Bidding file generation method and device, equipment and storage medium

Abstract

The invention relates to the field of artificial intelligence, and provides a bidding file generation method, device and equipment and a storage medium, the bidding file generation method comprises the following steps: obtaining a bidding file, and carrying out data enhancement processing on the bidding file to obtain a bidding enhancement file; processing the bidding file to obtain bidding data, and generating a prompt template based on the bidding data; inputting the bidding file, the bidding enhancement file and the prompt template into a preset bidding file generation model to obtain an outline file; wherein the bidding file generation model is obtained based on big language model training; and generating a target bidding file according to the outline file. According to the method, the bidding file, the bidding enhancement file and the prompt template are input into the preset bidding file generation model together for processing, so that the accuracy and efficiency of bidding file analysis and generation are improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular, to a method, device, equipment and storage medium for generating tender documents. Background Art

[0002] With the intensification of global competition and the expansion of project scale, the challenges faced by enterprises in participating in tender activities are increasing. Traditional bid evaluation methods usually rely on manual analysis of tenders, which is both time-consuming and laborious, and is also easily affected by human biases. The main characteristics of the bid evaluation method are manual work, a large number of documents, and the inherent risk of human errors. Effectively storing and managing tender documents, accurately interpreting the content of tender documents, giving reasonable suggestions for different documents, and more efficient work efficiency are all problems that enterprises urgently need to solve.

[0003] Currently, the application of AI in the field of tendering and bidding mainly focuses on single functions such as text classification, keyword extraction, and entity recognition, which limits the efficiency of comprehensive intelligent analysis. Summary of the Invention

[0004] The present invention provides a method, device, equipment and storage medium for generating tender documents to solve the defects in the prior art.

[0005] The present invention provides a method for generating tender documents, including: Obtaining a tender document, and performing data enhancement processing on the tender document to obtain a tender enhanced document; Processing the tender document to obtain tender data, and generating a prompt template based on the tender data; Inputting the tender document, the tender enhanced document and the prompt template into a preset tender document generation model to obtain an outline document; wherein, the tender document generation model is trained based on a large language model; Generating a target tender document according to the outline document.

[0006] According to the method for generating tender documents provided by the present invention, the processing the tender document to obtain tender data, and generating a prompt template based on the tender data includes: Based on the named entity recognition technology, extracting the named entities and entity relationships in the tender document, and extracting the context information in the tender document; Constructing an entity relationship graph according to the named entities and the entity relationships; Based on the entity relationship graph, extracting the structural information and legal information in the tender document through a graph neural network; Generate the prompt template based on the context information, the named entity, the structure information, and the legal information; wherein, the prompt template defines the task objective and generation requirements, and the prompt template includes a task definition, a generation prompt, precautions, and output examples.

[0007] According to a method for generating a tender document provided by the present invention, generating a target tender document according to the outline document includes: Obtain a successful tender document, and analyze the successful tender document to obtain successful tender factors; wherein, the successful tender factors include at least one of the technical solution details and the cost strategy. Adjust the outline document based on the successful tender factors to obtain the target tender document.

[0008] According to a method for generating a tender document provided by the present invention, after generating the target tender document according to the outline document, the method further includes: Obtain a personnel qualification library; wherein, the personnel qualification library includes multiple personnel qualification documents. Calculate the text similarity between each personnel qualification document and the tender document corresponding to the target tender document. Determine the tender participation personnel of the target tender document according to the text similarity.

[0009] According to a method for generating a tender document provided by the present invention, performing data enhancement processing on the tender document to obtain an enhanced tender document includes: Perform data enhancement processing on the tender document by synonym replacement and / or sentence pattern conversion to obtain an enhanced tender document.

[0010] According to a method for generating a tender document provided by the present invention, before performing data enhancement processing on the tender document to obtain an enhanced tender document, the method further includes: Remove irrelevant characters from the tender document and perform standardization processing on the tender document. Correct grammar errors and spelling mistakes in the tender document.

[0011] According to a method for generating a tender document provided by the present invention, after generating the target tender document according to the outline document, the method further includes: Real-time monitor industry trends to obtain a supplementary tender document. Adjust the target tender document based on the supplementary tender document.

[0012] The present invention also provides a tender document generation device, including: The first processing module is configured to obtain a bidding document and perform data augmentation processing on the bidding document to obtain an enhanced bidding document; The second processing module is configured to process the bidding document to obtain bidding data and generate a prompt template based on the bidding data; The input module is configured to input the bidding document, the enhanced bidding document, and the prompt template into a preset bidding document generation model to obtain an outline document; wherein, the bidding document generation model is trained based on a large language model; The generation module is configured to generate a target bidding document according to the outline document.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the method for generating a bidding document as described in any one of the above is implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method for generating a bidding document as described in any one of the above is implemented.

[0015] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, the method for generating a bidding document as described in any one of the above is implemented.

[0016] The method, device, equipment, and storage medium for generating a bidding document provided by the present invention perform data augmentation processing on the bidding document to obtain an enhanced bidding document. By data augmentation processing to increase bidding-related documents, the subsequent bidding document generation model can access more data from different angles and forms, improving the adaptability of the bidding document generation model to complex scenarios. The bidding document is processed to obtain bidding data, and a prompt template is generated based on the bidding data. There is a bidding document generation model pre-trained based on a large language model. The bidding document generation model intelligently analyzes the input document and automatically generates an outline document, and then optimizes the content according to the generated outline document to generate a target bidding document. The present invention improves the accuracy and efficiency of bidding document parsing and generation by inputting the bidding document, the enhanced bidding document, and the prompt template into a preset bidding document generation model for processing, reduces the dependence on labeled data, enhances the generalization ability of the model, simplifies the bidding document processing process, and reduces the necessity of manual intervention. The automated system generates a target bidding document, improving the processing efficiency and reducing the consumption of human resources, while reducing the document error rate and significantly improving the document quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] To more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0018] Figure 1 It is a schematic flowchart of the bidding document generation method provided by the present invention.

[0019] Figure 2 It is a schematic diagram of the bidding document generation method provided by the present invention.

[0020] Figure 3 It is a schematic structural diagram of the bidding document generation device provided by the present invention.

[0021] Figure 4 It is a schematic structural diagram of the electronic device provided by the present invention. Detailed implementation manners

[0022] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention in conjunction with the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0023] The user information (including but not limited to user personal information, award information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. And the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or refuse.

[0024] Figure 1 It is a flowchart of a bidding document generation method shown according to an exemplary embodiment. As Figure 1 shown, in an exemplary embodiment, the bidding document generation method includes steps 110 to 140, which are introduced in detail as follows.

[0025] Step 110, obtain a bidding document, and perform data enhancement processing on the bidding document to obtain a bid-enhanced document.

[0026] In an embodiment of the present invention, the original bidding data is pre - processed to obtain bidding documents. The bidding documents are documents related to bidding, which can be one document or multiple documents. For example, Figure 2 As shown, the bidding documents are processed by data augmentation to obtain enhanced bidding documents. By data augmentation, more documents related to bidding are added, so that the subsequent bidding document generation model can access data from more different angles and forms, improving the adaptability of the bidding document generation model to complex scenarios.

[0027] Step 120: Process the bidding documents to obtain bidding data, and generate a prompt template based on the bidding data.

[0028] In an embodiment of the present invention, the bidding documents are processed to obtain bidding data, such as company, time, location, value, etc., and a prompt template is generated based on the bidding data.

[0029] Step 130: Input the bidding documents, the enhanced bidding documents, and the prompt template into a preset bidding document generation model to obtain an outline document; wherein, the bidding document generation model is trained based on a large - language model.

[0030] In an embodiment of the present invention, a bidding document generation model is pre - trained based on a large - language model (Large Language Model, LLM). The bidding document generation model intelligently parses the input documents and automatically generates an outline document. The large - language model used in this embodiment of the present invention includes DeepSeek_r1_Distill_Qwen32B.

[0031] Step 140: Generate a target bidding document according to the outline document.

[0032] In an embodiment of the present invention, the generated outline document is optimized in content to generate a target bidding document.

[0033] In an embodiment of the present invention, by inputting the bidding documents, the enhanced bidding documents, and the prompt template into a preset bidding document generation model for processing together, the accuracy and efficiency of bidding document parsing and generation are improved, the dependence on labeled data is reduced, the generalization ability of the model is enhanced, at the same time, the bidding document processing flow is simplified, and the necessity of manual intervention is reduced. The automated system generates the target bidding document, improving the processing efficiency and reducing the consumption of human resources, while reducing the document error rate and significantly improving the document quality, thereby enhancing the image and reputation of the enterprise; and through automated optimization and high - quality bidding document generation, the enterprise can participate in more bidding projects, the average winning bid rate is effectively improved, and the market competitiveness is enhanced to a certain extent.

[0034] In an exemplary embodiment of the present invention, processing the bid documents to obtain bid data and generating a prompt template based on the bid data includes: Extracting named entities and entity relationships in the bid documents based on named entity recognition technology, and extracting context information in the bid documents; Constructing an entity relationship graph according to the named entities and the entity relationships; Extracting structural information and legal information in the bid documents through a graph neural network based on the entity relationship graph; Generating the prompt template based on the context information, the named entities, the structural information, and the legal information; wherein the prompt template defines a task objective and generation requirements, and the prompt template includes a task definition, a generation prompt, precautions, and an output example.

[0035] In the embodiment of the present invention, named entities and entity relationships are extracted from bid documents by using named entity recognition technology (NER). The named entities in this embodiment are key terms in this field. According to the named entities and entity relationships, an entity relationship graph is quickly constructed. At the same time, based on the constructed entity relationship graph, the entity relationships are dynamically adjusted through named entity recognition technology and a graph neural network (GNN) to reflect the latest document information and its associations. The structural information and legal information in the bid documents are captured through a graph neural network. Context information is extracted through Time-NLP and NER to optimize text understanding. For example, a construction plan for coping with specific seasonal climates is generated based on the date, or the content is adjusted according to the address. In the information extraction stage, special attention is paid to the dynamic update of the entity relationship graph, which includes automatically updating contract terms, responsible persons, legal obligations, etc. Through context analysis, information such as dates and locations is accurately extracted to provide context clues during the document generation process to improve the accuracy and relevance of the generated content.

[0036] A prompt template with a specific format and style is designed to guide the bid document generation model to generate information with a clear structure. Further, the results generated by the bid document generation model can be evaluated manually and automatically to iteratively optimize the prompt template. In view of the characteristics of large language models, multiple prompt templates are designed. These templates define task objectives and generation requirements, including not modifying the original document information, prohibiting the output of redundant information, etc. Each prompt template includes a task definition, a generation prompt, precautions, and an output example. By providing these standard guidelines, it is ensured that the format and content of the generated document are consistent with the expectations. After the prompt template is generated, it is evaluated manually to identify and optimize problems in the generation, and the effectiveness and reliability of the prompt template are improved through continuous iteration.

[0037] In an exemplary embodiment of the present invention, generating a target tender document according to the outline document includes: Obtain a tender success document, and analyze the tender success document to obtain tender success factors; wherein, the tender success factors include at least one of the technical solution details and cost strategy; Adjust the outline document based on the tender success factors to obtain the target tender document.

[0038] In the embodiment of the present invention, the tender document generation model generates a preliminary outline text, and then customizes and optimizes the content according to the tender success document and historical data. The generated outline document is generated according to the specific tender document requirements to ensure that it covers technical responses and business proposals. Using the tender success document, the outline document is automatically optimized through a machine learning model. Specifically, analyze the tender success factors in the tender success document, such as the technical solution details, cost strategy, etc., and adjust the content in the outline document accordingly to obtain the target tender document.

[0039] In an exemplary embodiment of the present invention, after generating the target tender document according to the outline document, the method further includes: Obtain a personnel qualification library; wherein, the personnel qualification library includes multiple personnel qualification documents; Calculate the text similarity between each personnel qualification document and the tender document corresponding to the target tender document; Determine the tender participation personnel of the target tender document according to the text similarity.

[0040] In the embodiment of the present invention, a company and a personnel qualification library are constructed, and the personnel qualification library is configured with qualification documents corresponding to multiple personnel. Through natural language processing technology and machine learning models for intelligent matching, the most suitable tender participation personnel combination for the target tender document is selected. Specifically, compare the personnel qualification documents in the qualification library with the qualification requirements in the tender document, and use text similarity calculation and machine learning to match the most relevant tender participation personnel. At the same time, analyze the personnel participating in successful tenders in history, determine the most effective qualification combination according to the analysis results, and use collaborative filtering technology to recommend the most likely winning qualification combination.

[0041] In the embodiment of the present invention, the employee qualification library is updated and verified regularly to ensure the accuracy of the information.

[0042] In an exemplary embodiment of the present invention, performing data enhancement processing on the tender document to obtain a tender enhanced document includes: Perform data enhancement processing on the tender document through synonym replacement and / or sentence pattern conversion to obtain a tender enhanced document.

[0043] In the embodiments of the present invention, after data preprocessing, techniques such as synonym replacement and sentence pattern conversion are applied for data augmentation. This enables the model to better handle diverse text input forms, thereby enhancing the text understanding ability.

[0044] In an exemplary embodiment of the present invention, before performing data augmentation processing on the tender document to obtain an enhanced tender document, the method further includes: Removing irrelevant characters from the tender document and performing standardization processing on the tender document; Correcting grammar errors and spelling mistakes in the tender document.

[0045] In the embodiments of the present invention, irrelevant characters in the tender document are removed, such as special symbols, headers and footers, extra spaces, etc., while the date format (ISO 8601) and currency units (such as USD) are standardized. The Grammarly API is integrated to correct grammar errors and spelling mistakes in the tender document.

[0046] When removing irrelevant characters from the tender document, read the tender document, define regular expressions for headers and footers, initialize an empty string or list, and save the processed tender document. Loop through each line or the entire text of the processed tender document, check whether the current line or text matches the regular expression of the header. If the current line or text matches the regular expression of the header, extract the header content and remove the header identifier, and save the processed content to a variable of the tender document, then traverse the next line or text; if the current line or text does not match the regular expression of the header, check whether the current line or text matches the regular expression of the footer. If the current line or text matches the regular expression of the footer, extract the footer content and remove the footer identifier, and save the processed content to a variable of the tender document, then traverse the next line or text; if the current line or text does not match the regular expression of the footer, directly save the current line or text to a variable of the tender document, then traverse the next line or text until all lines or text are traversed.

[0047] In an exemplary embodiment of the present invention, after generating the target tender document according to the outline document, the method further includes: Real-time monitoring of industry trends to obtain a supplementary tender document; Adjusting the target tender document based on the supplementary tender document.

[0048] In the embodiments of the present invention, industry trends are monitored in real time, any new supplementary tender documents are automatically integrated, and the target tender document is dynamically adjusted according to the supplementary tender document to ensure the up-to-dateness and accuracy of the tender outline.

[0049] The technical architecture for implementing the bid document generation method includes an infrastructure layer, a cloud-native layer, a model layer, an application technology layer, an application architecture layer, and an application layer. The infrastructure layer includes infrastructure such as GPU (Graphical Processing Unit), CPU (Central Processing Unit), RAM (Random Access Memory), HDD (Hard Disk Drive), and Network; the cloud-native layer deploys Docker and K&S (Kubernetes); the model layer provides access, management, computing power scheduling, and fine-tuning capabilities, and deploys various models such as large language models, image recognition OCR models, document understanding models, and multi-modal detection and analysis models; the application technology layer sets corresponding intelligent agents, and at the same time implements retrieval augmented generation (RAG), prompt engineering, chain-of-thought (COT), fine-turning, data scraping, data cleaning, data vectorization, and access control. The application architecture layer includes an engineering technology architecture, a business architecture, and a cloud-native architecture. The application layer is set according to specific needs. For example, in the present invention, bid documents need to be generated. Therefore, the application layer can implement applications such as on-line analysis processing (OLAP) to achieve enterprise-level document generation.

[0050] The bid document generation device provided by the present invention will be described below. The bid document generation device described below can be correspondingly referred to the bid document generation method described above. It should be noted that the device provided in the following embodiments and the method provided in the above embodiments belong to the same concept. The specific ways in which each module and unit perform operations have been described in detail in the method embodiments and will not be repeated here.

[0051] In an exemplary embodiment of the present invention, please refer to Figure 3 , Figure 3 which is a bid document generation device shown according to an exemplary embodiment and includes the following modules.

[0052] The first processing module 310 is configured to obtain a bid document and perform data enhancement processing on the bid document to obtain a bid enhanced document; The second processing module 320 is configured to process the bid document to obtain bid data and generate a prompt template based on the bid data; An input module 330, configured to input the tender document, the tender enhancement document, and the prompt template into a preset tender document generation model to obtain an outline document; wherein, the tender document generation model is trained based on a large language model; A generation module 340, configured to generate a target tender document according to the outline document.

[0053] In an exemplary embodiment of the present invention, the second processing module 320 includes: A first extraction sub-module, configured to extract named entities and entity relationships in the tender document based on named entity recognition technology, and extract context information in the tender document; A construction sub-module, configured to construct an entity relationship graph according to the named entities and the entity relationships; A second extraction sub-module, configured to extract structural information and legal information in the tender document through a graph neural network based on the entity relationship graph; A generation sub-module, configured to generate the prompt template based on the context information, the named entities, the structural information, and the legal information; wherein, the prompt template defines a task objective and generation requirements, and the prompt template includes a task definition, a generation prompt, precautions, and an output example.

[0054] In an exemplary embodiment of the present invention, the generation module 340 includes: An acquisition sub-module, configured to acquire a tender success document and analyze the tender success document to obtain tender success factors; wherein, the tender success factors include at least one of the technical solution detail degree and the cost strategy; An adjustment sub-module, configured to adjust the outline document based on the tender success factors to obtain the target tender document.

[0055] In an exemplary embodiment of the present invention, the tender document generation device further includes: An acquisition module, configured to acquire a personnel qualification library; wherein, the personnel qualification library includes a plurality of personnel qualification documents; A calculation module, configured to calculate the text similarity between each personnel qualification document and the tender document corresponding to the target tender document; A determination module, configured to determine the tender participation personnel of the target tender document according to the text similarity.

[0056] In an exemplary embodiment of the present invention, the first processing module 310 includes: An enhancement processing sub-module, configured to perform data enhancement processing on the tender document through synonym replacement and / or sentence pattern conversion to obtain a tender enhancement document.

[0057] In an exemplary embodiment of the present invention, the tender document generation device further includes: A removal module configured to remove irrelevant characters from the tender document and perform standardization processing on the tender document; A correction module configured to correct grammar errors and spelling mistakes in the tender document.

[0058] In an exemplary embodiment of the present invention, the tender document generation device further includes: A monitoring module configured to monitor industry dynamics in real time to obtain a tender supplementary document; An adjustment module configured to adjust the target tender document based on the tender supplementary document.

[0059] Figure 4 The schematic diagram of the physical structure of an electronic device is exemplified, as Figure 4 shown. The electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440. Among them, the processor 410, the communication interface 420, and the memory 430 complete mutual communication through the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute the tender document generation method, which includes: obtaining a tender document and performing data enhancement processing on the tender document to obtain a tender enhanced document; Processing the tender document to obtain tender data and generating a prompt template based on the tender data; Inputting the tender document, the tender enhanced document, and the prompt template into a preset tender document generation model to obtain an outline document; wherein, the tender document generation model is trained based on a large language model; Generating a target tender document according to the outline document.

[0060] In addition, when the logical instructions in the above-mentioned memory 430 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0061] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the tender document generation method provided by the above-mentioned various methods. The method includes: obtaining a tender document, and performing data enhancement processing on the tender document to obtain an enhanced tender document; Processing the tender document to obtain tender data, and generating a prompt template based on the tender data; Inputting the tender document, the enhanced tender document, and the prompt template into a preset tender document generation model to obtain an outline document; wherein, the tender document generation model is trained based on a large language model; Generating a target tender document according to the outline document.

[0062] On another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the tender document generation method provided by the above-mentioned various methods. The method includes: obtaining a tender document, and performing data enhancement processing on the tender document to obtain an enhanced tender document; Processing the tender document to obtain tender data, and generating a prompt template based on the tender data; Inputting the tender document, the enhanced tender document, and the prompt template into a preset tender document generation model to obtain an outline document; wherein, the tender document generation model is trained based on a large language model; Generating a target tender document according to the outline document.

[0063] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0064] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A bidding document generation method, characterized in that: include: Obtaining a bidding document, and performing data enhancement processing on the bidding document to obtain a bidding enhancement document; Processing the bidding document to obtain bidding data, and generating a prompt template based on the bidding data; Inputting the bidding document, the bidding enhancement file and the prompt template into a preset bidding document generation model to obtain an outline file; wherein the bidding document generation model is obtained based on the training of a large language model; Generate a target bidding document according to the outline document.

2. The bidding document generation method according to claim 1, characterized in that: The step of processing the bidding document to obtain bidding data and generating a prompt template based on the bidding data includes: Based on named entity recognition technology, extracting named entities and entity relationships in the bidding document, and extracting context information in the bidding document; Constructing an entity relationship graph according to the named entities and the entity relationships; Based on the entity relationship diagram, extracting structural information and legal information from the bidding document through a graph neural network; The prompt template is generated based on the context information, the named entity, the structural information and the legal information; wherein the prompt template defines the task goal and generation requirements, and the prompt template includes a task definition, generation prompts, precautions and output examples.

3. The bidding document generation method according to claim 1, characterized in that: The step of generating a target bidding document according to the outline document comprises: Obtaining a successful bidding document, and analyzing the successful bidding document to obtain a successful bidding factor; wherein the successful bidding factor includes at least one of the details of the technical solution and the cost strategy; The outline document is adjusted based on the bidding success factors to obtain the target bidding document.

4. The method for generating bidding documents according to claim 3, characterized in that: After generating the target bidding document according to the outline document, the method further comprises: Obtaining a personnel qualification database; wherein the personnel qualification database includes a plurality of personnel qualification files; Calculating the text similarity between each of the personnel qualification documents and the bidding document corresponding to the target bidding document; The bidding participants of the target bidding document are determined according to the text similarity.

5. The bidding document generation method according to claim 1, characterized in that: The step of performing data enhancement processing on the bidding document to obtain a bidding enhancement document includes: The bid document is subjected to data enhancement processing by synonym replacement and / or sentence conversion to obtain a bid enhancement document.

6. The bidding document generation method according to claim 1, characterized in that: Before performing data enhancement processing on the bidding document to obtain the bidding enhancement document, the method further includes: Removing irrelevant characters from the bidding document and standardizing the bidding document; Correct grammatical errors and spelling mistakes in said bid documents.

7. The bidding document generation method according to any one of claims 1 to 6, characterized in that: After generating the target bidding document according to the outline document, the method further comprises: Monitor industry trends in real time and obtain supplementary bidding documents; The target bid document is adjusted based on the bid supplement document.

8. A bidding document generating device, characterized in that: include: A first processing module is configured to obtain a bidding document and perform data enhancement processing on the bidding document to obtain a bidding enhancement document; a second processing module configured to process the bidding document to obtain bidding data and generate a prompt template based on the bidding data; An input module is configured to input the bidding document, the bidding enhancement file and the prompt template into a preset bidding document generation model to obtain an outline file; wherein the bidding document generation model is obtained based on the training of a large language model; A generating module is configured to generate a target bidding document according to the outline document.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the bidding document generation method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the bidding document generation method as described in any one of claims 1 to 7 is implemented.

Images