refer to Figure 4 , Figure 4 A schematic diagram of training data generation provided by an embodiment of the present invention, with the left side being a schematic diagram of forming a network topology intent recognition data set. For ease of understanding, please refer to Figure 5 , Figure 5 A schematic diagram of a target editing intent provided for an embodiment of the present invention, wherein the target editing intent includes 14 intent categories, including topology information operation, topology calculation, node operation, edge operation, and other intent categories.

[0071] S203: Perform task scheduling based on the target editing intent and the editing parameter information to obtain a task scheduling result.

[0072] This embodiment can decompose and arrange tasks according to the target editing intention, so that each subtask corresponds to a dedicated agent. The task arrangement result in this embodiment includes subtasks and editing agents corresponding to the subtasks.

[0073] S204: Determine whether the network topology information extraction has been completed.

[0074] S205: If not completed, call the network topology understanding large model to extract the network topology information in the network topology image to be edited.

[0075] In this embodiment, the process of extracting a multimodal network topology information data set required for training a large model for network topology understanding may include: setting a topology structure generation strategy, including a node generation strategy, an edge generation strategy, and an attribute generation strategy, and randomly generating a network topology structure according to the set strategy conditions. The structure data includes node information (name, attributes) and edge information (starting point, end point, weight, direction, attributes, etc.), and the network topology structure data is structured data; generating topology question and answer data based on the structured topology data, that is, generating three layers of questions and answers from bottom to top for the topology structure (that is, starting from the most basic nodes, edges, and attributes, and going up to part and all of the topology graph), wherein the bottom layer includes three types of question and answer data: nodes, edges, and attributes, and the middle layer includes three dimensions: nodes, edges, and attributes. The top layer contains all the information of nodes, edges and attributes in the question and answer. The generated structured network topology data is visualized by topological visualization method (for example, networkx library based on Python), and a topological image is generated. The topological image is combined with the generated topological question and answer data to form a multimodal network topology information extraction data. In order to strengthen the model's learning of the real network topology graph, image data based on the real network topology node icon is added, including images of various types of network node devices, and corresponding question and answer data is generated to form a training data based on the real network node. The above method is used iteratively to construct a data set of a certain scale to form a multimodal network topology information extraction data set. Please refer to Figure 4 , Figure 4 The right side is a schematic diagram of training data required to form a large model for understanding network topology provided by an embodiment of the present invention.

[0076] S206: If it is completed, the network topology image to be edited is edited using the editing agent according to the task scheduling result and the network topology information to obtain the edited network topology information.

[0077] The editing agent in this embodiment can be an agent for intent recognition, topology information extraction, topology calculation, node editing, edge editing, information viewing, general dialogue, etc.

[0078] S207: Generate a target topology image according to the edited network topology information.

[0079] S208 , extracting visual guidance information of the target topology image, and generating a target network topology image corresponding to the target topology image using a fusion method based on the visual guidance information.

[0080] The visual guidance information in this embodiment refers to the node target information extracted from the target topology image using an image recognition method, including the node icon and its location information. The target network topology image in this embodiment is fused by replacing the abstract node graphics in the original topology image with the node icons in the original network topology image, and completing the image fusion and generation based on the visual guidance information such as the size and position of the node, as well as the network topology structure information, to obtain the edited target network topology image.

[0081] The technical solution of the present invention aims at the network topology image editing process, constructs a full-process method from topology information extraction, topology intention recognition, topology calculation and editing to network topology image generation, and constructs an intelligent agent for each specific function, with the user's network topology intention recognition intelligent agent as the brain, to guide the topology information extraction, topology calculation, topology editing, image generation and other different task intelligent agents to collaborate and complete the network topology editing task. In order to generate the edited network topology image, an image generation method guided by topological structure information and visual information is adopted, and the edited topological structure information is used to generate the initial topology image in a more reasonable layout. At the same time, the visual information in the original network topology image is used as a guide to fuse the initial topology image and generate the edited network topology image, thereby improving the rationality and accuracy of image generation. Moreover, in order to realize interactive network topology image editing, it is necessary to extract the network topology information in the image, adopt an open source multimodal large model to understand the network topology image, and on this basis, adopt an image and question-answer generation method based on the network topology structure to construct a multimodal network topology information extraction dataset; according to the user's topology image editing or topology calculation needs, a prompt-guided intent recognition and information extraction method is adopted to construct a network topology intent recognition dataset, forming a complete topology image editing training dataset, which supports the training of the network topology intent large model and the multimodal network topology image understanding large model.

[0082] The beneficial effects of the technical solution of the present invention may include: (1) improving the accuracy and friendliness of topological information editing in network topological images. The proposed network topological image editing method based on a large model of intelligent agents uses a trained and fine-tuned network topological intention recognition large model to extract topological information, thereby improving the accuracy of topological information capture; the interactive network topological editing method based on the network topological intention large model further improves the accuracy and friendliness. (2) Improving the efficiency of topological image editing. In the editing process from image to image, information extraction based on a large model of network topological understanding is adopted to avoid manual methods. The interactive editing based on the large model of network topological intention further optimizes the editing process, thereby greatly improving the efficiency of topological image editing. (3) Improving the flexibility of network topological image editing. The designed agent-based topological editing method can accurately perceive user intentions and call related agents to execute user intentions. The execution algorithm involved in the agent has the advantages of being pluggable and easy to update, and its flexibility is improved.

[0083] The following is an introduction to a network topology image editing device provided by an embodiment of the present invention. The network topology image editing device described below and the network topology image editing method described above can be referred to each other.

[0084] Please refer to Figure 6 , Figure 6 A schematic diagram of the structure of a network topology image editing device provided by an embodiment of the present invention may include:

[0085] The intention recognition module 100 is used to recognize the network topology image to be edited and the network topology editing requirement information, and obtain the target editing intention and editing parameter information;

[0086] The first editing module 200 is used to edit the network topology information to be edited corresponding to the network topology image to be edited using an editing agent based on the target editing intention and the editing parameter information to obtain the edited network topology information; wherein the editing agent is a logic code including an editing function;

[0087] The second editing module 300 is used to generate a target topology image based on the edited network topology information, and generate a target network topology image corresponding to the target topology image.

[0088] Further, based on the above embodiment, the intention recognition module 100 may include:

[0089] The intention recognition unit is used to use the network topology intention recognition model to recognize the network topology image to be edited and the network topology editing requirement information to obtain the target editing intention and the editing parameter information.

[0090] Further, based on the above embodiment, the above intention recognition unit may include:

[0091] An intention recognition data set acquisition subunit is used to acquire an intention recognition data set; wherein the intention recognition data set is a dialogue sample including an input of a network topology image and network topology editing requirement information, and an output of editing intention and editing parameter information;

[0092] The network topology intention recognition model training subunit is used to train the large language model according to the intention recognition data set to obtain the network topology intention recognition model.

[0093] Further, based on any of the above embodiments, the first editing module 200 may include:

[0094] A network topology information determination unit to be edited, used to extract the network topology information in the network topology image to be edited by using a network topology understanding large model, so as to obtain the network topology information to be edited;

[0095] The editing agent-based editing unit is used to edit the network topology information to be edited using the editing agent based on the target editing intention and the editing parameter information to obtain the edited network topology information.

[0096] Further, based on any of the above embodiments, the first editing module 200 may include:

[0097] A subtask determining unit, configured to perform task arrangement based on the target editing intention and the editing parameter information to obtain a plurality of subtasks;

[0098] Based on the subtask editing unit, it is used to edit the network topology information to be edited by utilizing the editing agent corresponding to each subtask to obtain the edited network topology information.

[0099] Further, based on any of the above embodiments, the above-mentioned editing agent includes at least one of an intention recognition agent, a topology information extraction agent, a topology calculation agent, a node editing agent, an edge editing agent, an information viewing agent and a dialogue agent.

[0100] Further, based on any of the above embodiments, the second editing module 300 may include:

[0101] A node information determination unit, used to determine the node graph and position information corresponding to the target topology image;

[0102] A node icon determining unit, configured to determine a node icon corresponding to each node graphic based on the node graphic and the position information;

[0103] The target network topology image generating unit is used to replace the node graph in the target topology image with the node icon according to the location information to obtain the target network topology image.

[0104] It should be noted that the order of the modules and units in the above-mentioned network topology image editing device can be changed without affecting the logic.

[0105] The network topology image editing device provided by the embodiment of the present invention may include: an intention recognition module 100, which is used to recognize the network topology image to be edited and the network topology editing requirement information to obtain the target editing intention and editing parameter information; a first editing module 200, which is used to use an editing agent based on the target editing intention and the editing parameter information to edit the network topology information to be edited corresponding to the network topology image to be edited, and obtain the edited network topology information; wherein the editing agent is a logic code including an editing function; a second editing module 300, which is used to generate a target topology image based on the edited network topology information, and generate a target network topology image corresponding to the target topology image. Compared with the current manual editing of network topology images, the present application intelligently recognizes the target editing intention and editing parameter information, and then uses the editing agent to perform intelligent editing according to the target editing intention and editing parameter information, so that the efficiency and accuracy of topology image editing are greatly improved.

[0106] A network topology image editing device provided by an embodiment of the present invention is introduced below. The network topology image editing device described below and the network topology image editing method described above can refer to each other.

[0107] Please refer to Figure 7 , Figure 7 A schematic diagram of the structure of a network topology image editing device provided by an embodiment of the present invention may include:

[0108] A memory 10, used for storing computer programs;

[0109] The processor 20 is used to execute a computer program to implement the above-mentioned network topology image editing method.

[0110] The memory 10 , the processor 20 , and the communication interface 30 all communicate with each other via a communication bus 40 .

[0111] In the embodiment of the present invention, the memory 10 is used to store one or more programs, and the program may include program code, and the program code includes computer operation instructions. In the embodiment of the present invention, the memory 10 may store programs for implementing the following functions:

[0112] Identify the network topology image to be edited and the network topology editing requirement information, and obtain the target editing intention and editing parameter information;

[0113] Based on the target editing intent and the editing parameter information, the network topology information to be edited corresponding to the network topology image to be edited is edited using an editing agent to obtain the edited network topology information; wherein the editing agent is a logic code including an editing function;

[0114] A target topology image is generated based on the edited network topology information, and a target network topology image corresponding to the target topology image is generated.

[0115] In a possible implementation, the memory 10 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function, etc.; the data storage area may store data created during use.

[0116] In addition, the memory 10 may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include an NVRAM. The memory stores an operating system and operating instructions, executable modules or data structures, or a subset thereof, or an extended set thereof, wherein the operating instructions may include various operating instructions for implementing various operations. The operating system may include various system programs for implementing various basic tasks and processing hardware-based tasks.

[0117] The processor 20 may be a central processing unit (CPU), an application specific integrated circuit, a digital signal processor, a field programmable gate array or other programmable logic device, a microprocessor or any conventional processor, etc. The processor 20 may call a program stored in the memory 10 .

[0118] The communication interface 30 may be an interface of a communication module, and is used to connect to other devices or systems.

[0119] Of course, it should be noted that Figure 7 The structure shown does not constitute a limitation on the network topology image editing device in the embodiment of the present invention. In actual applications, the network topology image editing device may include Figure 7 More or fewer components than shown, or combinations of certain components.

[0120] The following is an introduction to a readable storage medium provided in an embodiment of the present invention. The readable storage medium described below and the network topology image editing method described above can be referenced to each other.

[0121] The present invention also provides a readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned network topology image editing method are implemented.

[0122] The readable storage medium may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0123] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0124] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0125] Finally, it should be noted that, in this article, relationships such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprises" or any other variations are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0126] The above is a detailed introduction to a network topology image editing method, device, equipment and readable storage medium provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A network topology image editing method, characterized in that: include: Identify the network topology image to be edited and the network topology editing requirement information, and obtain the target editing intention and editing parameter information; Based on the target editing intention and the editing parameter information, the network topology information to be edited corresponding to the network topology image to be edited is edited using an editing agent to obtain the edited network topology information; wherein the editing agent is a logic code including an editing function; A target topology image is generated based on the edited network topology information, and a target network topology image corresponding to the target topology image is generated.

2. The network topology image editing method according to claim 1, characterized in that: The network topology image to be edited and the network topology editing requirement information are identified to obtain the target editing intention and editing parameter information, including: The network topology image to be edited and the network topology editing requirement information are identified using a network topology intention recognition model to obtain the target editing intention and the editing parameter information.

3. The network topology image editing method according to claim 2, characterized in that: When the network topology intent recognition model is a large language model, the training process of the network topology intent recognition model includes: Acquire an intention recognition data set; wherein the intention recognition data set is a dialogue sample including an input of a network topology image and network topology editing requirement information, and an output of editing intention and editing parameter information; The large language model is trained according to the intent recognition data set to obtain the network topology intent recognition model.

4. The network topology image editing method according to claim 1, characterized in that: Based on the target editing intention and the editing parameter information, the network topology information to be edited corresponding to the network topology image to be edited is edited using an editing agent to obtain the edited network topology information, including: Extracting network topology information from the network topology image to be edited using a large network topology understanding model to obtain the network topology information to be edited; The network topology information to be edited is edited using the editing agent based on the target editing intention and the editing parameter information to obtain the edited network topology information.

5. The network topology image editing method according to any one of claims 1 to 4, characterized in that: Based on the target editing intention and the editing parameter information, the network topology information to be edited corresponding to the network topology image to be edited is edited using an editing agent to obtain the edited network topology information, including: Arrange tasks based on the target editing intention and the editing parameter information to obtain multiple subtasks; The network topology information to be edited is edited using the editing agent corresponding to each of the subtasks to obtain the edited network topology information.

6. The network topology image editing method according to claim 1, characterized in that: The editing agent includes at least one of an intention recognition agent, a topology information extraction agent, a topology calculation agent, a node editing agent, an edge editing agent, an information viewing agent and a dialogue agent.

7. The network topology image editing method according to claim 1, characterized in that: Generating a target topology image based on the edited network topology information, and generating a target network topology image corresponding to the target topology image, including: Determine the node graph and location information corresponding to the target topology image; Determine a node icon corresponding to each node graphic based on the node graphic and the position information; The node graphic in the target topology image is replaced by the node icon according to the location information to obtain the target network topology image.

8. A network topology image editing device, characterized in that: include: An intention recognition module is used to recognize the network topology image to be edited and the network topology editing requirement information, and obtain the target editing intention and editing parameter information; A first editing module, configured to edit the network topology information to be edited corresponding to the network topology image to be edited using an editing agent based on the target editing intent and the editing parameter information, to obtain the edited network topology information; wherein the editing agent is a logic code including an editing function; The second editing module is used to generate a target topology image based on the edited network topology information, and generate a target network topology image corresponding to the target topology image.

9. A network topology image editing device, characterized in that: include: Memory for storing computer programs; A processor, configured to execute the computer program to implement the steps of the network topology image editing method as claimed in any one of claims 1 to 7.

10. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the network topology image editing method according to any one of claims 1 to 7 are implemented.

11. A computer program product, characterized in that It comprises a computer program / instruction, which, when executed by a processor, implements the steps of the network topology image editing method as claimed in any one of claims 1 to 7.

Images