A design draft auditing method, device, equipment and medium

Abstract

The present application relates to the technical field of design draft processing, and particularly relates to a design draft auditing method and device, equipment and medium, data preprocessing is performed on an obtained design draft file to obtain target metadata corresponding to the design draft file, the target metadata is processed based on preset component recognition rules and hierarchical analysis algorithms to generate specific domain language data, the specific domain language data is audited in multiple dimensions to obtain an audit result of the design draft file, wherein the multi-dimensional auditing includes design element specification auditing, layer compliance auditing and rule violation item auditing. It can be seen that the present application generates target metadata through data preprocessing, converts the target metadata into DSL data based on preset component recognition rules and hierarchical analysis algorithms, and then utilizes multi-dimensional collaborative auditing of multi-dimensional auditing to realize multi-dimensional automatic auditing, effectively improve the efficiency and accuracy of design draft auditing, and improve the user experience.

Description

Technical Field

[0001] This invention relates to the field of design draft processing technology, and in particular to a design draft review method, apparatus, equipment and medium. Background Technology

[0002] During the development of digital products, technical design drafts, compliance reports, and other documents from various industries often require review. Design draft review is a key step in ensuring product visual consistency and development efficiency.

[0003] However, traditional design review methods typically involve manual review. Reviewers need to review a massive number of design drafts according to requirements. This manual review method is not only time-consuming and labor-intensive but also prone to errors, heavily relying on the reviewers' professional knowledge and experience. Although some design specification verification tools exist, most can only perform single-dimensional style checks and cannot cover complex scenarios such as layout rationality and component consistency, making it difficult to fundamentally solve the problems of review efficiency and collaboration costs. Therefore, how to effectively improve the efficiency and accuracy of design draft review is a pressing technical problem that needs to be solved. Summary of the Invention

[0004] Therefore, in order to address the above-mentioned technical problems, this invention provides a design draft review method, apparatus, equipment, and medium that can effectively improve the efficiency and accuracy of design draft review.

[0005] A first aspect of this application provides a design draft review method, the design draft review method comprising: The acquired design draft files are preprocessed to obtain the target metadata corresponding to the design draft files; Based on preset component identification rules and hierarchical parsing algorithms, the target metadata is transformed to generate domain-specific language data. The language data of the specific domain is reviewed in multiple dimensions to obtain the review results of the design draft file. The multi-dimensional review includes review of design element specifications, layer compliance, and review of violations.

[0006] A second aspect of this application provides a design draft review device, the design draft review device comprising: The processing module is used to preprocess the acquired design draft files to obtain the target metadata corresponding to the design draft files; The conversion module is used to convert the target metadata based on preset component recognition rules and hierarchical parsing algorithms to generate domain-specific language data. The review module is used to conduct multi-dimensional reviews of the language data in the specific domain and obtain the review results of the design draft file. The multi-dimensional review includes design element specification review, layer compliance review, and violation review.

[0007] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the design review method as described in the first aspect.

[0008] Fourthly, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, which, when executed by a processor, implements the design draft review method as described in the first aspect.

[0009] In summary, this invention provides a design draft review method, apparatus, device, and medium. It preprocesses the acquired design draft file to obtain target metadata corresponding to the design draft file. Based on preset component identification rules and hierarchical parsing algorithms, it transforms the target metadata to generate domain-specific language data. This domain-specific language data is then subjected to multi-dimensional review to obtain the review result of the design draft file. The multi-dimensional review includes design element specification review, layer compliance review, and violation review. Therefore, this application generates target metadata through data preprocessing, transforms the target metadata into DSL data based on preset component identification rules and hierarchical parsing algorithms, and then utilizes multi-dimensional collaborative review to achieve multi-dimensional automated review, effectively improving the efficiency and accuracy of design draft review and enhancing the user experience. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart illustrating a design draft review method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a design draft review device provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0013] It should be understood that, when used in this specification and the appended claims, terms include indicating the presence of the described feature, integral, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0014] It should also be understood that the terms used in this specification and the appended claims refer to any combination of one or more of the associated listed items and all possible combinations, and include such combinations.

[0015] As used in this specification and the appended claims, terms if can be interpreted in context as when... or once or in response to determination. Similarly, the phrase if determined or if matched to [described condition or event] can be interpreted in context as once determined or in response to determination or once matched to [described condition or event] or in response to matching to [described condition or event].

[0016] Furthermore, in the description of this invention and the appended claims, the terms first, second, third, etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0017] References to one or more embodiments described in this specification mean that a particular feature, structure, or characteristic described in connection with that embodiment is included in one or more embodiments of the invention. Therefore, phrases appearing in different parts of this specification as referring to one embodiment, some embodiments, some other embodiments, and others do not necessarily refer to the same embodiment, but rather mean one or more, but not all, embodiments, unless otherwise specifically emphasized. The terms include, comprise, have, and variations thereof mean including but not limited to, unless otherwise specifically emphasized.

[0018] It should be understood that the sequence number of each step in the following embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0019] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0020] See Figure 1 This is a flowchart illustrating a design draft review method according to an embodiment of the present invention, as shown below. Figure 1 As shown, this design review method can be implemented through the following steps.

[0021] S101: Perform data preprocessing on the acquired design draft file to obtain the target metadata corresponding to the design draft file.

[0022] S102: Based on preset component identification rules and hierarchical parsing algorithms, the target metadata is transformed to generate domain-specific language data.

[0023] S103: Perform multi-dimensional review on the specific domain language data to obtain the review results of the design draft file, wherein the multi-dimensional review includes design element specification review, layer compliance review, and violation item review.

[0024] In this embodiment, traditional manual review methods suffer from inefficiency and inaccuracy, heavily relying on the reviewers' expertise. Furthermore, existing technical tools can only perform single-dimensional style checks, failing to cover complex scenarios such as layout rationality and component consistency. The design draft file can be a currently being designed document or a design file obtained by calling the open API of mainstream design tools like Figma. The obtained design draft file undergoes data preprocessing to generate structured target metadata. This process extracts key feature information from the design draft and standardizes the data format, transforming non-standardized visual and interactive content into a machine-processable intermediate representation, thereby reducing the time-consuming and error-prone nature of manual parsing. Further, based on preset component recognition rules and hierarchical parsing algorithms, the target metadata is transformed into domain-specific language data. This domain-specific language data refers to a semantic representation specific to the design domain, which can be defined using XML format, such as using tags to describe component types and attributes, or organizing design elements through YAML files. Its main purpose is to translate design intent into machine-understandable language. Component identification rules are used to identify atomic and composite components in the design draft, while hierarchical parsing algorithms are used to determine the hierarchical connections between components. This transforms the design logic into a semantically clear, domain-specific language expression, providing a foundation for machine understanding in complex scenarios. Based on this, a multi-dimensional review is then performed on the DSL data to obtain the review results. This multi-dimensional review includes design element specification review, layer compliance review, and violation review. By simultaneously comparing with preset design specification standards, verifying layer naming protocols, and identifying potential risk content, a comprehensive review result is obtained.

[0025] In this embodiment, the synergistic effect of data preprocessing, generation of domain-specific language data, and multi-dimensional AI analysis avoids the time-consuming, labor-intensive, and error-prone nature of manual review. At the same time, it breaks through the limitations of single-dimensional inspection and achieves comprehensive coverage of complex scenarios such as layout rationality and component consistency, thereby improving the accuracy and reliability of design draft review.

[0026] In some embodiments described above in this application, one specific implementation of step S101 includes the following steps: Based on a preset feature processing model, key feature information is extracted from the design file to generate initial metadata, wherein the key feature information includes layer basic information, component type information, style parameter information, hierarchical relationship information, and interaction logic information. The initial metadata is converted into a data format to obtain the target metadata corresponding to the design file.

[0027] In this embodiment, the preset feature processing model refers to the algorithm model used to extract key feature information from the design draft file. It can be implemented using a convolutional neural network or a random forest model. The purpose is to ensure the complete extraction of key feature information. Here, we take the use of a pre-trained convolutional neural network as the feature processing model as an example to extract key feature information from the Figma format design draft file. Key feature information refers to the core data related to the review in the design draft file, including basic layer information (such as layer name and coordinate position), component type information (such as button or input box), style parameter information (such as font and color), hierarchical relationship information (such as nesting relationship), and interaction logic information (such as jump logic). Subsequently, the extracted initial metadata is converted into standardized JSON format target metadata through a JSON format converter to eliminate the risk of parsing errors caused by data format differences, so as to facilitate the use of subsequent component identification and review processes. Through the above technical solutions, this application can ensure that the key feature information of the design draft file is extracted completely and in a standardized format, so that the generated target metadata accurately represents the structure and behavior characteristics of the design draft, improves the accuracy of component-level parsing, the reliability of domain-specific language data generation, and the coverage depth of multi-dimensional review, thereby ensuring that the target metadata has high compatibility and operability in subsequent transformation and review steps, and provides reliable input for component-level parsing and multi-dimensional review.

[0028] In some embodiments described above in this application, one specific implementation of step S102 includes the following steps: Based on a pre-defined component feature library, a feature matching algorithm is used to identify atomic components and composite components in the design file. The hierarchical connection relationship between the atomic components and the composite components is determined using a hierarchical parsing algorithm; Based on the component hierarchical connection relationship, the target metadata and component attribute information are input into a specific language generation model to generate specific domain language data, wherein the component attribute information includes atomic component attribute information and composite component attribute information.

[0029] In this embodiment, during the design draft review process, the preset component feature library can specifically be a database storing feature vectors of common UI components, such as the style and structural features of components like buttons and text boxes. It can be implemented using a feature vector library built based on historical design draft data or a predefined component template library. Its purpose is to provide a unified recognition benchmark and avoid subjective bias in the recognition process. The feature matching algorithm can use an image recognition model based on convolutional neural networks to identify atomic components and composite components. The hierarchical parsing algorithm can use graph traversal algorithms or tree structure parsing algorithms to determine the parent-child relationship between components. The specific language generation model can be a rule-based template engine or a sequence-to-sequence model to convert target metadata and component attribute information into JSON format DSL data, where the component attribute information includes the size and color of atomic components and the layout parameters of composite components. First, based on a pre-defined component feature library, a feature matching algorithm is used to identify atomic and composite components in the design draft file, ensuring the standardization and accuracy of component identification. Then, a hierarchical parsing algorithm is used to determine the hierarchical connection relationships between atomic and composite components, automatically constructing the topological structure between components. Finally, based on these hierarchical connection relationships, target metadata and component attribute information are input into a specific language generation model to generate domain-specific language data. By fully utilizing the hierarchical relationship as the basis for generation, combined with the comprehensive input of target metadata and component attribute information, the DSL data accurately reflects the structural characteristics of the design draft. This sequential execution and information flow mechanism effectively solves the problems of ambiguous component identification and hierarchical structure errors, improving the accuracy of DSL data and the reliability of subsequent review. Through the above technical solution, this application effectively solves the problems of inaccurate identification and hierarchical errors caused by the lack of a clear mechanism in the component identification and hierarchical parsing process, ensuring the accuracy and structural integrity of DSL data generation, thereby improving the reliability and efficiency of design draft review.

[0030] In some embodiments described above in this application, one specific implementation of step S103 includes the following steps: The domain-specific language data is identified, and design elements existing in the domain-specific language data are extracted. The design elements include font specifications, spacing and position rules, color contrast, component standards, performance parameters, and naming protocols. The design elements are compared with the preset design specifications and standards, and based on the comparison results, the defective factors in the design elements that do not conform to the preset design specifications and standards are determined. Based on the aforementioned defect factors, the first review result of the design draft document is determined.

[0031] In this embodiment, identifying domain-specific language data refers to capturing the structured information of design elements through semantic parsing technology. This can be achieved using rule-based pattern matching methods or deep learning-based sequence labeling models. The aim is to efficiently extract key dimensions by leveraging the semantic expression characteristics of DSL data, avoiding the ambiguity caused by unstructured data in the original design draft. Comparing design elements with preset design specifications refers to an automated comparison mechanism, which can be implemented using threshold range verification or vector similarity calculation. The aim is to transform subjective experience judgment into objective quantitative analysis, eliminating review omissions caused by differences in standard interpretation. The preset design specifications are a set of exclusive specifications predefined to correspond to the design domain and application scenario of the design draft, rather than general design standards. Based on the comparison results, the defect factors in the design elements that do not conform to the preset design specifications are identified. Determining the first review result based on the defect factors refers to generating a review conclusion based on the defects. This can be achieved using classification decision trees or confidence-weighted scoring systems. The aim is to closely link the review results with specific problem points, providing a clear path for subsequent corrections. The process begins by identifying language data specific to a particular domain to extract design elements. DSL data is then used as a pre-processed semantic representation to ensure key dimensions such as font specifications and spacing rules are fully captured. Next, the extracted design elements are automatically compared with pre-defined standards, accurately identifying deviations through clearly defined thresholds. Finally, a first review result is generated based on the defect factors, ensuring the review conclusion directly reflects specific violations. This process, through the organic integration of structured data identification and systematic comparison, achieves a shift from extensive inspection to refined verification, ensuring the completeness of the review coverage and the operability of the results.

[0032] For example, when reviewing the interface design drafts of an e-commerce platform, language data specific to the domain is identified as containing font specification information (such as the font type being sans-serif and the size not conforming to preset specification requirements) and spacing rule information (such as element spacing exceeding the standard range). These design elements are automatically compared with preset design specification standards to determine the defect factors of font size and spacing not conforming to the specifications, and a review result is generated accordingly, clearly marking the specific location of the non-compliant design elements and the direction of correction. Through the above solution, this application can accurately locate design defects and achieve quantitative analysis, significantly improving the efficiency and accuracy of design draft review, and effectively solving the problem of defect omissions caused by review blind spots in complex scenarios.

[0033] In one specific embodiment of step S103 of the above-described embodiments of this application, the following step is further included: The specific domain language data is input into the text review model to obtain keywords of various layer types covered in the specific domain language data; Each layer type keyword is matched with a pre-defined name for each layer type to obtain the current layer type; The current layer type is checked for style compliance using an algorithm, resulting in a second review result for the design file.

[0034] In this embodiment, inputting domain-specific language data into the text review model refers to automatically parsing the semantic structure of the design draft using natural language processing technology. This can be achieved using a deep learning model based on the Transformer architecture, such as BERT or RoBERTa, with the aim of efficiently extracting layer type keywords, avoiding subjective errors from manual intervention, and ensuring comprehensive coverage. Matching each layer type keyword with each name of each pre-set layer type refers to performing precise comparison based on a pre-set standardized name library. This can be achieved using string matching algorithms or fuzzy matching algorithms, with the aim of resolving ambiguities in layer type identification and providing a reliable foundation for subsequent checks. Performing style compliance checks on the current layer type using an inspection algorithm refers to applying customized verification rules to the identified layer type. This can be achieved using a rule engine or machine learning model, with the aim of achieving targeted and in-depth review and ensuring the accuracy of style specification verification. First, domain-specific language data is input into the text review model to automatically extract layer type keywords. This is thanks to the structured text characteristics of DSL data, which allows for efficient parsing of the semantic information in the design draft. Next, the extracted keywords are matched against a pre-defined name database to accurately determine the current layer type, providing a reliable basis for type identification. Finally, a customized style compliance check algorithm is executed on the identified layer type. This algorithm reviews the layer according to pre-specified specifications for the current layer type to determine where it fails to comply, thus achieving in-depth verification of the specific layer type and obtaining the second review result of the design draft file. Through this sequential execution and information flow mechanism, the entire process organically integrates text parsing, type matching, and style checking, forming a complete automated layer compliance review chain.

[0035] It should be noted that, considering that different users may use different names for the same type of layer, and some users may not even be able to describe the layer type by name, but can only describe it by its color, shape, etc., the input text describing the scope of inspection is diverse and cannot be guaranteed to be standardized. Therefore, in this embodiment of the application, a text review model is used to perform semantic analysis on the language data of a specific domain to determine the layer type that needs to be reviewed, which is then used as the current layer type.

[0036] For example, when reviewing a user interface design draft, domain-specific language data is input into a text review model to identify layer type keywords such as "button" and "text_field". These keywords are matched against a pre-defined layer type name library to determine the types as "button" and "text box". Subsequently, for the "button" type, the algorithm verifies whether its corner radius, background color, and font size comply with design specifications. Simultaneously, for the "text box" type, its border style and padding parameters are checked, ultimately generating a review result report containing specific violations. Through the above technical solution, this application achieves automated identification and targeted checking of layer compliance review, accurately obtaining layer types and performing customized style verification, significantly reducing reliance on manual review, improving review efficiency, and effectively avoiding the omission of style violations in complex design scenarios.

[0037] In one specific embodiment of step S103 of the above-described embodiments of this application, the following step is further included: Based on a preset violation review model, the language data of the specific domain is reviewed to obtain the target violation content set and violation confidence score of the design draft file. The target violation content set includes a semantic violation content set, a keyword violation content set, and an image violation content set. Based on the target set of violations and the violation confidence score of the design draft file, a third review result of the design draft file is generated.

[0038] In this embodiment, a pre-defined violation review model receives domain-specific language data and leverages the characteristics of structured data input to achieve efficient parsing of the semantic logic of the design draft. This violation review model can employ a deep learning model based on the Transformer architecture. Subsequently, the model synchronously generates a target violation content set and a violation confidence score. The target violation content set is a categorized collection of violation information identified during the review process, which can be classified and stored according to three dimensions: semantic violations, keyword violations, and image violations. This violation review model refers to a dedicated analysis module for automatically detecting violation content in design drafts. It can be implemented using deep learning neural networks or a rule engine, and its purpose is to perform efficient semantic parsing and violation identification of structured DSL data. To ensure the review covers various scenarios, including logical inconsistencies in design intent, deviations in naming conventions, and compliance of visual elements, the violation confidence score is a quantitative indicator of the model's certainty regarding the identification results. It can be expressed as a probability value or a confidence percentage, aiming to objectively assess the severity of violations and reduce subjective bias. Finally, the system makes a comprehensive decision based on the type of violation and its confidence score, prioritizing high-risk violations through a confidence ranking mechanism. This forms a complete closed-loop review process from data input to result generation. Specifically, for each identified violation, the model calculates a confidence score ranging from 0.0 to 1.0. Ultimately, the system integrates the three types of violation sets with their corresponding confidence scores to generate a third review result containing violation type, location information, and risk level. Through this approach, this application achieves refined classification and quantitative assessment of violations in design drafts, accurately distinguishing different dimensions of violations and objectively measuring their severity. This effectively reduces omissions and misjudgments during the review process, improving the reliability of review results and the efficiency of subsequent corrections.

[0039] In some embodiments described above in this application, one specific embodiment following step S103 includes the following steps: Determine whether the review result meets the preset requirements for an approval result; If the review result meets the preset requirements for approval, the analysis data and component hierarchy of all design elements in the design draft file are applied and calculated, and the review indicator data is output in the form of a report. If the review result does not meet the preset requirements for the approval result, the review result will be corrected and feedback will be provided.

[0040] In this embodiment, since the review results include a first review result, a second review result, and a third review result, after obtaining the review results of the design draft files, the system first automatically evaluates the compliance of each review result sequentially based on preset standards. Specifically, the system calls a preset rule engine to sequentially determine the compliance of each review result. When each review result meets the preset requirements, the system automatically integrates the analysis data of the design elements and the component hierarchy, and outputs a structured report containing key indicators through a report generation engine. When a review result fails to meet the requirements, the system automatically pushes the problem details to the design collaboration platform and triggers a correction feedback mechanism, accurately pushing the problem to the relevant stage. This sequential execution decision framework ensures that the review results can be seamlessly connected to subsequent operations, avoiding delays and errors caused by manual intervention, thereby achieving automated closed-loop management of the review process. The process of determining whether the review results meet the preset requirements for approval refers to automatically assessing the compliance of the review results based on preset standards. This can be achieved using a rule matching engine or a threshold comparison algorithm, aiming to avoid the tedious process of repeated manual verification and ensure the objectivity and timeliness of the judgment process. Applying the analysis data and component hierarchy relationships of all design elements in the design draft files and outputting review indicator data in report form can be understood as transforming the review results into a structured report output. This can be achieved using templated report generation technology or data visualization tools, aiming to enable the development team to directly obtain actionable indicator data without additional parsing or organization, significantly improving data utilization efficiency. Providing feedback on the review results refers to automatically triggering a correction mechanism for results that do not meet the requirements. This can be achieved using an integrated issue tracking system or an automatic feedback interface, aiming to accurately relay issues to the relevant stages, avoiding issue backlog and communication delays, and accelerating the iterative optimization of the design drafts. The system first automatically assesses the compliance of the review results based on preset standards, triggering subsequent differentiated processing procedures. When the review results meet the preset requirements, the system automatically integrates the analysis data of design elements and component hierarchy relationships to generate a structured report. When the review results do not meet the requirements, the system immediately initiates a correction feedback mechanism, accurately pushing the issue to the relevant stage. Through this approach, this application achieves automatic judgment and differentiated response of review results, significantly shortening the review cycle, reducing the risk of errors due to human negligence, and improving data utilization efficiency and design iteration speed, thereby realizing automated closed-loop management of the review process.

[0041] In summary, this invention provides a design draft review method, apparatus, device, and medium. It preprocesses the acquired design draft file to obtain target metadata corresponding to the design draft file. Based on preset component identification rules and hierarchical parsing algorithms, it transforms the target metadata to generate domain-specific language data. This domain-specific language data is then subjected to multi-dimensional review to obtain the review result of the design draft file. The multi-dimensional review includes design element specification review, layer compliance review, and violation review. Therefore, this application generates target metadata through data preprocessing, transforms the target metadata into DSL data based on preset component identification rules and hierarchical parsing algorithms, and then utilizes multi-dimensional collaborative review to achieve multi-dimensional automated review, effectively improving the efficiency and accuracy of design draft review and enhancing the user experience.

[0042] Please see Figure 2 , Figure 2 This is a schematic diagram of the design draft review device provided in an embodiment of the present invention. This design draft review device corresponds one-to-one with the design draft review method in the above embodiments. Please refer to [link / reference] for details. Figure 1 as well as Figure 1 The relevant descriptions in the corresponding embodiments are shown below. For ease of explanation, only the parts relevant to this embodiment are shown. See also... Figure 2 The design draft review device 20 includes: a processing module 21, a conversion module 22, and a review module 23.

[0043] Processing module 21 is used to preprocess the acquired design draft file to obtain the target metadata corresponding to the design draft file; The conversion module 22 is used to convert the target metadata based on preset component recognition rules and hierarchical parsing algorithms to generate domain-specific language data. The review module 23 is used to conduct multi-dimensional review of the language data in the specific domain and obtain the review results of the design draft file. The multi-dimensional review includes design element specification review, layer compliance review, and violation review.

[0044] Optionally, the above processing module 21 is specifically used for: Based on a preset feature processing model, key feature information is extracted from the design file to generate initial metadata, wherein the key feature information includes layer basic information, component type information, style parameter information, hierarchical relationship information, and interaction logic information. The initial metadata is converted into a data format to obtain the target metadata corresponding to the design file.

[0045] Optionally, the conversion module 22 described above is specifically used for: Based on a pre-defined component feature library, a feature matching algorithm is used to identify atomic components and composite components in the design file. The hierarchical connection relationship between the atomic components and the composite components is determined using a hierarchical parsing algorithm; Based on the component hierarchical connection relationship, the target metadata and component attribute information are input into a specific language generation model to generate specific domain language data, wherein the component attribute information includes atomic component attribute information and composite component attribute information.

[0046] Optionally, the aforementioned audit module 23 is specifically used for: The domain-specific language data is identified, and design elements existing in the domain-specific language data are extracted. The design elements include font specifications, spacing and position rules, color contrast, component standards, performance parameters, and naming protocols. The design elements are compared with the preset design specifications and standards, and based on the comparison results, the defective factors in the design elements that do not conform to the preset design specifications and standards are determined. Based on the aforementioned defect factors, the review result of the design draft document is determined.

[0047] Optionally, the aforementioned audit module 23 is also used for: The specific domain language data is input into the text review model to obtain keywords of various layer types covered in the specific domain language data; Each layer type keyword is matched with a pre-defined name for each layer type to obtain the current layer type; The design file review result is obtained by checking the style compliance of the current layer type using an algorithm.

[0048] Optionally, the aforementioned audit module 23 is also used for: Based on a preset violation review model, the language data of the specific domain is reviewed to obtain the target violation content set and violation confidence score of the design draft file. The target violation content set includes a semantic violation content set, a keyword violation content set, and an image violation content set. Based on the target set of violations and the violation confidence score of the design draft file, the review result of the design draft file is generated.

[0049] Optionally, the audit module 23 mentioned above is specifically used for: Determine whether the review result meets the preset requirements for an approval result; If the review result meets the preset requirements for approval, the analysis data and component hierarchy of all design elements in the design draft file are applied and calculated, and the review indicator data is output in the form of a report. If the review result does not meet the preset requirements for the approval result, the review result will be corrected and feedback will be provided.

[0050] It should be noted that the information interaction and execution process between the above-mentioned units are based on the same concept as the method embodiments of the present invention. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0051] Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Figure 3 As shown, the computer device of this embodiment includes: at least one processor ( Figure 3 Only one is shown in the diagram), a memory, and a computer program stored in the memory and capable of running on at least one processor, wherein the processor executes the computer program to implement the steps in any of the above-described design draft review method embodiments.

[0052] This computer device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that... Figure 3 The examples of computer devices are merely examples and do not constitute a limitation on computer devices. Computer devices may include more or fewer components than shown in the illustration, or combinations of certain components, or different components, such as network interfaces, displays, and input systems.

[0053] In one embodiment, a computer-readable storage medium is provided that, when the instructions in the computer-readable storage medium are executed by a processor in a computer device, enables the computer device to perform the steps of any embodiment of the design draft review method disclosed in this invention, which will not be repeated here. The computer-readable storage medium may be non-volatile or volatile.

[0054] The processor referred to can be a CPU, but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0055] Memory includes readable storage media, internal memory, etc., wherein internal memory can be the RAM of a computer device, providing an environment for the operation of the operating system and computer-readable instructions stored in the readable storage media. The readable storage media can be the hard drive of the computer device, or in other embodiments, it can be an external storage device of the computer device, such as a plug-in hard drive, SmartMediaCard (SMC), SecureDigital (SD) card, or FlashCard. Furthermore, memory can include both internal storage units and external storage devices of the computer device. Memory is used to store the operating system, cooperative applications, bootloader, data, and other programs, such as program code of computer programs. Memory can also be used to temporarily store data that has been output or will be output.

[0056] Those skilled in the art will understand that implementing all or part of the processes in the above embodiments can be accomplished by a computer program instructing related hardware. This computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0057] Those familiar with the technical field will understand that, for ease of description and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this invention. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here. If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0058] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A design draft review method, characterized in that, The method includes: The acquired design draft files are preprocessed to obtain the target metadata corresponding to the design draft files; Based on preset component identification rules and hierarchical parsing algorithms, the target metadata is transformed to generate domain-specific language data. The language data of the specific domain is reviewed in multiple dimensions to obtain the review results of the design draft file. The multi-dimensional review includes review of design element specifications, layer compliance, and review of violations.

2. The design draft review method as described in claim 1, characterized in that, The step of preprocessing the acquired design draft file to obtain the target metadata corresponding to the design draft file includes: Based on a preset feature processing model, key feature information is extracted from the design file to generate initial metadata, wherein the key feature information includes layer basic information, component type information, style parameter information, hierarchical relationship information, and interaction logic information. The initial metadata is converted into a data format to obtain the target metadata corresponding to the design file.

3. The design draft review method as described in claim 1, characterized in that, The target metadata is transformed and processed based on preset component identification rules and hierarchical parsing algorithms to generate domain-specific language data, including: Based on a pre-defined component feature library, a feature matching algorithm is used to identify atomic components and composite components in the design file. The hierarchical connection relationship between the atomic components and the composite components is determined using a hierarchical parsing algorithm; Based on the component hierarchical connection relationship, the target metadata and component attribute information are input into a specific language generation model to generate specific domain language data, wherein the component attribute information includes atomic component attribute information and composite component attribute information.

4. The design draft review method as described in claim 1, characterized in that, The review results include a first review result. The multi-dimensional review of the language data in the specific domain to obtain the review results of the design draft file includes: The domain-specific language data is identified, and design elements existing in the domain-specific language data are extracted. The design elements include font specifications, spacing and position rules, color contrast, component standards, performance parameters, and naming protocols. The design elements are compared with the preset design specifications and standards, and based on the comparison results, the defective factors in the design elements that do not conform to the preset design specifications and standards are determined. Based on the aforementioned defect factors, the first review result of the design draft document is determined.

5. The design draft review method as described in claim 1, characterized in that, The review results include a second review result. The multi-dimensional review of the specific domain language data to obtain the review results of the design draft file includes: The specific domain language data is input into the text review model to obtain keywords of various layer types covered in the specific domain language data; Each layer type keyword is matched with a pre-defined name for each layer type to obtain the current layer type; The current layer type is checked for style compliance using an algorithm, resulting in a second review result for the design file.

6. The design draft review method as described in claim 1, characterized in that, The review results include a third review result. The multi-dimensional review of the language data in the specific domain to obtain the review results of the design draft file includes: Based on a preset violation review model, the language data of the specific domain is reviewed to obtain the target violation content set and violation confidence score of the design draft file. The target violation content set includes a semantic violation content set, a keyword violation content set, and an image violation content set. Based on the target set of violations and the violation confidence score of the design draft file, a third review result of the design draft file is generated.

7. The design draft review method as described in claim 1, characterized in that, After obtaining the review results of the design draft file, the process includes: Determine whether the review result meets the preset requirements for an approval result; If the review result meets the preset requirements for approval, the analysis data and component hierarchy of all design elements in the design draft file are applied and calculated, and the review indicator data is output in the form of a report. If the review result does not meet the preset requirements for the approval result, the review result will be corrected and feedback will be provided.

8. A design draft review device, characterized in that, include: The processing module is used to preprocess the acquired design draft files to obtain the target metadata corresponding to the design draft files; The conversion module is used to convert the target metadata based on preset component recognition rules and hierarchical parsing algorithms to generate domain-specific language data. The review module is used to conduct multi-dimensional reviews of the language data in the specific domain and obtain the review results of the design draft file. The multi-dimensional review includes design element specification review, layer compliance review, and violation review.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the design draft review method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the design draft review method as described in any one of claims 1 to 7.

Images