Intelligent tax handling method based on tax industry large model

Through the intelligent tax payment method based on the big model of the tax industry, the problem of inefficiency of traditional tax payment methods is solved, real-time tax consultation, automatic form filling and risk prediction are realized, and tax payment efficiency and tax-related compliance are improved.

CN118941400APending Publication Date: 2024-11-12INSPUR SOFTWARE TECH CO LTD
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411009881.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Traditional tax payment methods are inefficient, and taxpayers find it difficult to quickly deal with and avoid tax-related issues when dealing with tax-related matters.

Method used

Using smart tax payment methods based on big models of the tax industry, deep learning technology is built through steps such as data collection and preprocessing, model training and optimization, intelligent consulting services, smart tax-related services, risk prediction and prompts, and exclusive tax services. Natural language processing system, providing real-time tax consultation, automatic form filling and risk prediction.

Benefits of technology

It has improved the tax department's intelligent tax service capabilities, improved the tax payment efficiency of taxpayers, avoided tax-related risks, and enhanced tax-related compliance and sense of gain.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941400A_ABST
    Figure CN118941400A_ABST
Patent Text Reader

Abstract

The invention particularly relates to an intelligent tax handling method based on a tax industry large model. According to the intelligent tax handling method based on the tax industry large model, tax data are collected, and processing cleaning, classification labeling and formatting processing are carried out; constructing a tax industry large model by using a deep learning technology, and continuously training, optimizing and adjusting the parameters and the structure of the model; real-time tax consultation and answering are provided for taxpayers based on the trained tax industry large model; the tax industry large model automatically generates a tax-related acceptance form, and automatically fills in form information for the taxpayer to confirm and tutorize the taxpayer to supplement the tax-related form; and corresponding risk prompts and suggestions are provided for taxpayers, and personalized tax services are provided. According to the intelligent tax handling method based on the tax industry large model, the intelligent tax service capability level of the tax department is improved, the tax handling efficiency of the taxpayer is improved, and meanwhile, the tax-related risk of the taxpayer can be avoided, so that the tax-related compliance and the obtaining feeling of the taxpayer are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of big model technology, and in particular to a smart tax handling method based on a big model of the tax industry. Background Art

[0002] At present, with the rapid development of computer hardware technology and the improvement of computing infrastructure, the continuous advancement of machine learning technologies such as deep learning, and the application and promotion of big models in various industries, the tax authorities are also accelerating the construction of smart taxation and focusing on improving tax service capabilities.

[0003] In order to solve the problems of low efficiency and complex business in traditional tax filing methods, and taxpayers' inability to quickly handle and avoid tax-related issues when handling tax-related matters, the present invention proposes a smart tax filing method based on a big model of the tax industry. Summary of the invention

[0004] In order to remedy the deficiencies of the prior art, the present invention provides a simple and efficient intelligent tax filing method based on a large tax industry model.

[0005] The present invention is achieved through the following technical solutions:

[0006] A smart tax handling method based on a tax industry big model includes the following steps:

[0007] Step S1, data collection and preprocessing;

[0008] Collect tax data, including but not limited to tax rules, taxpayer information, tax-related records and taxpayer tax-related behavior data, and process, clean, classify and format the collected tax data to form a high-quality data set;

[0009] In step S1, a large amount of tax data is obtained by using database connection, API service and web crawler technology.

[0010] Check the collected original tax data, eliminate redundant data, supplement missing data, standardize data format, remove noise, HTML tags, special characters and outliers, desensitize related sensitive information, complete format conversion and normalization of original data content, unify data format content, and ensure data quality and accuracy.

[0011] In step S1, the cleaned tax data is analyzed, and meaningful feature identifiers are extracted by customizing the supervised learning task, and labels are created with tax professional terms to annotate tax-related knowledge so that the label data can reflect the characteristics, patterns and laws of the tax data;

[0012] The labeled tax data is standardized and normalized and organized into a high-quality data set. At the same time, the data set is split into training data set, verification data set and test data set for the training, verification and testing of large models in the tax industry.

[0013] Step S2: model training and optimization;

[0014] Use deep learning technology to build a large model for the tax industry, and continuously train the large model for the tax industry, optimize and adjust the parameters and structure of the model so that it can accurately understand the patterns and laws in the tax data; at the same time, use optimization algorithms to improve the training efficiency and performance of the model;

[0015] In step S2, a language model is built using Transformer and natural language technology to perform natural language processing, speech recognition is performed on the tax-related claims raised by taxpayers, semantic analysis and context information are performed in combination with natural language understanding technology, semantic understanding and context understanding are performed in combination with business scenarios, and user intentions are judged; data analysis and calculation are performed through a large model of the tax industry, and the optimal calculation results are fed back in the form of voice, text and tables.

[0016] In step S2, a large amount of training data sets are used to train the tax industry large model, and the model parameters are optimized and adjusted according to the task requirements to reduce the minimum prediction error;

[0017] Use validation data sets to monitor the performance of the model during training and evaluate the performance of the model for further optimization and adjustment, so that the tax industry big model can correctly understand the patterns and rules in tax data;

[0018] Finally, the tax industry model is further evaluated and verified through the test data set until it meets business needs.

[0019] In step S2, the performance of the tax industry large model is checked regularly, and new tax data is collected for retraining and optimization of the model, and the model is continuously updated and adjusted to adapt to new business needs.

[0020] Step S3: Intelligent consulting service;

[0021] Provide taxpayers with real-time tax consultation and answers based on the trained tax industry big model;

[0022] When taxpayers raise tax-related issues in the form of voice or text, the language model identifies the taxpayer's intention through voice recognition, semantic analysis, contextual information, business scenarios and semantic understanding. At the same time, through the tax industry big model data analysis and calculation, the optimal results are generated and fed back to the taxpayer to help taxpayers solve tax problems;

[0023] Step S4: Smart tax-related services;

[0024] The tax industry big model automatically identifies the taxpayer's registration information, identification information and tax-related requirements, automatically generates a tax-related acceptance form based on the taxpayer's tax-related demands, and automatically fills in the basic information, identification information and invoicing information in the form based on the taxpayer's tax-related information for the taxpayer to confirm; at the same time, it guides the taxpayer to supplement the tax-related forms; Step S5, risk prediction and prompts;

[0025] Through the analysis of tax data, the tax industry big model combines the tax-related data reported by taxpayers to predict the existing tax risks and provide taxpayers with corresponding risk warnings and suggestions;

[0026] Step S6: Specialized tax services;

[0027] Supports personalized customization and provides personalized tax services based on taxpayers' needs and preferences.

[0028] The system for realizing the intelligent tax handling method based on the tax industry big model of the present invention comprises a data collection and preprocessing module, a model training and optimization module, an intelligent consulting module, an intelligent tax-related module, a risk prediction and prompting module and a dedicated tax module;

[0029] The data collection and preprocessing module is responsible for collecting tax data, including but not limited to tax rules, taxpayer information, tax-related records and taxpayer tax-related behavior data, and processing, cleaning, classifying, labeling and formatting the collected tax data to form a high-quality data set;

[0030] The model training and optimization module is responsible for building a large tax industry model using deep learning technology, and continuously training the large tax industry model, optimizing and adjusting the parameters and structure of the model so that it can accurately understand the patterns and laws in the tax data; at the same time, it uses optimization algorithms to improve the training efficiency and performance of the model;

[0031] The intelligent consulting module is responsible for providing taxpayers with real-time tax consultation and answers based on the trained tax industry big model;

[0032] When taxpayers raise tax-related issues in the form of voice or text, the language model identifies the taxpayer's intention through voice recognition, semantic analysis, contextual information, business scenarios and semantic understanding. At the same time, through the tax industry big model data analysis and calculation, the optimal results are generated and fed back to the taxpayer to help taxpayers solve tax problems;

[0033] The smart tax-related module is responsible for automatically identifying the taxpayer's registration information, identification information and tax-related requirements based on the trained tax industry model, automatically generating tax-related acceptance forms based on the taxpayer's tax-related demands, and automatically filling in the basic information, identification information and invoicing information in the form based on the taxpayer's tax-related information for the taxpayer to confirm; at the same time, it guides the taxpayer to supplement the tax-related forms;

[0034] The risk prediction and warning module is responsible for predicting the existing tax risks based on the tax industry big model and the tax-related data reported by the taxpayers, and providing the taxpayers with corresponding risk warnings and suggestions;

[0035] The dedicated tax module supports personalized customization and provides personalized tax services based on the needs and preferences of taxpayers.

[0036] A smart tax handling device based on a large tax industry model, characterized in that it includes a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the above-mentioned method steps when executing the computer program.

[0037] A readable storage medium, characterized in that: a computer program is stored on the readable storage medium, and the computer program implements the above method steps when executed by a processor.

[0038] The beneficial effect of the present invention is that the intelligent tax filing method based on the big model of the tax industry has improved the intelligent tax service capability of the tax department, improved the tax filing efficiency of taxpayers, and at the same time avoided taxpayers' tax-related risks, thereby improving taxpayers' tax-related compliance and sense of gain. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Attached Figure 1 Schematic diagram of the data collection and preprocessing method of the present invention.

[0041] Attached Figure 2 This is a schematic diagram of the autonomous tax service system based on the tax industry big model of the present invention. DETAILED DESCRIPTION

[0042] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0043] The smart tax handling method based on the tax industry big model includes the following steps:

[0044] Step S1, data collection and preprocessing;

[0045] The intelligent tax service system based on the tax industry big model construction requires a huge data set for training and optimization. Therefore, it is necessary to first collect tax data, including but not limited to tax rules, taxpayer information, tax-related records and taxpayer tax-related behavior data, and process, clean, classify and format the collected tax data to form a high-quality data set;

[0046] In step S1, a large amount of tax data is obtained by using database connection, API service and web crawler technology.

[0047] Check the collected original tax data, eliminate redundant data, supplement missing data, standardize data format, remove noise, HTML tags, special characters and outliers, desensitize related sensitive information, complete format conversion and normalization of original data content, unify data format content, and ensure data quality and accuracy.

[0048] In step S1, the cleaned tax data is analyzed, and meaningful feature identifiers are extracted by customizing the supervised learning task, and labels are created with tax professional terms to annotate tax-related knowledge so that the label data can reflect the characteristics, patterns and laws of the tax data;

[0049] The labeled tax data is standardized and normalized and organized into a high-quality data set. At the same time, the data set is split into training data set, verification data set and test data set for the training, verification and testing of large models in the tax industry.

[0050] Step S2: model training and optimization;

[0051] Use deep learning technology to build a large model for the tax industry, and continuously train the large model for the tax industry, optimize and adjust the parameters and structure of the model so that it can accurately understand the patterns and laws in the tax data; at the same time, use optimization algorithms to improve the training efficiency and performance of the model;

[0052] In step S2, a language model is built using Transformer and natural language technology to perform natural language processing, speech recognition is performed on the tax-related claims raised by taxpayers, semantic analysis and context information are performed in combination with natural language understanding technology, semantic understanding and context understanding are performed in combination with business scenarios, and user intentions are judged; data analysis and calculation are performed through a large model of the tax industry, and the optimal calculation results are fed back in the form of voice, text and tables.

[0053] In step S2, a large amount of training data sets are used to train the tax industry large model, and the model parameters are optimized and adjusted according to the task requirements to reduce the minimum prediction error;

[0054] Use validation data sets to monitor the performance of the model during training and evaluate the performance of the model for further optimization and adjustment, so that the tax industry big model can correctly understand the patterns and rules in tax data;

[0055] Finally, the tax industry model is further evaluated and verified through the test data set until it meets business needs.

[0056] In step S2, as tax policies and businesses change, the performance of the tax industry large model is regularly checked, and new tax data is collected for retraining and optimization of the model, and the model is continuously updated and adjusted to adapt to new business needs.

[0057] Step S3: Intelligent consulting service;

[0058] Provide taxpayers with real-time tax consultation and answers based on the trained tax industry big model;

[0059] When taxpayers raise tax-related issues in the form of voice or text, the language model identifies the taxpayer's intention through voice recognition, semantic analysis, contextual information, business scenarios and semantic understanding. At the same time, through the tax industry big model data analysis and calculation, the optimal results are generated and fed back to the taxpayer to help taxpayers solve tax problems;

[0060] Step S4: Smart tax-related services;

[0061] The tax industry big model automatically identifies the taxpayer's registration information, identification information and tax-related requirements, automatically generates tax-related acceptance forms based on the taxpayer's tax-related demands, and automatically fills in the basic information, identification information and invoicing information in the form based on the taxpayer's tax-related information for the taxpayer to confirm; at the same time, it guides taxpayers to supplement tax-related forms; through intelligent analysis and processing, it improves the accuracy and efficiency of tax-related business handling, while reducing the burden on taxpayers;

[0062] Step S5: risk prediction and prompting;

[0063] By analyzing tax data, the tax industry big model combines the tax-related data reported by taxpayers to predict existing tax risks and provide taxpayers with corresponding risk warnings and suggestions; this helps taxpayers to promptly identify and resolve potential problems and avoid tax violations and fines.

[0064] Step S6: Specialized tax services;

[0065] Supports personalized customization and provides personalized tax services tailored to the needs and preferences of taxpayers. For example, it recommends suitable tax preferential policies to taxpayers and provides customized tax planning suggestions.

[0066] The system for realizing the intelligent tax handling method based on the tax industry big model includes a data collection and preprocessing module, a model training and optimization module, an intelligent consulting module, an intelligent tax-related module, a risk prediction and prompting module, and a dedicated tax module;

[0067] The data collection and preprocessing module is responsible for collecting tax data, including but not limited to tax rules, taxpayer information, tax-related records and taxpayer tax-related behavior data, and processing, cleaning, classifying, labeling and formatting the collected tax data to form a high-quality data set;

[0068] The model training and optimization module is responsible for building a large tax industry model using deep learning technology, and continuously training the large tax industry model, optimizing and adjusting the parameters and structure of the model so that it can accurately understand the patterns and laws in the tax data; at the same time, it uses optimization algorithms to improve the training efficiency and performance of the model;

[0069] The intelligent consulting module is responsible for providing taxpayers with real-time tax consultation and answers based on the trained tax industry big model;

[0070] When taxpayers raise tax-related issues in the form of voice or text, the language model identifies the taxpayer's intention through voice recognition, semantic analysis, contextual information, business scenarios and semantic understanding. At the same time, through the tax industry big model data analysis and calculation, the optimal results are generated and fed back to the taxpayer to help taxpayers solve tax problems;

[0071] The smart tax-related module is responsible for automatically identifying the taxpayer's registration information, identification information and tax-related requirements based on the trained tax industry model, automatically generating tax-related acceptance forms based on the taxpayer's tax-related demands, and automatically filling in the basic information, identification information and invoicing information in the form based on the taxpayer's tax-related information for the taxpayer to confirm; at the same time, it guides the taxpayer to supplement the tax-related forms;

[0072] The risk prediction and warning module is responsible for predicting the existing tax risks based on the tax industry big model and the tax-related data reported by the taxpayers, and providing the taxpayers with corresponding risk warnings and suggestions;

[0073] The dedicated tax module supports personalized customization and provides personalized tax services based on the needs and preferences of taxpayers.

[0074] The intelligent tax handling device based on the tax industry big model includes a memory and a processor; the memory is used to store computer programs, and the processor is used to implement the above-mentioned method steps when executing the computer program.

[0075] The readable storage medium stores a computer program, and when the computer program is executed by a processor, the above method steps are implemented.

[0076] In summary, the smart tax administration method based on the tax industry big model collects a large amount of tax data and forms a high-quality training data set after data cleaning, labeling and conversion; uses deep learning and other machine learning technologies to build a tax industry big model, and continuously trains and optimizes the parameters and structure of the model to enable it to more accurately understand the patterns and laws in the tax data; based on the trained tax big model, it can provide taxpayers with smart tax-related services:

[0077] Support taxpayers to provide real-time tax consultation and answer services; support automatic identification of taxpayers' tax information and tax-related requirements and automatically generate tax-related scenarios and tax-related forms to assist taxpayers in completing tax-related business processing; support the prediction of taxpayers' tax-related risks and provide response plans; support taxpayers to fully enjoy tax preferential policies and generate customized tax planning suggestions for taxpayers; improve taxpayers' tax handling efficiency, avoid tax-related risks, and allow taxpayers to enjoy the benefits brought by tax-related service reforms, thereby improving taxpayers' tax-related compliance and sense of gain.

[0078] This intelligent tax filing method based on the tax industry big model can improve tax filing efficiency, avoid potential risks, and enjoy tax incentives.

[0079] For example, when taxpayers handle VAT declaration business, they handle business through the system of intelligent service system based on the big model of tax industry. They only need to say "I want to handle VAT declaration" by voice. The system will automatically display the VAT declaration form based on the characteristics of the taxpayer and business needs. At the same time, the form will be pre-filled with the basic information of the taxpayer and the invoicing data for the taxpayer to confirm. After the taxpayer confirms that it is correct, submit the declaration. The system will automatically verify the relevant data and risk check. After the verification is passed, the business handling is completed. If the taxpayer changes the pre-filled data by himself and triggers an abnormal warning, the system will automatically prompt the taxpayer to remind the problem and assist the taxpayer to deal with the existing tax-related risk issues in a timely manner.

[0080] The above is a detailed introduction to a smart tax payment method based on a tax industry big model in an example of the present invention. This section uses specific examples to illustrate the principles and implementation methods of the invention. The above examples are only used to help understand the core ideas of the present invention. Without departing from the principles of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

Claims

1. A smart tax handling method based on a tax industry big model, characterized by: The following steps are involved: Step S1, data collection and preprocessing; Collect tax data, including but not limited to tax rules, taxpayer information, tax-related records and taxpayer tax-related behavior data, and process, clean, classify and format the collected tax data to form a high-quality data set; Step S2: model training and optimization; Use deep learning technology to build a large model for the tax industry, and continuously train the large model for the tax industry, optimize and adjust the parameters and structure of the model so that it can accurately understand the patterns and laws in the tax data; at the same time, use optimization algorithms to improve the training efficiency and performance of the model; Step S3: Intelligent consulting service; Provide taxpayers with real-time tax consultation and answers based on the trained tax industry big model; Taxpayers raise tax-related issues in the form of voice or text. The language model identifies the taxpayer's intention through voice recognition, semantic analysis, contextual information, business scenarios and semantic understanding. At the same time, through the tax industry big model data analysis and calculation, the generated results are fed back to the taxpayer to help taxpayers solve tax problems; Step S4: Smart tax-related services; The tax industry big model automatically identifies the taxpayer's registration information, identification information and tax-related requirements, automatically generates a tax-related acceptance form based on the taxpayer's tax-related demands, and automatically fills in the basic information, identification information and invoicing information in the form based on the taxpayer's tax-related information for the taxpayer to confirm; at the same time, it guides the taxpayer to supplement the tax-related forms; Step S5, risk prediction and prompts; Through the analysis of tax data, the tax industry big model combines the tax-related data reported by taxpayers to predict the existing tax risks and provide taxpayers with corresponding risk warnings and suggestions; Step S6: Specialized tax services; Supports personalized customization and provides personalized tax services based on taxpayers' needs and preferences.

2. The method for intelligent tax administration based on the tax industry big model according to claim 1 is characterized by: In step S1, tax data is obtained by using database connection, API service and web crawler technology. Check the collected original tax data, eliminate redundant data, supplement missing data, standardize data format, remove noise, HTML tags, special characters and outliers, desensitize related sensitive information, complete format conversion and normalization of original data content, unify data format content, and ensure data quality and accuracy.

3. The method for intelligent tax administration based on the tax industry big model according to claim 1 is characterized by: In step S1, the cleaned tax data is analyzed, and meaningful feature identifiers are extracted by customizing the supervised learning task, and labels are created with tax professional terms to annotate tax-related knowledge so that the label data can reflect the characteristics, patterns and laws of the tax data; The labeled tax data is standardized and normalized and organized into a high-quality data set. At the same time, the data set is split into training data set, verification data set and test data set for the training, verification and testing of large models in the tax industry.

4. The method for intelligent tax administration based on the tax industry big model according to claim 1 is characterized by: In step S2, a language model is built using Transformer and natural language technology to perform natural language processing, speech recognition is performed on the tax-related claims raised by taxpayers, semantic analysis and context information are performed in combination with natural language understanding technology, semantic understanding and context understanding are performed in combination with business scenarios, and user intentions are judged; Data analysis and calculation are performed through the big model of the tax industry, and the calculation results are fed back in the form of voice, text and tables.

5. The method for intelligent tax administration based on the tax industry big model according to claim 1 is characterized by: In step S2, the tax industry large model is trained using the training data set, and the model parameters are optimized and adjusted according to the task requirements to reduce the minimum prediction error; Use validation data sets to monitor the performance of the model during training and evaluate the performance of the model for further optimization and adjustment, so that the tax industry big model can correctly understand the patterns and rules in tax data; Finally, the tax industry model is further evaluated and verified through the test data set until it meets business needs.

6. The method for intelligent tax administration based on the tax industry big model according to claim 5 is characterized by: In step S2, the performance of the tax industry large model is checked regularly, and new tax data is collected for retraining and optimization of the model, and the model is continuously updated and adjusted to adapt to new business needs.

7. A system for implementing the intelligent tax handling method based on the tax industry big model as described in any one of claims 1 to 6, characterized in that: Including data collection and preprocessing module, model training and optimization module, intelligent consulting module, Smart tax-related module, risk prediction and reminder module and exclusive taxation module; The data collection and preprocessing module is responsible for collecting tax data, including but not limited to tax rules, taxpayer information, tax-related records and taxpayer tax-related behavior data, and processing, cleaning, classifying, labeling and formatting the collected tax data to form a high-quality data set; The model training and optimization module is responsible for building a large tax industry model using deep learning technology, and continuously training the large tax industry model, optimizing and adjusting the parameters and structure of the model so that it can accurately understand the patterns and laws in the tax data; at the same time, it uses optimization algorithms to improve the training efficiency and performance of the model; The intelligent consulting module is responsible for providing taxpayers with real-time tax consultation and answers based on the trained tax industry big model; Taxpayers raise tax-related issues in the form of voice or text. The language model identifies the taxpayer's intention through voice recognition, semantic analysis, contextual information, business scenarios and semantic understanding. At the same time, through the tax industry big model data analysis and calculation, the generated results are fed back to the taxpayer to help taxpayers solve tax problems; The smart tax-related module is responsible for automatically identifying the taxpayer's registration information, identification information and tax-related requirements based on the trained tax industry model, automatically generating tax-related acceptance forms based on the taxpayer's tax-related demands, and automatically filling in the basic information, identification information and invoicing information in the form based on the taxpayer's tax-related information for the taxpayer to confirm; at the same time, it guides the taxpayer to supplement the tax-related forms; The risk prediction and warning module is responsible for predicting the existing tax risks based on the tax industry big model and the tax-related data reported by the taxpayers, and providing the taxpayers with corresponding risk warnings and suggestions; The dedicated tax module supports personalized customization and provides personalized tax services based on taxpayers' needs and preferences.

8. A smart tax handling device based on a tax industry big model, characterized by: The method comprises a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the method according to any one of claims 1 to 6 when executing the computer program.

9. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Citation Information

Cited By

  • Large-model-driven tax customer service intelligent training and visual pairing system

    CN119784394A

  • Tax AI agent construction method based on large language model

    CN119961396A

  • Finance and tax training data processing method and system for large language model

    CN120030342A