Method and device for realizing automatic approval of informatization project

Abstract

The invention relates to the technical field of artificial intelligence large models, and particularly provides a method and device for realizing automatic examination and approval of informatization projects, and the method comprises the following steps: S1, integrating computing power resources and constructing a computing platform; s2, integrating the government affair knowledge base and historical project data; s3, training and optimizing a Hainot big model; s4, constructing an examination and approval model and realizing functions; and S5, testing and optimizing. Compared with the prior art, the whole efficiency of informatization construction can be improved, more accurate and efficient services can be provided for departments, and work is promoted to develop in the more intelligent and convenient direction.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence large model technology, and specifically provides a method and device for realizing automatic approval of information technology projects. Background Art

[0002] With the rapid development and continuous progress of computer technology and the widespread provision of high-performance computing resources, large model technology has emerged and flourished. The rise of this technology not only marks a major leap in the field of artificial intelligence, but also heralds a new era for data processing and analysis capabilities.

[0003] With the rapid development of computer technology, breakthroughs at the hardware level are particularly critical. High-performance computing clusters, advanced data centers, and GPU computing cards customized for deep learning provide solid hardware support for the training of large-scale models. The continuous optimization and upgrading of these hardware resources have made it more efficient and faster than ever to process massive amounts of data and perform complex computing tasks.

[0004] At the same time, advances in software have also injected strong momentum into the development of large model technology. The continuous improvement of deep learning frameworks, such as TensorFlow and PyTorch, has made model development, training, and deployment more convenient and flexible. These frameworks not only provide a rich algorithm library and tool set, but also support distributed training, greatly improving the efficiency and scale of model training.

[0005] Driven by both hardware and software, big model technology has begun to show its strong potential and advantages. By using large-scale data sets for training, big models can learn more complex and sophisticated feature representations, thereby performing well in many application scenarios. Whether in the fields of natural language processing, computer vision or speech recognition, big models have demonstrated excellent performance and broad application prospects.

[0006] In the process of establishing and building government projects, the review process plays a vital role. It is not only related to the scientificity, feasibility and compliance of the project, but also an important guarantee to ensure that the project can accurately meet government needs and effectively promote the effectiveness of information construction. In order to ensure the comprehensiveness and depth of the review, it is usually necessary to gather a large number of people from different fields and with different professional backgrounds to participate in a meticulous, comprehensive and thoughtful review of the project content.

[0007] Secondly, the professional quality and practical experience of the reviewers are crucial to the accuracy and effectiveness of project reviews. They not only need to have solid professional knowledge and be able to deeply understand the technical details and implementation plans of the project; they also need to have rich practical experience and be able to provide practical suggestions and improvement measures for the project based on past success stories and failure lessons. At the same time, keen policy insight is also one of the indispensable abilities of reviewers, which can help them accurately grasp policy orientations and development trends and ensure that the project can keep up with the trend of the times.

[0008] However, a comprehensive review of a project requires a lot of time and energy from the reviewers. This is not only because the project content itself may involve multiple fields and multiple levels, requiring the reviewers to conduct in-depth and detailed analysis and research; but also because government projects are often highly complex and uncertain, requiring the reviewers to constantly track the progress of the project and adjust the review strategy and methods in a timely manner.

[0009] Therefore, faced with the high time cost, high manpower consumption and the need to improve approval efficiency in the process of government project approval, how to reasonably arrange the time and energy of reviewers while ensuring the quality of project review has become an urgent problem to be solved. Summary of the invention

[0010] The present invention aims at the above-mentioned deficiencies of the prior art and provides a method for realizing automatic approval of information technology projects with strong practicability.

[0011] A further technical task of the present invention is to provide a device that is reasonably designed, safe and applicable for realizing automatic approval of information technology projects.

[0012] The technical solution adopted by the present invention to solve its technical problem is:

[0013] A method for realizing automatic approval of information technology projects comprises the following steps:

[0014] S1. Integration of computing resources and construction of computing platform;

[0015] S2, integration of government knowledge base and historical project data;

[0016] S3, Hainuo large model training and optimization;

[0017] S4, approval model construction and function implementation;

[0018] S5. Testing and optimization.

[0019] Furthermore, in step S1, it includes:

[0020] S1-1, computing power resource integration;

[0021] First, we integrate the diverse computing resources of servers and GPU computing cards, and effectively schedule and allocate these computing resources by building a resource management system.

[0022] S1-2, computing platform construction;

[0023] Based on the integrated computing resources, a computing platform is built, and the necessary software environment is deployed on the computing platform, including the deep learning framework and database management system.

[0024] Furthermore, in step S2, it includes:

[0025] S2-1, Construction of government affairs knowledge base;

[0026] Collect a large amount of knowledge and information related to government information projects, systematically organize and classify this knowledge, and form a structured government knowledge base;

[0027] S2-2, historical project data integration;

[0028] Collect a large amount of historical project data from relevant departments and institutions, including project application materials, approval process records and review results, clean, deduplicate and standardize the data, and format and convert the data according to the needs of model training.

[0029] Furthermore, in step S3, it includes:

[0030] S3-1. Based on deep learning technology, we built the preliminary architecture of the Hairuo model, used the integrated government knowledge base and historical project data to conduct customized training on the model, and improved the accuracy and generalization ability of the model by adjusting model parameters and optimizing loss functions;

[0031] S3-2, fine tuning;

[0032] Use cross-validation and A / B testing to conduct a comprehensive evaluation of the fine-tuned model, and further optimize and improve the model based on the evaluation results.

[0033] Furthermore, in step S4, it includes:

[0034] S4-1, approval model construction;

[0035] The trained Hairuo big model is combined with the approval process to build an automatic approval model for government information projects. Based on the approval model, it provides project establishment basis, plan key points extraction, plan construction content, security protection and confidentiality evaluation, and plan content comparison.

[0036] S4-2, Opening of API interface;

[0037] Provide an open API interface and write detailed API documentation and sample codes.

[0038] Furthermore, in step S5, it includes:

[0039] S5-1, system testing;

[0040] S5-2, system optimization;

[0041] Further, in step S5-1, the response time and throughput performance indicators of the system are comprehensively tested, and the security of the system is tested;

[0042] In step S5-2, the algorithm is optimized and the resource allocation strategy is adjusted to improve the operating efficiency and stability of the system, collect user feedback and opinions, and carefully optimize and improve the system's interface design and operating procedures.

[0043] A device for realizing automatic approval of information technology projects, comprising: at least one memory and at least one processor;

[0044] The at least one memory is used to store a machine-readable program;

[0045] The at least one processor is used to call the machine-readable program to execute a method for realizing automatic approval of information technology projects.

[0046] Compared with the prior art, the method and device for realizing automatic approval of information technology projects of the present invention have the following outstanding beneficial effects:

[0047] The present invention utilizes the current implementation method to greatly improve the review efficiency, reduce human intervention and errors, and ensure the objectivity and accuracy of the review results. It effectively reduces the workload of reviewers and reduces time costs, allowing them to devote more energy to in-depth research and discussion of complex and difficult projects, further improving the quality of review. This can not only improve the overall efficiency of information construction, but also provide departments with more accurate and efficient services, and promote the development of work in a more intelligent and convenient direction. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] 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.

[0049] Attached Figure 1 The invention is a framework diagram of a method for realizing automatic approval of information technology projects. DETAILED DESCRIPTION

[0050] In order to enable those skilled in the art to better understand the solution of the present invention, the present invention is further described in detail below in conjunction with specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] A best embodiment is given below:

[0052] like Figure 1 As shown, a method for realizing automatic approval of information technology projects in this embodiment has the following steps:

[0053] S1. Integration of computing resources and construction of computing platform;

[0054] include:

[0055] S1-1, computing power resource integration;

[0056] First, we integrate various computing resources such as high-performance servers and GPU computing cards. Hardware resources are the basis for supporting the training and operation of Hairuo's large models, ensuring the performance requirements of the models when processing large amounts of data and complex calculations.

[0057] By building a resource management system, we can effectively schedule and allocate these computing resources. This includes dynamically adjusting resource usage according to different stages of model training to optimize training efficiency and cost.

[0058] S1-2, computing platform construction;

[0059] Based on the integrated computing resources, an efficient and scalable computing platform is built. The platform supports advanced functions such as multi-node parallel computing and distributed storage, providing technical support for the training and deployment of Hairuo large models.

[0060] The necessary software environment, including deep learning framework, database management system, etc., is deployed on the computing platform to provide comprehensive support for model training and application.

[0061] S2, integration of government knowledge base and historical project data;

[0062] include:

[0063] S2-1, Construction of government affairs knowledge base;

[0064] By consulting policies and regulations, industry standards, and professional literature, we collect a large amount of knowledge and information related to government information projects, systematically organize and classify this knowledge, and form a structured government knowledge base. This helps the model better understand and apply professional knowledge in the field of government affairs during the training process.

[0065] S2-2, historical project data integration;

[0066] Collect a large amount of historical project data from government departments and related institutions, including project application materials, approval process records, review results, etc.

[0067] The data was preprocessed by cleaning, deduplication, and labeling to ensure data quality and consistency. At the same time, the data was properly formatted and converted according to the needs of model training.

[0068] S3, Hainuo large model training and optimization;

[0069] include:

[0070] S3-1, customized training;

[0071] Based on deep learning technology, the preliminary architecture of the Hairuo model was constructed. This architecture combines the characteristics of government information projects and has powerful data processing and analysis capabilities.

[0072] The integrated government affairs knowledge base and historical project data are used to conduct in-depth customized training of the model. The accuracy and generalization ability of the model are improved by adjusting model parameters, optimizing loss functions and other strategies.

[0073] S3-2, fine tuning;

[0074] The model was fine-tuned for key information such as extraction of key points of the plan, judgment of the basis for project establishment, and main contents of construction. This helps the model to more accurately understand and identify the core points of the project during the approval process.

[0075] The fine-tuned model was comprehensively evaluated through cross-validation, A / B testing, etc. Based on the evaluation results, the model was further optimized and improved.

[0076] S4, approval model construction and function implementation;

[0077] include:

[0078] S4-1, approval model construction;

[0079] The trained Hairuo model was combined with the approval process to build an automatic approval model for government information projects. The model can automatically handle the uploading, parsing and preprocessing of project documents.

[0080] Based on the approval model, it provides functions such as project establishment basis, plan key points extraction, plan construction content, security protection and confidentiality evaluation, plan content comparison, etc. It helps the reviewers to have a more comprehensive understanding of the project during the approval process and improve the accuracy and efficiency of the approval.

[0081] S4-2, API interface open;

[0082] In order to promote the expansion of functions and convenient integration with third-party applications, an open API interface is provided to facilitate third-party applications to quickly integrate and call the functions of the approval model.

[0083] We also write detailed API documentation and sample codes to provide comprehensive guidance and support for third-party developers.

[0084] S5. Testing and optimization;

[0085] include:

[0086] S5-1, system testing;

[0087] We conduct detailed tests on the functions of the automatic approval system to ensure that each function can be executed accurately and correctly, thereby ensuring the functional integrity and correctness of the system. At the same time, we conduct comprehensive tests on key performance indicators such as the system's response time and throughput to scientifically evaluate the system's performance and ensure that it can still operate stably under high-load environments. In addition, we also conduct rigorous tests on the security of the system, covering multiple aspects such as data encryption and permission control, aiming to ensure the security and reliability of the system in all aspects.

[0088] S5-2, system optimization;

[0089] Based on the test results, we deeply analyzed the performance bottlenecks of the system and made targeted optimizations. By optimizing algorithms and adjusting resource allocation strategies, we significantly improved the operating efficiency and stability of the system. In addition, we also widely collected user feedback and opinions, and made detailed optimizations and improvements to the system's interface design, operating procedures, etc., making the system more in line with user habits and further improving the system's ease of use and user experience.

[0090] Based on the above method, a device for realizing automatic approval of information technology projects in this embodiment includes: at least one memory and at least one processor;

[0091] The at least one memory is used to store a machine-readable program;

[0092] The at least one processor is used to call the machine-readable program to execute a method for realizing automatic approval of information technology projects.

[0093] The above-mentioned specific implementations are only specific cases of the present invention. The patent protection scope of the present invention includes but is not limited to the above-mentioned specific implementations. Any technical solutions that conform to the above-mentioned specific implementations of the present invention and any appropriate changes or substitutions made by ordinary technicians in the relevant technical field shall fall within the patent protection scope of the present invention.

[0094] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for realizing automatic approval of information technology projects, characterized in that: The steps are as follows: S1. Integration of computing resources and construction of computing platform; S2, integration of government knowledge base and historical project data; S3, Hainuo large model training and optimization; S4, approval model construction and function implementation; S5. Testing and optimization.

2. A method for realizing automatic approval of information technology projects according to claim 1, characterized in that: In step S1, it includes: S1-1, computing power resource integration; First, we integrate the diverse computing resources of servers and GPU computing cards, and effectively schedule and allocate these computing resources by building a resource management system. S1-2, computing platform construction; Based on the integrated computing resources, a computing platform is built, and the necessary software environment is deployed on the computing platform, including the deep learning framework and database management system.

3. A method for realizing automatic approval of information technology projects according to claim 2, characterized in that: In step S2, it includes: S2-1, Construction of government affairs knowledge base; Collect a large amount of knowledge and information related to government information projects, systematically organize and classify this knowledge, and form a structured government knowledge base; S2-2, historical project data integration; Collect a large amount of historical project data from relevant departments and institutions, including project application materials, approval process records and review results, clean, deduplicate and standardize the data, and format and convert the data according to the needs of model training.

4. A method for realizing automatic approval of information technology projects according to claim 3, characterized in that: In step S3, it includes: S3-1. Based on deep learning technology, we built the preliminary architecture of the Hairuo model, used the integrated government knowledge base and historical project data to conduct customized training on the model, and improved the accuracy and generalization ability of the model by adjusting model parameters and optimizing loss functions; S3-2, fine tuning; Use cross-validation and A / B testing to conduct a comprehensive evaluation of the fine-tuned model, and further optimize and improve the model based on the evaluation results.

5. A method for realizing automatic approval of information technology projects according to claim 4, characterized in that: In step S4, it includes: S4-1, approval model construction; The trained Hairuo big model is combined with the approval process to build an automatic approval model for government information projects. Based on the approval model, it provides project establishment basis, plan key points extraction, plan construction content, security protection and confidentiality evaluation, and plan content comparison. S4-2, Opening of API interface; Provide an open API interface and write detailed API documentation and sample codes.

6. A method for realizing automatic approval of information technology projects according to claim 5, characterized in that: In step S5, it includes: S5-1, system testing; S5-2. System optimization.

7. A method for realizing automatic approval of information technology projects according to claim 6, characterized in that: In step S5-1, the response time and throughput performance indicators of the system are comprehensively tested, and the security of the system is tested; In step S5-2, the algorithm is optimized and the resource allocation strategy is adjusted to improve the operating efficiency and stability of the system, collect user feedback and opinions, and carefully optimize and improve the system's interface design and operating procedures.

8. A device for realizing automatic approval of information technology projects, characterized in that: include: at least one memory and at least one processor; The at least one memory is used to store a machine-readable program; The at least one processor is configured to call the machine-readable program to execute the method according to any one of claims 1 to 7.

Images