Biding document generation method and system based on large model

Abstract

The invention relates to the technical field of document generation, in particular to a bidding document generation method and system based on a large model. The method comprises the steps of obtaining current enterprise data, extracting enterprise keywords for each piece of current data, and establishing an enterprise knowledge base; obtaining a bid invitation file, extracting bid invitation keywords, and generating a structured demand list; obtaining bidding keywords, identifying corresponding enterprise keywords and corresponding current data, dynamically generating bidding document contents, and outputting the bidding document contents; the system comprises an enterprise keyword extraction module, a bidding keyword extraction module and a bidding document dynamic generation module. According to the method, enterprise keyword recognition is performed through the structured demand list generated for the bidding keyword, and the bidding document is generated, so that the bidding document content can be dynamically generated according to the bidding document demand; the problem that the enterprise is in a passive position in the bidding process and misses business opportunities due to the fact that the requirement that the enterprise efficiently and accurately deals with the bidding demand cannot be met is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of document generation, and particularly to a tender generation method and system based on a large model. Background Art

[0002] In today's highly competitive business environment, it is common for enterprises to participate in various project tenders to obtain business opportunities. However, during the tendering process, enterprises face numerous challenges. On the one hand, enterprises have accumulated a large amount of data, including historical project cases, qualification documents, and technical parameters, etc. These data are scattered in different departments, systems, and documents, lacking effective integration and management. When preparing tenders for specific tender projects, enterprise employees often need to spend a great deal of time and effort searching, screening, and organizing relevant information among numerous documents. This not only has low efficiency but also easily leads to information omission or errors, affecting the quality of the tender and the success rate of the tendering. On the other hand, tender documents usually have complex structures and diverse requirements. The tenderer will put forward various forms of requirements for the technical, commercial, and other aspects of the project. These requirements may be hidden in the large amount of text in the tender documents, making it difficult for enterprises to quickly and accurately extract and understand the key information and transform it into tender content that meets the tender requirements.

[0003] Traditional manual processing methods cannot meet the requirements of enterprises to efficiently and accurately respond to tender needs, resulting in enterprises being in a passive position during the tendering process and missing business opportunities. Therefore, it is very necessary to propose a tender generation method and system that can dynamically generate tender content according to the requirements of tender documents. Summary of the Invention

[0004] The purpose of the present invention is to provide a tender generation method and system based on a large model, aiming to solve the technical problem in the prior art that it cannot meet the requirements of enterprises to efficiently and accurately respond to tender needs, resulting in enterprises being in a passive position during the tendering process and missing business opportunities.

[0005] To achieve the above purpose, a tender generation method based on a large model adopted by the present invention includes the following steps: Obtain the current data of the enterprise, extract enterprise keywords for each piece of current data, and establish an enterprise knowledge base; Obtain the tender document, extract tender keywords, and generate a structured requirement list; Obtain the tender keywords, identify the corresponding enterprise keywords and their corresponding current data, dynamically generate tender content, and output.

[0006] Among them, in the step of obtaining the current data of the enterprise, extracting enterprise keywords for each piece of current data, and establishing an enterprise knowledge base: Collect the current data of the enterprise and detect the integrity and accuracy of the current data; the enterprise knowledge base includes historical project cases, qualification documents, and technical parameters; Use a large model to extract enterprise keywords for each piece of collected current data; Associate the extracted keywords with the corresponding data and store them in the database to form an enterprise knowledge base.

[0007] Among them, in the step of collecting the current data of the enterprise and detecting the integrity and accuracy of the current data, the process of integrity detection is: Check the fields included in each piece of data and output the detection results.

[0008] Among them, in the step of collecting the current data of the enterprise and detecting the integrity and accuracy of the current data, the process of accuracy detection is: For numerical data, check whether the data is within a reasonable range; For text data, check whether there are grammar, logic, and contradictory information in the data and output the detection results.

[0009] Among them, in the step of using a large model to extract enterprise keywords for each piece of collected current data: For historical project cases, enterprise keywords include project name, project type, key technologies, and project achievements; for qualification documents, enterprise keywords include qualification name, issuing agency, and validity period; for technical parameters, enterprise keywords include product model, performance indicators, and technical specifications.

[0010] Among them, in the step of obtaining the bidding documents, extracting bidding keywords, and generating a structured requirements list: Obtain the bidding documents and use a large model to extract bidding keywords from the bidding documents; the bidding keywords include project name, bidding scope, technical requirements, and commercial terms; According to the extracted bidding keywords, structure the requirements in the bidding documents to form a requirements list.

[0011] Among them, before the step of structuring the requirements in the bidding documents to form a requirements list: Dynamically adjust the priority of the bidding keywords.

[0012] Among them, in the step of obtaining the bidding keywords, identifying the corresponding enterprise keywords and their corresponding current data, dynamically generating the tender content, and outputting: Obtain the bidding keywords and use a keyword matching algorithm to match the bidding keywords with the enterprise keywords in the enterprise knowledge base; Dynamically generate the tender content according to the matched enterprise data and bidding requirements.

[0013] After the step of dynamically generating the tender content according to the matched enterprise data and tender requirements: Review and proofread the tender, and output the tender content.

[0014] The present invention also provides a tender generation system based on a large model, including an enterprise keyword extraction module, a tender keyword extraction module, and a tender dynamic generation module; wherein: The enterprise keyword extraction module is used to obtain the current enterprise data, extract enterprise keywords for each piece of current data, and establish an enterprise knowledge base; wherein the enterprise knowledge base includes historical project cases, qualification documents, and technical parameters; The tender keyword extraction module is used to obtain the tender document, extract tender keywords, and generate a structured requirement list; The tender dynamic generation module is used to obtain the tender keywords, identify the corresponding enterprise keywords and their corresponding current data, dynamically generate the tender content, and output it.

[0015] A method and system for generating a tender based on a large model according to the present invention respectively adopt the enterprise keyword extraction module, the tender keyword extraction module, and the tender dynamic generation module to perform the following steps: obtain the current enterprise data, extract enterprise keywords for each piece of current data, and establish an enterprise knowledge base; obtain the tender document, extract tender keywords, and generate a structured requirement list; obtain the tender keywords, identify the corresponding enterprise keywords and their corresponding current data, dynamically generate the tender content, and output it; through the structured requirement list generated for the tender keywords, perform enterprise keyword identification, and generate a tender, so as to be able to dynamically generate the tender content according to the requirements of the tender document; avoid the situation that the requirements for the enterprise to efficiently and accurately respond to the tender cannot be met, resulting in the enterprise being in a passive position during the tender process and missing business opportunities. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0017] Figure 1 is a flowchart of the steps of the method for generating a tender based on a large model of the present invention.

[0018] Figure 2 is a flowchart of the steps of S100 of the present invention.

[0019] Figure 3It is the flowchart of step S200 of the present invention.

[0020] Figure 4 It is the flowchart of step S300 of the present invention.

[0021] Figure 5 It is the structural schematic diagram of the bid document generation system based on the large model of the present invention.

[0022] Figure 6 It is the structural schematic diagram of the electronic device of the present invention.

[0023] 401 - Enterprise keyword extraction module, 402 - Bidding keyword extraction module, 403 - Bid document dynamic generation module. Detailed implementation mode

[0024] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all the implementation manners consistent with the present application.

[0025] The terms used in the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms of "a", "the" and "said" used in the present application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0026] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to determining".

[0027] Please refer to Figures 1 to 4 , the present invention provides a bid document generation method based on a large model, including the following steps: S100: Obtain the current enterprise data, extract enterprise keywords for each piece of current data, and establish an enterprise knowledge base; wherein the enterprise knowledge base includes historical project cases, qualification documents, and technical parameters.

[0028] In this embodiment, the current enterprise data is obtained, enterprise keywords are extracted for each piece of current data, and an enterprise knowledge base is established. The specific process is as follows: S101: Collect the current data of the enterprise and detect the integrity and accuracy of the current data; the enterprise knowledge base includes historical project cases, qualification documents, and technical parameters. S102: Use a large model to extract enterprise keywords for each piece of collected current data. S103: Associate the extracted keywords with the corresponding data and store them in the database to form an enterprise knowledge base.

[0029] In the above process: Collect data such as historical project cases, qualification documents, and technical parameters from various internal systems of the enterprise (such as ERP, CRM, file servers, etc.) and paper documents. For historical project cases, collect information such as project name, project cycle, project achievements, and key technologies; for qualification documents, collect information such as qualification name, issuing agency, validity period, and qualification level; for technical parameters, collect information such as product model, performance indicators, and technical specifications. At the same time, for historical project cases, enterprise keywords include project name, project type, key technology, and project achievements; for qualification documents, enterprise keywords include qualification name, issuing agency, and validity period; for technical parameters, enterprise keywords include product model, performance indicators, and technical specifications.

[0030] Perform integrity and accuracy detection on the above-mentioned current data of the enterprise respectively.

[0031] Among them, in the process of integrity detection: Check the fields included in each piece of data and output the detection results. For example: for historical project cases, fields such as project name, project cycle, and project achievements must be included; for qualification documents, fields such as qualification name, issuing agency, and validity period must be included. If a piece of data lacks a necessary field, it is determined that the data is incomplete.

[0032] Example formula: Let the data field set be F = {f1, f2,..., f n}, and the actual field set of a piece of data be F actual = {f a1 , f a2 ,..., f am} (m ≤ n). If there exists and , then the data is incomplete.

[0033] Among them, in the process of accuracy detection: For numerical data, check whether it is within a reasonable range. For example, whether the performance indicator values in technical parameters conform to industry common sense or product specification requirements.

[0034] For text data, check whether it conforms to grammar rules, is logically reasonable, and whether there are obvious errors or contradictory information. For example, whether the validity period in the qualification documents is reasonable (the start date is earlier than the end date), whether the description of project achievements is clear and accurate, etc.

[0035] Example formula (taking numerical data as an example): Let the reasonable range of a certain technical parameter be [a, b], and the actual value of this parameter be x. If x <a 或 x>If it is b, then the data is inaccurate.

[0036] Use a large model to extract enterprise keywords for each piece of currently collected data; for historical project cases, enterprise keywords include project name, project type, key technologies, and project achievements; for qualification documents, enterprise keywords include qualification name, issuing agency, and validity period; for technical parameters, enterprise keywords include product model, performance indicators, and technical specifications.

[0037] During the use of the large model, clean the collected data by removing punctuation marks, special characters, stop words (such as meaningless words like "de", "shi", "he", etc.). Perform word segmentation to split the text into meaningful words. For example, use a Chinese word segmentation tool (such as Jieba) to segment Chinese text.

[0038] Adopt the TF-IDF algorithm to extract keywords. Among them, the TF-IDF algorithm: Term Frequency (TF): Calculate the frequency of a certain word appearing in a document. The formula is:

[0039] Where n i,j is the number of times the word t i appears in the document d j and is the total number of times all words appear in the document d j

[0040] Inverse Document Frequency (IDF): Measure the general importance of a word. The formula is:

[0041] Where |D| is the total number of documents, is the number of documents containing the word t i

[0042] TF-IDF value: TF - IDF i,j =TF i,j × IDF i The larger the TF-IDF value, the more important the word is in the current document and the more likely it is to be a keyword.

[0043] According to the TF-IDF values, select the words with higher scores as keywords. For example, the calculated TF-IDF values of some words are as follows: Word: Detection; TF-IDF value: 0.55; ​​Word: Image; TF-IDF value: 0.6; Word: Analysis; TF-IDF value: 0.45; Word: In; TF-IDF value: 0.1; Word: Of; TF-IDF value: 0.05.

[0044] Observing the above TF-IDF values, it can be found that words such as "in" and "of" have significantly lower values and are usually stop words, contributing less to the expression of the document theme. A threshold can be set, such as 0.5, to screen out words with TF-IDF values higher than 0.5 as keywords. After screening, "detection" and "image" are obtained as enterprise keywords.

[0045] Assign a unique identifier (such as ID) to each piece of data, and associate the extracted keywords with this identifier. For example, for a piece of historical project case data with an ID of "001", the extracted keywords are "project innovation", "efficient implementation", etc., and an association relationship is established between the keywords and "001".

[0046] Select a suitable database for storage, such as the relational database MySQL. In the relational database, two tables can be created. One table is used to store the basic information of the data (such as project name, project cycle, etc.), and the other table is used to store keyword information, and the two tables are associated through a foreign key. For example, the basic information table of the data contains fields: ID, project name, project cycle, etc.; the keyword information table contains fields: ID (associated with the ID in the basic information table of the data), keywords, etc. Store the associated data and keywords in the database to form an enterprise knowledge base for convenient subsequent query and use.

[0047] S200: Obtain the bidding documents, extract the bidding keywords, and generate a structured requirements list.

[0048] In this embodiment, obtain the bidding documents, extract the bidding keywords, and generate a structured requirements list. The specific process is as follows: S201: Obtain the bidding documents and use a large model to extract the bidding keywords from the bidding documents; the bidding keywords include project name, bidding scope, technical requirements, and commercial terms; S202: Dynamically adjust the priority of the bidding keywords; S203: According to the extracted bidding keywords, structure and organize the requirements in the bidding documents to form a requirements list.

[0049] In the above process: Obtain the bidding documents and use a large model to extract the bidding keywords from the bidding documents; the process of the large model extracting the bidding keywords from the bidding documents is the same as that in step S100; the bidding keywords include project name, bidding scope, technical requirements, and commercial terms.

[0050] The right of priority is determined according to the following evaluation criteria: Importance evaluation: Evaluate the importance of keywords based on factors such as their positions in the tender documents, frequencies of occurrence, and degrees of association with the core content of the project. For example, the project name and tender scope are usually crucial for the overall understanding of the project and should be given higher priorities; while some keywords related to auxiliary technical details may have relatively lower priorities.

[0051] Timeliness evaluation: Consider the time requirements in different parts of the tender documents, such as the bid deadline, project implementation time, etc. If a keyword is closely related to the time requirement, such as "The bid deadline is [specific date]", then the task or requirement corresponding to this keyword has high timeliness and its priority should be appropriately increased.

[0052] Risk evaluation: Analyze the risks that the content involved in the keywords may bring. For example, if the key indicators in the technical requirements cannot be met, it may lead to bid failure or difficulties in project implementation. Then the tasks corresponding to these keywords have high risks and need to be focused on and their priorities increased.

[0053] During the tender preparation process, continuously pay attention to whether there are updates or supplementary instructions in the tender documents. If the tenderer modifies or emphasizes certain content, it is necessary to adjust the priorities of relevant keywords in a timely manner. For example, if the tenderer raises the requirements for a certain technical indicator in the supplementary instructions, the priorities of the keywords related to this indicator should be increased accordingly.

[0054] The priorities of keywords are also adjusted based on the discussions and experiences of the tender team. The members of the tender team may have different understandings and focuses on the tender documents. Through collective discussions and feedback, the importance of keywords can be evaluated more comprehensively, thereby dynamically adjusting the priorities. For example, if the team members find that there may be greater difficulties in the actual operation of a certain business clause, the priorities of the keywords related to this clause should be increased.

[0055] For example: Based on the three evaluation criteria of importance, timeliness, and risk, a comprehensive priority adjustment formula can be constructed. Let the priority of the keyword be P, the score of importance evaluation be I, the score of timeliness evaluation be T, the score of risk evaluation be R, and a weight be assigned to each criterion, which are set as w I 、w T 、w R respectively, and w I +w T +w R = 1. Then the priority P adjustment formula can be expressed as: P = w I × I + w T × T + w R ×R Description of index scores: Importance assessment score I: Quantitative scoring can be carried out based on the position of the keyword in the tender document, the frequency of occurrence, and the degree of association with the core content of the project. For example, the more important the position (such as the title, the beginning of a paragraph, etc.), the higher the score; the higher the frequency of occurrence, the higher the score; the closer the association with the core content of the project, the higher the score. A full score (such as 10 points) can be set and scored according to the specific situation.

[0056] Timeliness assessment score T: Score according to the degree of closeness of the keyword to the time requirement. If the keyword is directly related to key time nodes such as the tender deadline and the project implementation time, the score is relatively high; if it is only indirectly related or irrelevant, the score is relatively low. Similarly, a full score (such as 10 points) is set for scoring.

[0057] Risk assessment score R: Score by analyzing the possible risk level of the content involved in the keyword. The higher the risk, the higher the score; the lower the risk, the lower the score. The full score can be set to 10 points.

[0058] Weight description Weight distribution: weight w I 、w T 、w R The specific values of can be adjusted according to the characteristics and actual needs of the project. For example, if the project has very strict time requirements, then the weight w of the timeliness assessment can be appropriately increased T ; if the project has high technical difficulty and high risk, then the weight w of the risk assessment can be appropriately increased R .

[0059] For example: Suppose there is a tender for a software development project, and now it is necessary to adjust the priorities of several keywords, namely "project name", "tender deadline", and "system performance indicators".

[0060] After analysis, if it is considered that the importance, timeliness, and risk have the same impact on the project, so w I = 0.4, w T = 0.3, w R = 0.3.

[0061] Score the indicators for each keyword: Project name: Importance assessment score I: The project name is the core identifier of the project, which is crucial for the overall understanding of the project. It appears in an important position in the tender document, is closely related to the core content of the project, and can be scored 9 points.

[0062] Timeliness assessment score T: The project name has no direct relation to the time requirement and can be scored 2 points.

[0063] Risk assessment score R: The project name itself does not bring obvious risks and can be scored 1 point.

[0064] Tender deadline: Importance assessment score I: The tender deadline is a key piece of information in the tender process, but it is not as core as the project name for the overall understanding of the project and can be scored 7 points.

[0065] Timeliness assessment score T: It is directly related to the key time node of the tender and is very important, and can be scored 9 points.

[0066] Risk assessment score R: Missing the tender deadline will lead to the failure of the tender, with a relatively high risk, and can be scored 8 points.

[0067] System performance indicators: Importance assessment score I: The system performance indicators are key requirements for software projects and have a greater impact on the successful implementation of the project, and can be scored 8 points.

[0068] Timeliness assessment score T: It has no direct relation to the time requirement and can be scored 3 points.

[0069] Risk assessment score R: If the system performance indicators are not met, it may lead to the failure of project acceptance or unstable operation, with a relatively high risk, and can be scored 9 points.

[0070] Calculate the priority of each keyword Project name: P 项目名称 =0.4×9 + 0.3×2 + 0.3×1 = 3.6 + 0.6 + 0.3 = 4.5 Tender deadline: P 投标截止日期 =0.4×7 + 0.3×9 + 0.3×8 = 2.8 + 2.7 + 2.4 = 7.9 System performance indicators: P 系统性能指标 =0.4×8 + 0.3×3 + 0.3×9 = 3.2 + 0.9 + 2.7 = 6.8 During the tender preparation process, the tenderer issued a supplementary notice, raising the requirements for the system performance indicators. At this time, re-evaluate the score of the system performance indicators: Importance assessment score I: Due to the increased requirements, the system performance indicators have become more critical and are adjusted to 9 points.

[0071] Timeliness assessment score T: Still not directly related to the time requirements and is adjusted to 3 points.

[0072] Risk assessment score R: The risk of not meeting the new requirements is higher and is adjusted to 10 points.

[0073] Recalculate the priority of the system performance indicators: P 系统性能指标（调整后） = 0.4×9 + 0.3×3 + 0.3×10 = 3.6 + 0.9 + 3 = 7.5 The adjusted priority order is: Bid deadline (7.9) > System performance indicators (7.5) > Project name (4.5). The bidding team can arrange the work priorities and time allocation more reasonably according to this priority order.

[0074] Classify the requirements in the bidding documents according to the extracted bidding keywords. For example, classify the requirements related to the project name into one category, the requirements related to the bidding scope into another category, and so on, to form categories such as project name requirements, bidding scope requirements, technical requirement requirements, and commercial term requirements. For each major category, it can be further subdivided. For example, technical requirement requirements can be subdivided into performance indicator requirements, quality standard requirements, test and acceptance requirements, etc.; commercial term requirements can be subdivided into payment method requirements, delivery date requirements, after-sales service requirements, etc.

[0075] Accurately extract the requirement content: For the requirements under each classification, accurately extract the relevant content from the bidding documents and describe it clearly and accurately. For example, for the performance indicator requirements in the technical requirements, clarify the specific indicator values and units; for the payment method requirements in the commercial terms, elaborate on the time nodes, proportions, and methods of payment.

[0076] Organize the requirements in a unified format, such as using a table form to clearly list information such as requirement category, requirement description, priority, etc. For example, create a table with columns such as "Requirement Category", "Requirement Description", "Priority", etc., and fill in the sorted requirements item by item in the table.

[0077] S300: Obtain the bidding keywords, identify the corresponding enterprise keywords and their corresponding current data, dynamically generate the tender content, and output.

[0078] In this embodiment, obtain the bidding keywords, identify the corresponding enterprise keywords and their corresponding current data, dynamically generate the tender content, and output. The specific process is as follows: S301: Obtain tender keywords, and use keyword matching algorithms to match the tender keywords with the enterprise keywords in the enterprise knowledge base; S302: Dynamically generate tender content according to the matched enterprise data and tender requirements; S303: Review and proofread the tender, and output the tender content.

[0079] In the above process: Standardize the obtained tender keywords, such as unifying case, removing special characters, etc., for more accurate matching later. Extract keywords related to the enterprise's business and advantages from the enterprise knowledge base, and these keywords should be able to accurately reflect the enterprise's core competitiveness. For example, for a software development enterprise, the enterprise keywords may include "software development", "system integration", "artificial intelligence algorithm", "big data analysis", etc.

[0080] The keyword matching algorithms include string-based matching, semantic-based matching, and hybrid matching; among them: Algorithms based on string matching: Such as the simple string matching algorithm (determine whether there is a match by comparing characters one by one), the KMP algorithm (an efficient string matching algorithm that reduces unnecessary comparisons by preprocessing the pattern string), etc. These algorithms can directly compare the strings of tender keywords and enterprise keywords to judge whether there is a matching relationship.

[0081] Algorithms based on semantic matching: Considering that tender keywords and enterprise keywords may have differences in expression but the same semantics, semantic matching algorithms can be used. For example, use word vector models (such as Word2Vec, GloVe, etc.) to convert tender keywords and enterprise keywords into vector representations, and then calculate the similarity between vectors (such as cosine similarity), and judge whether there is a match according to the similarity threshold.

[0082] Hybrid matching algorithm: Combine the advantages of string matching and semantic matching algorithms. First, perform string matching to quickly screen out possible keyword pairs, and then perform semantic matching on these keyword pairs to improve the accuracy and efficiency of matching.

[0083] Tender requirement classification: Further classify tender requirements. For example, classify technical requirements into functional requirements, performance requirements, security requirements, etc.; classify commercial terms into price terms, payment terms, delivery terms, etc.

[0084] Determination of requirement weights: Determine the weights of different requirements according to the key points in the tender document and the enterprise's advantages. For example, if the tender document has higher requirements for technical performance, the weight of technical requirements should be increased accordingly.

[0085] Data screening: Based on the matched enterprise keywords, screen out enterprise data related to the bidding requirements from the enterprise knowledge base. For example, if specific technical indicators are mentioned in the bidding requirements, then screen out the relevant data in the enterprise products that meet these technical indicators.

[0086] Data integration: Integrate the screened enterprise data to form content that meets the bidding requirements. For example, organically combine information such as the product advantages, successful cases, and qualification certificates of the enterprise with the technical requirements and business terms in the bidding requirements to highlight the competitiveness of the enterprise.

[0087] According to the type and characteristics of the bidding project, select a suitable tender document template. The tender document template should include basic structures such as a cover page, table of contents, and main text, as well as common chapters such as technical solutions, business quotations, and enterprise qualifications.

[0088] Fill the integrated enterprise data into the corresponding positions in the tender document template. For example, in the technical solution chapter, describe in detail how the enterprise's products or services meet the technical requirements in the bidding requirements; in the business quotation chapter, give a reasonable quotation according to the bidding requirements and the enterprise's cost situation.

[0089] According to the special requirements of the bidding project and the actual situation of the enterprise, make personalized adjustments to the generated tender document content. For example, if the bidding document requires providing specific supporting materials, then add these materials in the corresponding positions in the tender document; if the enterprise has unique advantages in certain aspects, then highlight these advantages in the tender document.

[0090] Review and proofread the tender document. Have professional personnel within the enterprise, such as technical experts, business personnel, and legal personnel, review the generated tender document. The review content includes whether the tender document content meets the bidding requirements, whether the technical solution is feasible, whether the business quotation is reasonable, whether the enterprise qualifications are true and valid, and whether the tender document format is standardized.

[0091] Carefully proofread the text in the tender document to check for typos, grammar errors, improper use of punctuation marks, etc.

[0092] Proofread the data in the tender document to ensure the accuracy and consistency of the data. For example, check whether data such as technical indicators, quotation amounts, and dates are correct.

[0093] Check whether the format of the tender document meets the requirements, such as whether the font, font size, line spacing, page margins, etc. are unified, whether the title levels are clear, and whether the chart numbers are correct.

[0094] According to the requirements of the bidding document, convert the reviewed and proofread tender document into the specified format. For example, if the bidding document requires submitting a tender document in PDF format, then convert the Word format tender document into PDF format.

[0095] Meanwhile, to protect the security of the tender document content, the tender document is encrypted. At the same time, the tender document and other relevant materials (such as power of attorney, copies of enterprise qualifications, etc.) are packaged to form a complete tender document package.

[0096] Corresponding to the embodiments of the tender document generation method based on the large model described above, the present application also provides embodiments of a tender document generation system based on the large model.

[0097] Figure 5 is a block diagram of a bid generation system based on a large model shown according to an exemplary embodiment. Refer to Figure 5 , the system may include: an enterprise keyword extraction module 401, a tender keyword extraction module 402, and a bid dynamic generation module 403; where: The enterprise keyword extraction module 401 is used to obtain the current enterprise data, extract enterprise keywords for each piece of current data, and establish an enterprise knowledge base; where the enterprise knowledge base includes historical project cases, qualification documents, and technical parameters; The tender keyword extraction module 402 is used to obtain the tender documents, extract tender keywords, and generate a structured requirement list; The bid dynamic generation module 403 is used to obtain the tender keywords, identify the corresponding enterprise keywords and their corresponding current data, dynamically generate the bid content, and output it.

[0098] In this embodiment, the enterprise keyword extraction module 401 obtains the current enterprise data, extracts enterprise keywords for each piece of current data, and establishes an enterprise knowledge base; where the enterprise knowledge base includes historical project cases, qualification documents, and technical parameters; the tender keyword extraction module 402 obtains the tender documents, extracts tender keywords, and generates a structured requirement list; the bid dynamic generation module 403 obtains the tender keywords, identifies the corresponding enterprise keywords and their corresponding current data, dynamically generates the bid content, and outputs it; by generating a structured requirement list for the tender keywords, performing enterprise keyword identification, and generating a bid, it is possible to dynamically generate the bid content according to the requirements of the tender documents; avoiding the situation where the enterprise is in a passive position and misses business opportunities due to its inability to meet the requirements of efficient and accurate response to tender needs.

[0099] Regarding the system in the above embodiment, the specific ways in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0100] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial descriptions of the method embodiments. 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 application. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0101] Correspondingly, this application also provides an electronic device, including: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the method for generating a tender document based on a large model as described above. As Figure 6 shown, it is a hardware structure diagram of a device with any data processing capability where the tender document generation system based on a large model provided by an embodiment of the present invention is located. Except for Figure 6 the processors, memory, and network interfaces shown, any device with data processing capability where the device in the embodiment is located usually includes other hardware according to the actual functions of the device with any data processing capability, which will not be elaborated here.

[0102] Correspondingly, this application also provides a computer-readable storage medium, on which computer instructions are stored, and when the instructions are executed by a processor, the method for generating a tender document based on a large model as described above is implemented. The computer-readable storage medium may be an internal storage unit of any device with data processing capability described in any of the foregoing embodiments, such as a hard disk or a memory. The computer-readable storage medium may also be an external storage device, such as a plug-in hard disk, a Smart Media Card (SMC), an SD card, a Flash Card, etc. equipped on the device. Further, the computer-readable storage medium may also include both the internal storage unit of any device with data processing capability and the external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capability, and may also be used to temporarily store the data that has been output or will be output.

[0103] Those skilled in the art will readily think of other implementation schemes of this application after considering the specification and practicing the content disclosed herein. This application aims to cover any variations, uses, or adaptive changes of this application, and these variations, uses, or adaptive changes follow the general principles of this application and include the common general knowledge or conventional technical means in the technical field not disclosed in this application.

[0104] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A tender document generation method based on a large model, characterized in that, It includes the following steps: Obtain the current data of the enterprise, extract enterprise keywords for each piece of current data, and establish an enterprise knowledge base; Obtain the tender documents, extract tender keywords, and generate a structured requirement list; Obtain the tender keywords, identify the corresponding enterprise keywords and their corresponding current data, dynamically generate the tender content, and output it.

2. The method for generating tender documents based on a large model according to claim 1, wherein In the step of obtaining the current data of the enterprise, extracting enterprise keywords for each piece of current data, and establishing an enterprise knowledge base: Collect the current data of the enterprise and perform integrity and accuracy detection on the current data; Among them, the enterprise knowledge base includes historical project cases, qualification documents, and technical parameters; Use a large model to extract enterprise keywords for each piece of collected current data; Associate the extracted keywords with the corresponding data and store them in the database to form an enterprise knowledge base.

3. The method for generating a tender document based on a large model according to claim 2, wherein, In the step of collecting the current data of the enterprise and performing integrity and accuracy detection on the current data, the process of integrity detection is: Check the fields included in each piece of data and output the detection results.

4. The method for generating a tender document based on a large model according to claim 2, wherein In the step of collecting the current data of the enterprise and performing integrity and accuracy detection on the current data, the process of accuracy detection is: For numerical data, check whether the data is within a reasonable range; For text data, check whether there are grammar, logic, and contradictory information, and output the detection results.

5. The method for generating tender documents based on a large model according to claim 2, wherein, In the step of using a large model to extract enterprise keywords for each piece of collected current data: For historical project cases, enterprise keywords include project name, project type, key technologies, and project achievements; for qualification documents, enterprise keywords include qualification name, issuing agency, and validity period; For technical parameters, enterprise keywords include product model, performance indicators, and technical specifications.

6. The method for generating a tender document based on a large model according to claim 1, wherein, In the step of obtaining the tender documents, extracting tender keywords, and generating a structured requirement list: Obtain the tender documents and use a large model to extract tender keywords from the tender documents; among them, the tender keywords include project name, tender scope, technical requirements, and commercial terms; According to the extracted tender keywords, structurally organize the requirements in the tender documents to form a requirement list.

7. The method for generating tender documents based on a large model according to claim 6, wherein, Before the step of structurally organizing the requirements in the tender documents according to the extracted tender keywords to form a requirement list: Dynamically adjust the priority of the tender keywords.

8. The method for generating tender documents based on a large model according to claim 1, wherein In the step of obtaining the tender keywords, identifying the corresponding enterprise keywords and their corresponding current data, dynamically generating the tender content, and outputting it: Obtain the tender keywords and use a keyword matching algorithm to match the tender keywords with the enterprise keywords in the enterprise knowledge base; Dynamically generate the tender content according to the matched enterprise data and tender requirements; Review and proofread the tender and output the tender content.

9. The method for generating a tender document based on a large model according to claim 8, wherein, After the step of dynamically generating the tender content according to the matched enterprise data and tender requirements: Review and proofread the tender and output the tender content.

10. A tender document generation system based on a large model, which is applied to the tender document generation method based on a large model as described in claim 1, and is characterized in that, It includes an enterprise keyword extraction module, a tender keyword extraction module, and a tender dynamic generation module; among them: The enterprise keyword extraction module is used to obtain the current data of the enterprise, extract enterprise keywords for each piece of current data, and establish an enterprise knowledge base; the enterprise knowledge base includes historical project cases, qualification documents, and technical parameters; The tender keyword extraction module is used to obtain the tender documents, extract tender keywords, and generate a structured requirement list; The tender document dynamic generation module is used to obtain the tender keywords, identify the corresponding enterprise keywords and their corresponding current data, dynamically generate the tender document content, and output it.

Images