Electronic data acquisition and analysis plug-in generation method and system based on AI large model

Abstract

The invention discloses an electronic data acquisition and analysis plug-in generation method and system based on an AI large model, and the method comprises the steps: obtaining an electronic data acquisition demand and a cue word frame inputted by a user, and enabling the cue word frame to comprise role setting, target description, skill endowing and limiting conditions; processing the data according to the cue word framework and the workflow arrangement mode to generate data suitable for large model processing; inputting the processed data and the cue word into an AI large model, and generating a plug-in code snippet aiming at the electronic data acquisition requirement by the AI large model; integrating the generated code snippets into a complete electronic data acquisition plug-in code by using a code merging tool; running the electronic data on the plug-in platform to obtain a plug-in code, verifying the correctness of the function of the plug-in code, and if the requirement is not met, adjusting the cue word and repeating the steps; and after verification is completed, the generated electronic data is subjected to plug-in deployment application. According to the method, standardization and automation of electronic data acquisition plug-in development can be realized, and the electronic data acquisition working efficiency and the code quality are improved.

Description

Technical Field

[0001] This invention relates to the technical field of computer electronic data acquisition, and in particular to a method and system for generating electronic data acquisition and analysis plugins based on AI large models. Background Technology

[0002] In current electronic data acquisition and analysis work, plugin development faces problems such as slow iteration speed and complex development processes. Traditionally, both simple and complex plugins rely on manual coding, which is not only inefficient but also makes it difficult to guarantee code standardization and maintainability. Furthermore, when facing new electronic data acquisition needs, the process from R&D research to plugin deployment often takes nearly a week, severely impacting the timeliness of electronic data acquisition work. In addition, when parsing data related to illegal applications, traditional methods require searching and interpreting the meaning of each field individually, which is inefficient and difficult to guarantee accuracy. Current electronic data acquisition and analysis plugin development solutions have the following shortcomings:

[0003] 1. Low development efficiency: Traditional electronic data acquisition and analysis plugin development relies heavily on manual coding. From requirements analysis to code writing and testing, each step requires a significant investment of time and effort. Developing complex plugins often involves lengthy development cycles, failing to meet rapidly changing electronic data acquisition needs.

[0004] 2. Inconsistent code quality: Manually written code is greatly affected by the developer's personal technical level and programming habits, making it difficult to guarantee the code's standardization, readability, and maintainability.

[0005] 3. Lack of intelligence and adaptability: Traditional technologies struggle to automatically generate plugin code based on users' natural language descriptions, requiring users to possess professional programming knowledge and skills. Furthermore, when faced with newly emerging illegal applications, traditional technologies cannot quickly adapt to changes in their data structures and encryption methods, necessitating extensive new research and development.

[0006] Therefore, there is an urgent need for a technology that can standardize and automate the development of electronic data acquisition and analysis plugins to improve the efficiency and quality of electronic data acquisition work. Summary of the Invention

[0007] To address the aforementioned technical problems in the existing technology, this invention proposes a method and system for generating electronic data acquisition and analysis plugins based on AI large models, thereby solving the above-mentioned technical problems.

[0008] According to a first aspect of the present invention, a method for generating electronic data acquisition and analysis plugins based on AI large models is proposed, comprising:

[0009] S1: Obtain the user's input electronic data acquisition requirements and prompt word framework, which includes role setting, goal description, skill assignment and constraints;

[0010] S2: Process the data according to the prompt word framework and workflow arrangement to generate data suitable for large model processing;

[0011] S3: Input the processed data and prompts into the AI ​​big model, which will generate plug-in code snippets for electronic data acquisition needs;

[0012] S4: Use a code merging tool to integrate the generated code snippets into complete electronic data acquisition plugin code;

[0013] S5: Run the electronic data acquisition plugin code on the plugin platform to verify the correctness of its function. If it does not meet the requirements, adjust the prompt words and repeat the above steps.

[0014] S6: After verification, deploy the generated electronic data acquisition plugin into the application.

[0015] In some specific embodiments, the prompt word framework in S1 specifically includes: the roles are set as a combination of professional parsing experts and Python data parsing experts for the type of application to be parsed, and the specific professional fields are determined according to the nature of the application and the characteristics of the data; the goal is described as generating Python code that can accurately parse various fields in the application configuration file according to the provided sample code framework structure; the skills are provided including proficiency in the advanced features of the Python programming language and the use of common libraries, mastery of various data parsing and processing techniques and algorithms, and a deep understanding and accurate application of configuration file reading and parsing methods.

[0016] In some specific embodiments, the restrictions include prohibiting changes to the sample code framework, avoiding the use of specific keywords that are likely to cause confusion or errors, ensuring that all field values ​​are stored in the prescribed string format, and requiring the generation of complete and uninterrupted code. For the generation of financial data electronic data acquisition plugins, the restrictions also include compliance with financial data security and compliance standards.

[0017] In some specific embodiments, the workflow orchestration methods specifically include:

[0018] Determine the format type of input data, and distinguish between plain text, database files, binary data, and niche and unconventional data formats;

[0019] For niche and unconventional data formats, start the conversion program to convert them to JSON format. The conversion process follows the principle of consistency in data structure and semantics.

[0020] For binary data, key information is extracted based on the type of binary data and the application scenario, and then converted into a standard, readable text format.

[0021] In some specific embodiments, the AI ​​large model includes a natural language understanding layer, a code generation core layer, and a data adaptation and optimization layer. The natural language understanding layer is configured to receive user input, including a prompt word framework containing role settings, target descriptions, skill assignments, and constraints, as well as processed data. It uses deep learning algorithms and pre-trained language models to perform semantic analysis and intent recognition on the prompt words, extract key information and task requirements, and convert them into a structured representation suitable for internal model processing. The code generation core layer is configured to generate plug-in code fragments that meet the needs of electronic data acquisition based on the output of the natural language understanding layer. It utilizes a knowledge base containing a large number of code templates and algorithm logic related to electronic data acquisition, combined with a generative model architecture of deep learning, and learns and adapts to different application scenarios and data types. The data adaptation and optimization layer dynamically adapts to the characteristics and format of the input data during the code generation process. It extracts and optimizes features of the pre-processed results of niche and unconventional data formats and binary data, providing high-quality input that meets data specifications and task requirements for the code generation core layer. At the same time, it verifies and adjusts the compatibility between the code and the data after the code is generated.

[0022] In some specific embodiments, S5 specifically includes: deploying the merged electronic data acquisition plugin code to the plugin platform for running and debugging, simulating the actual electronic data acquisition working environment and data input, and observing whether the output results meet the expected electronic data acquisition goals and data parsing requirements; if the debugging results do not meet expectations, establishing a feedback mechanism, automatically backtracking to the prompt word framework or workflow orchestration steps, prompting the user to adjust the electronic data acquisition requirements, optimize the prompt word content, or improve the workflow processing flow according to the error information and debugging results, and restarting the entire method process until a plugin that meets the electronic data acquisition requirements is generated.

[0023] According to a second aspect of the invention, a computer-readable storage medium is provided on which one or more computer programs are stored, which, when executed by a computer processor, implement the method described above.

[0024] According to a third aspect of the present invention, an electronic data acquisition and analysis plug-in generation system based on an AI large model is proposed, comprising:

[0025] The user interaction module is configured to obtain electronic data acquisition requirements and prompt word frameworks for user input. The prompt word framework includes role setting, goal description, skill assignment, and constraints.

[0026] The data processing and workflow management module is configured to process data based on the prompt word framework and workflow arrangement to generate data suitable for large model processing.

[0027] The code merging and verification module is configured to input processed data and prompts into an AI big model, which then generates plugin code snippets tailored to the electronic data acquisition (EDI) requirements. A code merging tool is used to integrate these code snippets into a complete EDI plugin code. The EDI plugin code is then run on the plugin platform to verify its functionality. After verification, the generated EDI plugin is deployed and applied.

[0028] In some specific embodiments, the prompt word framework specifically includes: the roles are set as a combination of professional parsing experts and Python data parsing experts for the type of application to be parsed, with the specific professional fields determined according to the nature of the application and the characteristics of the data; the goal is described as generating Python code that can accurately parse various fields in the application configuration file according to the provided sample code framework structure; the skills provided include proficiency in the advanced features of the Python programming language and the use of common libraries, mastery of various data parsing and processing techniques and algorithms, and a deep understanding and accurate application of configuration file reading and parsing methods; the constraints include prohibiting changes to the sample code framework, avoiding the use of specific keywords that are likely to cause confusion or errors, ensuring that all field values ​​are stored in the specified string format, and requiring the generation of complete and missing code. For the generation of financial data electronic data acquisition plugins, the constraints also include compliance with financial data security and compliance standards.

[0029] In some specific embodiments, the workflow orchestration method includes: determining the format type of the input data, distinguishing between plain text, database files, binary data, and niche unconventional data formats; for niche unconventional data formats, starting a conversion program to convert them into JSON format, with the conversion process following the principle of consistency between data structure and semantics; for binary data, extracting key information based on the type and application scenario of the binary data, and converting it into a standard readable text format.

[0030] In some specific embodiments, the AI ​​large model includes a natural language understanding layer, a code generation core layer, and a data adaptation and optimization layer. The natural language understanding layer is configured to receive user input, including a prompt word framework containing role settings, target descriptions, skill assignments, and constraints, as well as processed data. It uses deep learning algorithms and pre-trained language models to perform semantic analysis and intent recognition on the prompt words, extract key information and task requirements, and convert them into a structured representation suitable for internal model processing. The code generation core layer is configured to generate plug-in code fragments that meet the needs of electronic data acquisition based on the output of the natural language understanding layer. It utilizes a knowledge base containing a large number of code templates and algorithm logic related to electronic data acquisition, combined with a generative model architecture of deep learning, and learns and adapts to different application scenarios and data types. The data adaptation and optimization layer dynamically adapts to the characteristics and format of the input data during the code generation process. It extracts and optimizes features of the pre-processed results of niche and unconventional data formats and binary data, providing high-quality input that meets data specifications and task requirements for the code generation core layer. At the same time, it verifies and adjusts the compatibility between the code and the data after the code is generated.

[0031] In some specific embodiments, the code merging and verification module specifically includes: deploying the merged electronic data acquisition plugin code to the plugin platform for running and debugging, simulating the actual electronic data acquisition working environment and data input, and observing whether the output results meet the expected electronic data acquisition goals and data parsing requirements; if the debugging results do not meet expectations, a feedback mechanism is established to automatically backtrack to the prompt word framework or workflow orchestration steps, prompting the user to adjust the electronic data acquisition requirements, optimize the prompt word content, or improve the workflow processing flow based on the error information and debugging results, and restarting the entire method process until a plugin that meets the electronic data acquisition requirements is generated.

[0032] This invention proposes a method and system for generating electronic data acquisition and analysis plugins based on AI large-scale models, which has the following technical effects:

[0033] 1. Automated Plugin Generation Mechanism: Based on a large AI model, a unique electronic data acquisition (EDI) code generation framework is constructed. Through the design of a specialized EDI agent prompting framework and code examples, the system automatically generates executable plugin code from the user's natural language description of their EDI requirements. This approach significantly reduces the workload of manual coding; the manual coding percentage for complex plugins drops from 100% to 20%, and simple plugins can even be entirely generated by machines. This significantly improves plugin development efficiency and changes the traditional development model that relies heavily on manual coding.

[0034] 2. Prompt Templates and Optimization Methods: The proposed prompt templates cover multiple dimensions, including roles, objectives, skills, and sample code, and can guide large-scale models to generate plugin code for various applications. Furthermore, through continuous optimization of prompts and samples, the large-scale model can automatically understand configuration information and accurately extract the meaning of internal fields. It can automatically determine the application type based on input data, effectively addressing the diversity of non-compliant applications and overcoming the limitations of previous prompt design for single functions or specific applications.

[0035] 3. Workflow Orchestration for Complex Data: Addressing the challenge of parsing diverse data formats involved in complex application plugins, an innovative workflow-based orchestration approach is adopted. For niche and unconventional data formats, they are first converted to common formats before analysis. For binary data, manual pre-processing is performed before code generation. Simultaneously, code merging tools are used to handle the parsing of multiple configuration files, ensuring comprehensive and accurate parsing of complex application data. This solves the difficulties of traditional methods in handling complex data formats and improves the completeness and accuracy of electronic data acquisition and parsing.

[0036] 4. Rapid Response to Needs: When dealing with situations where the product does not support electronic data acquisition in real-world scenarios, leveraging intelligent agents and existing infrastructure, and through simple adjustments to prompts or workflows, a significant amount of coding and debugging is eliminated. This drastically reduces plugin deployment time from nearly a week to one or two days, greatly improving the responsiveness to electronic data acquisition needs and enhancing its practicality and flexibility in real-world application scenarios. Attached Figure Description

[0037] The accompanying drawings are included to provide a further understanding of the embodiments and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments and, together with the description, serve to explain the principles of the invention. Other embodiments and many anticipated advantages of the embodiments will be readily recognized as they become better understood through reference to the following detailed description. Other features, objects, and advantages of this application will become more apparent from reading the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0038] Figure 1 This is a flowchart of an embodiment of the electronic data acquisition and analysis plugin generation method based on an AI large model, according to one of the present applications.

[0039] Figure 2 This is a flowchart of a specific embodiment of the electronic data acquisition and analysis plugin generation method based on an AI large model, according to this application.

[0040] Figure 3 This is a flowchart illustrating a specific embodiment of the workflow orchestration method of this application;

[0041] Figure 4This is a flowchart illustrating the generation and implementation of an electronic data acquisition and analysis plugin according to a specific embodiment of this application.

[0042] Figure 5 This is a schematic diagram of an electronic data acquisition plugin for a soft router, based on a specific embodiment of this application.

[0043] Figure 6 This is a schematic diagram of the electronic data acquisition result interface of a specific embodiment of this application;

[0044] Figure 7 An embodiment of this application generates a system architecture diagram for an AI-based large-scale model-based electronic data acquisition and analysis plugin;

[0045] Figure 8 A schematic diagram of the structure of a computer system used to implement the electronic device of the present application. Detailed Implementation

[0046] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0047] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0048] Figure 1 A flowchart illustrating a method for generating an AI-based large-scale electronic data acquisition and analysis plugin according to an embodiment of this application is shown. Figure 1 As shown, the method includes the following steps:

[0049] S1: Obtain the user's input electronic data acquisition requirements and prompt word framework. The prompt word framework includes role setting, goal description, skill assignment, and constraints.

[0050] In a specific embodiment, the prompt word framework includes: a combination of professional parsing experts and Python data parsing experts, with the specific professional fields determined based on the application type and data characteristics; the objective is to generate Python code that can accurately parse various fields in the application configuration file, following the provided sample code framework structure; skills include proficiency in advanced features of the Python programming language and the use of common libraries, mastery of various data parsing and processing techniques and algorithms, and a deep understanding and accurate application of configuration file reading and parsing methods; and restrictions include prohibiting changes to the sample code framework, avoiding the use of specific keywords that are likely to cause confusion or errors, ensuring that all field values ​​are stored in the specified string format, and requiring the generation of complete and undamaged code. For the generation of financial data electronic data acquisition plugins, restrictions also include compliance with financial data security and compliance standards.

[0051] S2: Process the data according to the prompt word framework and workflow arrangement to generate data suitable for large model processing.

[0052] In a specific embodiment, the workflow orchestration method includes: determining the format type of the input data, distinguishing between ordinary text, database files, binary data, and niche unconventional data formats; for niche unconventional data formats, starting a conversion program to convert them into JSON format, with the conversion process following the principle of consistency between data structure and semantics; for binary data, extracting key information based on the type and application scenario of the binary data, and converting it into a standard readable text format.

[0053] S3: Input the processed data and prompts into the AI ​​model, which will then generate plugin code snippets tailored to the electronic data acquisition needs.

[0054] In a specific embodiment, the AI ​​large model includes a natural language understanding layer, a code generation core layer, and a data adaptation and optimization layer. The natural language understanding layer is configured to receive user input, including a prompt word framework containing role settings, target descriptions, skill assignments, and constraints, as well as processed data. It uses deep learning algorithms and pre-trained language models to perform semantic analysis and intent recognition on the prompt words, extracting key information and task requirements, and converting them into a structured representation suitable for internal model processing. The code generation core layer is configured to generate plug-in code fragments that meet the needs of electronic data acquisition based on the output of the natural language understanding layer. It utilizes a knowledge base containing a large number of code templates and algorithm logic related to electronic data acquisition, combined with a generative model architecture of deep learning, and learns and adapts to different application scenarios and data types. The data adaptation and optimization layer dynamically adapts to the characteristics and format of the input data during the code generation process. It extracts and optimizes features of the pre-processed results of niche and unconventional data formats and binary data, providing high-quality input that meets data specifications and task requirements for the code generation core layer. At the same time, it verifies and adjusts the compatibility between the code and the data after code generation.

[0055] S4: Use a code merging tool to integrate the generated code snippets into complete electronic data acquisition plugin code.

[0056] S5: Run the electronic data acquisition plugin code on the plugin platform to verify the correctness of its function. If it does not meet the requirements, adjust the prompt words and repeat the above steps.

[0057] In a specific embodiment, the merged electronic data acquisition plugin code is deployed to the plugin platform for running and debugging, simulating the actual electronic data acquisition working environment and data input, and observing whether the output results meet the expected electronic data acquisition goals and data parsing requirements. If the debugging results do not meet expectations, a feedback mechanism is established to automatically backtrack to the prompt word framework or workflow orchestration steps, prompting the user to adjust the electronic data acquisition requirements, optimize the prompt word content, or improve the workflow processing flow based on the error information and debugging results, and restart the entire method process until a plugin that meets the electronic data acquisition requirements is generated.

[0058] S6: After verification, deploy the generated electronic data acquisition plugin into the application.

[0059] Compared with existing technologies, this method for generating electronic data acquisition and analysis plugins based on AI large models has the following characteristics:

[0060] 1. Significantly improve development efficiency: Based on an AI big data model, an electronic data acquisition code generation framework is built. By designing an electronic data acquisition intelligent agent prompt word framework and code examples, it is possible to automatically generate electronic data acquisition requirements from user natural language descriptions to runnable plug-in code.

[0061] 2. Powerful Complex Data Processing Capabilities: Employing a workflow-based orchestration approach, niche and unconventional data formats are first converted to the commonly used JSON format before being submitted to large-scale models for analysis. For binary data files, they are first manually processed and exported into a readable text format, and then plugin code is generated based on prompts and sample code templates. This approach can comprehensively and accurately parse complex application data, solving the difficulties of traditional methods in handling complex data formats.

[0062] 3. Guaranteed code quality: The code generated by the large model has high standardization and consistency, reducing code quality issues caused by differences in human programming habits. The generated code is easier to maintain and upgrade, reducing later maintenance costs.

[0063] 4. Possesses intelligent and adaptive capabilities: By optimizing prompts and samples, the large model can automatically understand configuration information and accurately extract the meaning of internal fields, and can automatically determine the application type based on input data. Furthermore, when dealing with emerging applications, leveraging the intelligent agent and existing working foundation, it can quickly adapt to new changes and meet new electronic data acquisition needs by simply fine-tuning prompts or workflows.

[0064] Figure 2 This is a flowchart of a specific embodiment of the electronic data acquisition and analysis plugin generation method based on an AI large model, as shown below. Figure 2 As shown, the solution mainly includes a user module, a data processing module, an AI large-scale model system, and a plugin generation and debugging module. The user module comprises two components: electronic data acquisition requirement input and prompt word writing. This is the user-interactive part, where users clarify the specific requirements of their electronic data acquisition task and write prompt words according to the prescribed method. The data processing module includes components for data format conversion and manual binary data processing, respectively, for non-standard data formats and binary data. This module preprocesses data that cannot be directly processed by the large-scale model, making it usable by the model. The AI ​​large-scale model system consists of three components: prompt word parsing, a code generation engine, and data format judgment. Prompt word parsing understands the meaning of the user-written prompt words, data format judgment provides a decision-making basis for subsequent data processing, and the code generation engine generates corresponding parsed code based on the parsed prompt words and processed data. The plugin generation and debugging module includes a code merging tool and a plugin platform. The code merging tool combines code fragments generated by the large-scale model into complete plugin code, while the plugin platform is used to debug and verify the generated plugins, ensuring their correctness and effectiveness. The architecture diagram illustrates the collaborative relationships and data flow between the various modules, demonstrating how the system gradually generates usable electronic data acquisition plugins from user input.

[0065] In a specific implementation, the user first writes prompts containing roles, objectives, skills, and sample code based on their eDLC requirements and inputs them into the AI-based big data model system. The system processes the input data according to the prompt template and workflow orchestration. For different data formats, after conversion or pre-processing according to corresponding rules, the big data model generates corresponding parsing code snippets. These code snippets are combined into complete eDLC plugin code using a merging tool. Finally, the plugin is debugged and verified on the plugin platform. If it does not meet expectations, feedback is provided to adjust the prompts or further optimize the processing flow until a satisfactory eDLC plugin is generated.

[0066] In a specific embodiment, the electronic data acquisition agent prompting word framework includes role setting, clear objectives, skill assignment, and constraints. Wherein:

[0067] Role settings include: a custom component and a Python data parsing expert. The custom component assigns roles with corresponding domain-specific parsing capabilities based on the type of application being parsed. For example, for parsing virtual machine configuration files, the role would be set as "Virtual Machine Configuration File and Python Data Parsing Expert."

[0068] The objectives include: generating accurate Python code based on the configuration file information provided by the user and following the framework provided in the example code; ensuring that the generated code follows the framework structure of the example code provided by the user and is complete; being able to parse the fields in the configuration file according to your understanding; for custom parts, the objectives need to be defined in conjunction with the type of application being parsed, such as parsing virtual machine configuration files, which can be defined as parsing the main configuration of the virtual machine, and generating the table fields according to the corresponding content of the file.

[0069] The skills acquired include: understanding the meaning of input data fields and proficiently using Python for data parsing and encoding; the ability to deeply analyze the meaning of each field in the input data, accurately determine its data type and business purpose, and generate table structures and field codes according to the example code framework. Therefore, specific skills are assigned: proficiency in the Python programming language, familiarity with data parsing and processing techniques, understanding and application of configuration file reading and parsing methods, and custom skills tailored to specific application types, such as proficiency in parsing virtual machine configuration files, including understanding the format and data content of virtual machine configuration files.

[0070] The constraints include: To prevent the agent from generalizing its output, certain conditions need to be imposed to ensure the accuracy of the output code: Do not change the framework structure of the provided sample code; do not use custom module keywords such as config, file, db, and error; ensure that the value of each field is entered into the database as a string; code cannot be omitted and must be generated completely; the custom part mainly parses the constraints that need to be added according to the specific application type.

[0071] In a specific implementation, the learning sample is designed as follows: Due to Python's concise syntax, ease of understanding, and rapid code writing and debugging, along with its numerous high-level abstractions and rich library support, the sample code design includes the following aspects:

[0072] ① Define the table structure of the plugin, including table name and description information, column name and description information, and call the electronic data acquisition software data operation interface to create database tables according to the defined table structure;

[0073] ② Define an initialization method that triggers the electronic data acquisition software to call initialization operations when the electronic data acquisition plugin is running, such as creating a database based on the defined table structure;

[0074] ③ Traverse the file system and select files to be analyzed based on the characteristics of the files to be parsed, such as matching specified folder directories or file names, and files with certain file extensions, etc.

[0075] ④ Define the entry function for parsing the file, call the created application parsing class, and implement the specific parsing process in the application parsing class. Perform corresponding operations on the selected files, obtain information, insert it into the database, and update the number of displayed node records.

[0076] ⑤ The main interfaces for interfacing with the electronic data acquisition plugin engine are as follows:

[0077] Interface 1: Describes the definition of data items in a database table.

[0078] forensics.config.FieldCfg(name,description,data_type='text');

[0079] Interface 2: Describes plugin configuration

[0080] forensics.config.PluginCfg(name,version='1.0.0');

[0081] Interface 3: Describes the database table definition

[0082] forensics.config.TableCfg(name,description);

[0083] Interface 4: Database table insertion and retrieval operations

[0084] forensics.db.TableStore(table_name);

[0085] Interface 5: Get the full file path, file extension, and file path to the specified directory for export.

[0086] forensics.file.FmFile(file_id:int);

[0087] Interface 6: Manages plugin functionality, builds tree structures, and saves data to the database.

[0088] forensics.plugin.FmPlugin(plugin_id).

[0089] Figure 3 A flowchart illustrating a specific embodiment of the workflow orchestration method of this application is shown, as follows: Figure 3 As shown, workflow orchestration methods specifically include:

[0090] Identify data formats in complex application plugins, including plain text, database files, binary data, and niche, unconventional data formats.

[0091] For niche and unconventional data formats, write code to convert them to JSON format for large-scale model analysis. Ensure that the conversion is accurate and meets the requirements.

[0092] For binary data, since large models cannot handle it, manual preliminary analysis is used to export it into a readable text format to ensure the integrity and accuracy of the information.

[0093] Based on the processed data and prompts, input the large model to generate plugin code according to the example code template.

[0094] If the application needs to parse multiple configuration files, repeat the above steps for each file to generate code, and then use a code merging tool to merge them into complete plugin code.

[0095] Run and debug the merged plugin code on the plugin platform to see if the output results meet expectations.

[0096] Figure 4 A flowchart illustrating the implementation of an electronic data acquisition and analysis plugin according to a specific embodiment of this application is shown, such as... Figure 4 As shown, the specific steps for generating the electronic data acquisition and analysis plugin include:

[0097] The process begins with the user inputting their electronic data acquisition needs and prompts; this is the trigger point for the entire process. The user creates corresponding prompts based on the actual electronic data acquisition task and expected results, providing guidance for subsequent processing.

[0098] Data Format Determination Branch: The system will then determine the data format, classifying it into two categories: normal and non-normal. This step is crucial because different data formats require different processing methods to ensure that the large model can accurately generate parsed code.

[0099] Code generation and merging path: If the data format is standard, the large model will directly generate parsing code based on the prompts, and then this code will be integrated by the code merging tool. If the data format is non-standard, data format conversion or manual pre-processing will be performed first, and the processed results will then be used by the large model to generate code and enter the merging stage.

[0100] Verification and feedback loop: The merged code enters the plugin platform for debugging and verification. If the verification result does not meet expectations, the process will backtrack to the user input stage, prompting the user to adjust the electronic data acquisition requirements or prompt words, and restart the entire process until a plugin that meets the requirements is generated.

[0101] This invention provides a method for generating electronic data acquisition and analysis plugins based on an AI-powered large-scale model. This method enables rapid development of electronic data acquisition and analysis plugins, automates plugin code generation, and improves electronic data acquisition efficiency. Taking the parsing of a specific application as an example, prompts are written based on its data characteristics and electronic data acquisition objectives. For instance, the intelligent agent is positioned as an expert in the application's domain, with the goal of extracting specific key data using Python code. This agent is endowed with the skills to understand the application's data structure and encoding conversion, and sample code from similar applications is provided. For unconventional data formats in the application, a conversion program is written to convert them to JSON format before inputting them into the large-scale model. If binary data files exist, key information is first manually analyzed and converted into text format. The electronic data acquisition and analysis capabilities of the large-scale model then begin generating the electronic data acquisition plugin code. (Reference) Figure 5 The diagram shown illustrates an electronic data acquisition plugin for a soft router according to a specific embodiment of this application. Figure 6 The diagram shown is an electronic data acquisition result interface of a specific embodiment of this application. Currently, the related technologies of this invention have been applied to the development of electronic data acquisition plugins such as chat applications and soft routers, and have achieved good results.

[0102] Figure 7 This application illustrates an embodiment of an electronic data acquisition and analysis plugin generation system architecture diagram based on an AI large model, as shown below. Figure 7As shown, the system includes a user interaction module 701, a data processing and workflow management module 702, and a code merging and verification module 703. The user interaction module 701 is configured to acquire the user's input of electronic data acquisition requirements and a prompting framework, which includes role settings, target descriptions, skill assignments, and constraints. The data processing and workflow management module 702 is configured to process the data according to the prompting framework and workflow arrangement to generate data suitable for large-scale model processing. The code merging and verification module 703 is configured to input the processed data and prompts into the AI ​​large-scale model, which generates plugin code snippets for the electronic data acquisition requirements. A code merging tool is used to integrate the generated code snippets into complete electronic data acquisition plugin code. The electronic data acquisition plugin code is run on the plugin platform to verify its functionality. After verification, the generated electronic data acquisition plugin is deployed and applied.

[0103] In specific embodiments, the AI-based large-scale model-based electronic data acquisition and analysis plugin development method and system of this application is not only a new technology for developing electronic data acquisition and analysis plugins, but also a new model for developing electronic data acquisition plugins. It achieves 80% automation of electronic data acquisition plugin code writing, effectively addressing the rapidly evolving needs for parsing illegal applications and significantly reducing human development costs. Simultaneously, it transforms the traditional machine code-based interaction mode, allowing electronic data acquisition plugins to be created directly through natural language, greatly lowering the coding threshold and enabling non-R&D personnel to participate in electronic data acquisition plugin writing. A comparison with traditional development models is shown in the table below:

[0104]

[0105] The following is for reference. Figure 8 It shows a schematic diagram of the structure of a computer system suitable for implementing the electronic device of the present application. Figure 8 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0106] like Figure 8 As shown, the computer system includes a central processing unit (CPU) 801, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 802 or programs loaded from storage section 808 into random access memory (RAM) 803. RAM 803 also stores various programs and data required for the operation of system 800. CPU 801, ROM 802, and RAM 803 are interconnected via bus 804. Input / output (I / O) interface 805 is also connected to bus 804.

[0107] The following components are connected to I / O interface 805: an input section 806 including a keyboard, mouse, etc.; an output section 807 including a liquid crystal display (LCD) and speakers, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN card and a modem, etc. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to I / O interface 805 as needed. A removable medium 811, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 810 as needed so that computer programs read from it can be installed into storage section 808 as needed.

[0108] Specifically, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 809, and / or installed from removable medium 811. When the computer program is executed by central processing unit (CPU) 801, it performs the functions defined in the methods of this application. It should be noted that the computer-readable storage medium of this application can be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable storage medium other than a computer-readable storage medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. Program code contained on a computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0109] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0110] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0111] The modules described in the embodiments of this application can be implemented in software or in hardware.

[0112] In another aspect, this application also provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: acquire user-inputted electronic data acquisition requirements and a prompt word framework, the prompt word framework including role settings, target descriptions, skill assignments, and constraints; process the data according to the prompt word framework and workflow arrangement to generate data suitable for large-scale model processing; input the processed data and prompt words into the AI ​​large-scale model, which generates plugin code snippets for the electronic data acquisition requirements; integrate the generated code snippets into complete electronic data acquisition plugin code using a code merging tool; run the electronic data acquisition plugin code on the plugin platform to verify its functional correctness; if it does not meet the requirements, adjust the prompt words and repeat the above steps; after verification, deploy the generated electronic data acquisition plugin for application.

[0113] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A method for generating electronic data acquisition and analysis plugins based on AI large-scale models, characterized in that, include: S1: Obtain the user's input electronic data acquisition requirements and prompt word framework, wherein the prompt word framework includes role setting, target description, skill assignment and restriction conditions; S2: Process the data according to the prompt word framework and workflow arrangement to generate data suitable for large model processing; S3: Input the processed data and the prompt words into the AI ​​big model, and the AI ​​big model generates plug-in code snippets for electronic data acquisition needs; S4: Use a code merging tool to integrate the generated code snippets into complete eData Acquisition plugin code; S5: Run the electronic data acquisition plugin code on the plugin platform to verify the correctness of its function. If it does not meet the requirements, adjust the prompt words and repeat the above steps. S6: After verification, deploy the generated electronic data acquisition plugin into the application.

2. The method for generating an electronic data acquisition and analysis plugin based on an AI large model according to claim 1, characterized in that, The prompt word framework in S1 specifically includes: the role setting is a combination of a professional parsing expert and a Python data parsing expert, with the specific professional field determined according to the nature of the application and data characteristics; the goal description is to generate Python code that can accurately parse various fields in the application configuration file according to the provided sample code framework structure; the skills include proficiency in the advanced features of the Python programming language and the use of common libraries, mastery of various data parsing and processing techniques and algorithms, and a deep understanding and accurate application of configuration file reading and parsing methods.

3. The method for generating an electronic data acquisition and analysis plugin based on an AI large model according to claim 1, characterized in that, The restrictions include prohibiting changes to the sample code framework, avoiding the use of specific keywords that are likely to cause confusion or errors, ensuring that all field values ​​are stored in the prescribed string format, and requiring the generation of complete and undamaged code. For the generation of financial data electronic data acquisition plugins, the restrictions also include compliance with financial data security and compliance standards.

4. The method for generating an electronic data acquisition and analysis plugin based on an AI large model according to claim 1, characterized in that, The workflow orchestration methods specifically include: Determine the format type of input data, and distinguish between plain text, database files, binary data, and niche and unconventional data formats; For the aforementioned niche and unconventional data formats, a conversion program is initiated to convert them into JSON format, and the conversion process follows the principle of consistency in data structure and semantics; For the binary data, key information is extracted based on the type and application scenario of the binary data, and then converted into a standard readable text format.

5. The method for generating an electronic data acquisition and analysis plugin based on an AI large model according to claim 4, characterized in that, The AI ​​big model includes a natural language understanding layer, a code generation core layer, and a data adaptation and optimization layer. The natural language understanding layer is configured to receive the prompt word framework containing role settings, target descriptions, skill assignments, and constraints input by the user, as well as the processed data. It uses deep learning algorithms and pre-trained language models to perform semantic analysis and intent recognition on the prompt words, extract key information and task requirements, and convert them into a structured representation suitable for internal processing of the model. The core code generation layer is configured based on the output of the natural language understanding layer. It utilizes a knowledge base containing a large number of electronic data acquisition-related code templates and algorithmic logic, combined with a generative model architecture based on deep learning. Through learning and adaptation to different application scenarios and data types, it generates plug-in code fragments that meet the needs of electronic data acquisition. The data adaptation and optimization layer dynamically adapts to the characteristics and format of the input data during the code generation process. It extracts and optimizes features of the niche and unconventional data formats and the preprocessed results of the binary data, providing the core code generation layer with high-quality input that meets data specifications and task requirements. At the same time, it verifies and adjusts the compatibility between the code and the data after the code is generated.

6. The method for generating an electronic data acquisition and analysis plugin based on an AI large model according to claim 1, characterized in that, S5 specifically includes: deploying the merged electronic data acquisition plugin code to the plugin platform for running and debugging, simulating the actual electronic data acquisition working environment and data input, and observing whether the output results meet the expected electronic data acquisition goals and data parsing requirements; if the debugging results do not meet expectations, establishing a feedback mechanism, automatically backtracking to the prompt word framework or workflow orchestration steps, prompting the user to adjust the electronic data acquisition requirements, optimize the prompt word content, or improve the workflow processing flow according to the error information and debugging results, and restarting the entire method process until a plugin that meets the electronic data acquisition requirements is generated.

7. A computer-readable storage medium having one or more computer programs stored thereon, characterized in that, When the one or more computer programs are executed by a computer processor, they perform the method according to any one of claims 1-6.

8. A plugin generation system for electronic data acquisition and analysis based on an AI large model, characterized in that, include: The user interaction module is configured to obtain electronic data acquisition requirements and prompt word frameworks for user input. The prompt word framework includes role setting, target description, skill assignment, and constraints. The data processing and workflow management module is configured to process data according to the prompt word framework and workflow arrangement method to generate data suitable for large model processing; The code merging and verification module is configured to input the processed data and the prompt words into the AI ​​big model, and the AI ​​big model generates plug-in code snippets for electronic data acquisition needs; The generated code snippets are integrated into a complete electronic data acquisition plugin code using a code merging tool; the electronic data acquisition plugin code is run on the plugin platform to verify the correctness of its function; after verification, the generated electronic data acquisition plugin is deployed and applied.

9. The electronic data acquisition and analysis plug-in generation system based on an AI large model according to claim 8, characterized in that, The prompt word framework specifically includes: the roles are set as a combination of professional analysis experts and Python data analysis experts for the type of application to be analyzed, with the specific professional fields determined according to the nature of the application and data characteristics; the goal is described as generating Python code that can accurately parse various fields in the application configuration file according to the provided sample code framework structure; the skills include proficiency in the advanced features of the Python programming language and the use of common libraries, mastery of various data analysis and processing techniques and algorithms, and a deep understanding and accurate application of configuration file reading and parsing methods; the constraints include prohibiting changes to the sample code framework, avoiding the use of specific keywords that are likely to cause confusion or errors, ensuring that all field values ​​are stored in the specified string format, and requiring the generation of complete and missing code. For the generation of financial data electronic data acquisition plugins, the constraints also include compliance with financial data security and compliance standards.

10. The electronic data acquisition and analysis plug-in generation system based on an AI large model according to claim 8, characterized in that, The workflow orchestration method specifically includes: determining the format type of the input data, distinguishing between ordinary text, database files, binary data, and niche unconventional data formats; for the niche unconventional data formats, starting a conversion program to convert them into JSON format, with the conversion process following the principle of consistency between data structure and semantics; for the binary data, extracting key information based on the type and application scenario of the binary data, and converting it into a standard readable text format.

11. The electronic data acquisition and analysis plug-in generation system based on an AI large model according to claim 10, characterized in that, The AI ​​big model includes a natural language understanding layer, a code generation core layer, and a data adaptation and optimization layer. The natural language understanding layer is configured to receive the prompt word framework containing role settings, target descriptions, skill assignments, and constraints input by the user, as well as the processed data. It uses deep learning algorithms and pre-trained language models to perform semantic analysis and intent recognition on the prompt words, extract key information and task requirements, and convert them into a structured representation suitable for internal processing of the model. The core code generation layer is configured based on the output of the natural language understanding layer. It utilizes a knowledge base containing a large number of electronic data acquisition-related code templates and algorithmic logic, combined with a generative model architecture based on deep learning. Through learning and adaptation to different application scenarios and data types, it generates plug-in code fragments that meet the needs of electronic data acquisition. The data adaptation and optimization layer dynamically adapts to the characteristics and format of the input data during the code generation process. It extracts and optimizes features of the niche and unconventional data formats and the preprocessed results of the binary data, providing the core code generation layer with high-quality input that meets data specifications and task requirements. At the same time, it verifies and adjusts the compatibility between the code and the data after the code is generated.

12. The electronic data acquisition and analysis plug-in generation system based on an AI large model according to claim 8, characterized in that, The code merging and verification module specifically includes: deploying the merged electronic data acquisition plugin code to the plugin platform for running and debugging, simulating the actual electronic data acquisition working environment and data input, and observing whether the output results meet the expected electronic data acquisition goals and data parsing requirements; if the debugging results do not meet expectations, a feedback mechanism is established to automatically backtrack to the prompt word framework or workflow orchestration steps, prompting the user to adjust the electronic data acquisition requirements, optimize the prompt word content, or improve the workflow processing flow based on the error information and debugging results, and restarting the entire method flow until a plugin that meets the electronic data acquisition requirements is generated.

Images