Bidding document intelligent auditing method based on block chain

Abstract

The invention discloses an intelligent bidding document auditing method based on a block chain, and belongs to the technical field of data processing, and the method specifically comprises the following steps: analyzing a technical clause dependency relationship by means of a natural language processing technology, and generating a semantic link; establishing a mathematical constraint network and a dynamic verification condition for the parameter data; extracting spatio-temporal information from the qualification file, and dynamically binding the spatio-temporal information with project geographic coordinates; then, scanning a semantic link breakpoint matching missing pattern library, positioning a parameter value domain conflict range, verifying qualification file validity and a space-time matching degree, and obtaining a file defect marked with a risk level; secondly, extracting peripheral semantic features of defect nodes, screening candidate terms from a knowledge base, carrying out progressive transformation, retaining a framework, injecting quantitative indexes, and adding territory restriction clauses; and finally, simulating different correction combinations in a virtual environment, comprehensively evaluating technical feasibility, cost fluctuation and risk probability, and screening out an optimal correction scheme.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to an intelligent audit method for tender documents based on blockchain. Background Art

[0002] In the current wave of digitalization, big data technology has become a key force driving the transformation and upgrading of various industries. The field of bidding and tendering has also accumulated a vast amount of document data, covering technical details, parameter information, qualification certificates, etc. Properly processing and analyzing this data is crucial for enterprises to control costs and stand out in the fierce competition.

[0003] Currently, the defect audit of tender documents relies to a large extent on manual operations. Auditors need to rely on professional knowledge to study technical terms word by word, manually check various parameter data, and carefully verify the authenticity and compliance of qualification documents.

[0004] However, this traditional manual audit mode exposes great drawbacks, requiring high labor costs and time costs. Moreover, due to the limited speed of manual audits, it is difficult to complete a comprehensive and detailed audit in a short time when faced with urgent tender projects. In addition, manual audits are extremely vulnerable to subjective factors and it is difficult to ensure the accuracy and consistency of audit results. Therefore, an optimized intelligent audit method for tender documents is needed as an aid to tender document audits. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent audit method for tender documents based on blockchain to solve the following technical problems:

[0006] The traditional manual audit mode requires high labor costs and time costs, and manual audits are extremely vulnerable to subjective factors.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] An intelligent audit method for tender documents based on blockchain includes the following steps:

[0009] Obtain and analyze the tender documents uploaded by the user, perform dependency parsing on the technical terms through natural language processing technology, identify the logical dependency relationships between the terms and generate weighted semantic links; establish a mathematical constraint network for the parameter data, and generate dynamic verification conditions by analyzing the functional relationships between the values; extract spatio-temporal information from the qualification documents and dynamically bind the spatio-temporal information to the geographical information coordinates of the current project;

[0010] Scan the breakpoints of the semantic link and match the missing pattern library; according to the established constraint equations, locate the range with conflicts in the parameter value range; verify the validity of the qualification documents, and at the same time check the matching degree between the qualification documents and the project spatio-temporal coordinates to obtain the document defects marked with risk levels;

[0011] For the said document defects, extract the semantic features within two levels around the defect nodes, including the topic keywords of the associated clauses, the fluctuation range of the reference parameters, and the regional policy constraint conditions;

[0012] According to the said semantic features, screen the candidate clauses in the knowledge base with a semantic matching degree exceeding the set threshold, and progressively transform the candidate clauses. First, retain the basic framework of the candidate clauses, then inject the quantitative indicators matching the current parameters, and finally append the restrictive clauses conforming to the regional characteristics of the project;

[0013] Based on the transformed candidate clauses, amend the tender documents, synchronously simulate the amendment combinations of different candidate clauses and different amendment directions in a virtual test environment, and screen out the optimal amendment plan by comprehensively evaluating the technical feasibility, cost fluctuation trend, and risk probability of different amendment combinations.

[0014] As a further solution of the present invention: the specific analysis process of the technical clauses, parameter data, and qualification documents is as follows:

[0015] When performing dependency syntactic analysis on the technical clause text, first identify the subject-predicate-object structure to form the logical backbone, extract the conditional adverbial, temporal adverbial, and reference documents as constraint branches, and generate technical clause nodes; when establishing a dynamic mathematical constraint network for the parameter table data, construct a difference equation for the linear relationship and establish a fuzzy equation set with a tolerance interval for the non-linear relationship to generate parameter nodes, and add historical fluctuation range labels to each parameter node; when performing optical character recognition on the qualification document image, use adaptive threshold segmentation technology to locate the key information area, correct the distorted text through perspective transformation, and establish a dynamic association between the extracted issuing agency, validity period information and the project location coordinates and time axis in the geographic information system to form a verifiable spatio-temporal validity label to obtain qualification nodes; a two-way verification channel is established between the parameter nodes and the technical clause nodes, and the qualification nodes are dynamically bound to the spatio-temporal coordinate axes to form a three-dimensional network structure.

[0016] As a further solution of the present invention: the detection process of the said document defects is as follows:

[0017] For the technical clauses, analyze the interruption position of the logical backbone, locate the associated nodes upstream and downstream of the missing points through semantic trajectory tracking technology, match the high-frequency completion pattern library in the knowledge base, and at the same time verify whether the integrity threshold of the constraint branches meets the standard;

[0018] For parameter data, the interval propagation algorithm is used to calculate the conflict value range, construct a parameter influence conduction path model, display the associated influence range of conflicting parameters in the form of a heat map, and mark the key conflict areas through dynamic coloring;

[0019] For qualification documents, connect to the electronic certificate library for real-time online verification, extract the issuing agency, validity period, and permitted scope of the qualification documents, calculate the overlapping ratio of the qualification validity period and the project cycle, generate a timeliness warning mark for documents with an overlapping ratio lower than the legal requirements, and highlight them with a flashing icon in the three-dimensional defect map.

[0020] As a further solution of the present invention: The progressive transformation specifically includes:

[0021] Retain the core terms and logical structure of the candidate clauses, delete the redundant descriptions irrelevant to the current project, and adjust the expression order of the clauses according to the semantic environment of the defect nodes;

[0022] Generate numerical values with tolerance intervals according to the fluctuation range of the parameter constraint network, add dynamic calculation formula annotations to the key parameters, and the numerical adjustment range of the key parameters meets the requirements of the technical specifications;

[0023] Retrieve the latest policy documents of the project location, analyze the influence weight of regional characteristics on the technical solution, generate constraint conditions including keywords such as environmental protection requirements, construction time limits, and regional preferences for material procurement, and seamlessly embed the new sub-clauses into the original clause framework through semantic fusion technology to maintain the logical coherence of the text.

[0024] As a further solution of the present invention: The process of amending the tender document based on the transformed candidate clauses is as follows:

[0025] Split the amendment process into independent verification units. The independent verification unit includes a technical feasibility evaluation module, a cost impact calculation module, and a legal risk analysis module; When constructing a parallel test scenario in a virtual environment, load the complete project background data for each scenario, including construction environment parameters, market price fluctuation curves, and legal norm versions; When multiple amendment combinations are run synchronously, the technical feasibility evaluation module simulates the impact of parameter adjustment during construction on the technological process, the cost impact calculation module tracks the chain fluctuations of material, labor, and management costs, and the legal risk analysis module scans the compatibility between clause changes and the latest regulations; After simulating and generating the technical parameter fluctuation cloud map, cost change curve, and risk probability distribution map, calculate the comprehensive scores of each plan through a weighted scoring algorithm, and finally select the amendment combination with the highest score and generate an execution instruction.

[0026] As a further solution of the present invention: When the user amends the tender document according to the amendment plan, an irreversible amendment track record is generated and the associated weight of the knowledge base is updated.

[0027] As a further solution of the present invention, the process of updating the association weights of the knowledge base is as follows:

[0028] Convert the successfully corrected cases selected by the user into structured feature vectors, including four dimensions: project type, defect features, correction solutions, and implementation effects; compare the new cases with the knowledge base nodes through a similarity matching algorithm, enhance the association weights of the cases with a matching degree higher than the threshold, and improve their priorities in subsequent complementary recommendations; conduct feature analysis on the correction patterns that cause secondary problems, extract the key factors leading to the problems, and reduce their association weights; establish a fast response link for high-frequency correction patterns. When detecting defects with a similarity exceeding the set value to historical cases, automatically skip the regular retrieval process, directly call the verified effective solutions, and inject the current project feature parameters.

[0029] As a further solution of the present invention: when multiple users modify the same tender document simultaneously, first mark the influence scope of each correction operation, record the modification time, operator identity, and change content; secondly, detect the logical conflicts in the overlapping areas, analyze the contradictions in technical parameters, cost calculation conflicts, and differences in legal bases of the conflicting clauses; then retrieve the historical arbitration case library, match the conflict resolution path with the highest similarity to generate a recommendation list, and the recommendations include compromise solutions, parameter compromise values, and legal interpretation instructions; finally, record the adopted decision-making solutions and update the conflict resolution rule library, establish a priority response strategy for high-frequency conflict patterns, and automatically recommend optimized solutions when the same conflict type is detected subsequently.

[0030] Advantages of the present invention:

[0031] The cost control method of the present invention based on big data analysis deeply analyzes the logical dependence relationship of technical clauses through natural language processing technology to generate semantic links, comprehensively establishes a mathematical constraint network of parameter data to accurately locate the conflict value range, dynamically binds the spatio-temporal information of qualification documents with the geographical information coordinates of the project to achieve multi-dimensional verification, can also extract semantic features to screen and gradually transform candidate clauses, and comprehensively evaluate and screen the optimal correction solutions in a virtual test environment. It can not only greatly improve the accuracy and depth of tender document analysis, but also efficiently and intelligently complement and correct document defects, effectively reduce costs and avoid risks, meet the cost control and risk avoidance needs in the enterprise tender process, and significantly improve the efficiency and quality of tender work. Brief Description of the Drawings

[0032] The present invention will be further described below with reference to the accompanying drawings.

[0033] Figure 1 It is a flow diagram of the present invention. Detailed Embodiments

[0034] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0035] Please refer to Figure 1 As shown, the present invention is an intelligent review method for tender documents based on blockchain, including the following steps:

[0036] After the user uploads the tender document, the system immediately starts the analysis process. With the help of advanced natural language processing technology, it deeply analyzes the technical terms and meticulously analyzes the dependency relationships between the terms. In this process, through complex algorithm models, it accurately identifies the complex logical dependencies between the terms, and then generates a semantic link with weights. The weights are assigned based on multiple factors such as the tightness of the association and importance between the terms, so as to intuitively present the strength of the logical relationship between the terms. For the parameter data part, a comprehensive and dynamic mathematical constraint network is fully constructed. Using professional mathematical analysis means, it deeply analyzes the potential functional relationships between the values, and thus generates highly targeted dynamic verification conditions. The dynamic verification conditions are not fixed, but will be adjusted in real time with the changes of the parameters and the actual situation of the project. For the qualification documents, specific information extraction technology is used to accurately capture the spatio-temporal information therein, including key information such as the issuance time of the document, the start and end time of the validity period, and the geographical scope of the business license, and these spatio-temporal information are skillfully and dynamically bound to the geographical information coordinates of the current project to comprehensively strengthen the infrastructure of the tender document analysis.

[0037] After the basic data processing is completed, it enters the key defect detection link. The system will accurately scan the semantic link generated in the early stage and keenly capture the break points therein. Once a break is found, it will quickly match with the internally constructed missing pattern library. The missing pattern library is summarized based on a large amount of historical tender document data and industry experience, covering various common logical missing situations. At the same time, according to the constraint equations established in the early stage, professional mathematical algorithms are used to accurately locate the range of conflicts in the parameter value domain, and carefully identify the value intervals where the parameters are contradictory and do not conform to the actual business logic. In terms of the verification of qualification documents, the system rigorously verifies their effectiveness, not only verifying the authenticity of the documents, but also synchronously checking the matching degree between the qualification documents and the project spatio-temporal coordinates. For example, it judges whether the validity period of the qualification document covers the project execution period, whether the permitted scope matches the project location and business type, etc., and finally generates a document defect report with the risk level clearly marked, intuitively showing various problems and their severity in the tender document.

[0038] In the face of the detected document defects, the present invention adopts a unique processing strategy. The system will intelligently extract the semantic features within two levels around the defect nodes, deeply mine the subject keywords of the associated clauses, so as to clarify the business field and core points of the clause. At the same time, accurately obtain the fluctuation range of the reference parameters, and the determination of this fluctuation range comprehensively considers various factors such as historical data, industry standards, and actual project requirements. In addition, it will deeply analyze the regional policy constraints, and widely collect relevant policy regulations, industry norms, etc. of the project location.

[0039] Based on the extracted semantic features, the system efficiently screens from a vast knowledge base to find candidate clauses whose semantic matching degree exceeds a pre-set threshold. The knowledge base stores a large amount of sorted and classified clause information, covering common and special clauses in various industries and various projects. After screening out the candidate clauses, progressive transformation is carried out on them. First, carefully retain the basic framework of the candidate clauses to ensure the stability of the core logic and key architecture of the clauses. Then, skillfully inject quantitative indicators that are precisely matched with the current parameters, and these quantitative indicators are carefully calculated and set according to the actual parameter requirements of the project, industry standards, and cost control objectives. Finally, deeply investigate the regional characteristics of the project and append restrictive sub-clauses that conform to the local actual situation, such as considering local environmental protection policies, labor regulations, etc., to make the clauses more in line with the actual project scenario.

[0040] After completing the transformation of the candidate clauses, the tender documents are amended based on these transformed clauses. To ensure the scientificity and reliability of the amendment plan, the present invention constructs a virtual test environment. In this environment, multiple amendment combinations of different candidate clauses and different amendment directions are synchronously simulated. Through a series of professional evaluation modules, comprehensively evaluate the technical feasibility of different amendment combinations, simulate the actual impact of parameter adjustment and technical clause change on the overall technological process during construction; closely track the cost fluctuation trend, and carefully calculate the chain fluctuations of materials, labor, and management costs, etc. caused by the amendment; comprehensively scan the legal risk probability, and rigorously verify the compatibility of clause changes with the latest laws and regulations. After comprehensive simulation and deduction, generate an intuitive technical parameter fluctuation cloud map, a clear cost change curve, and an accurate risk probability distribution map. With the help of a professional weighted scoring algorithm, comprehensively consider various factors, accurately calculate the comprehensive scores of each plan, and finally screen out the optimal amendment plan with the highest score from numerous plans, providing scientific and reliable guidance for the improvement of the tender documents, and helping enterprises effectively control costs and accurately avoid risks during the tender process.

[0041] In a preferred embodiment of the present invention, the specific analysis process for technical clauses, parameter data, and qualification documents is as follows:

[0042] When performing dependency syntactic analysis on the text of technical terms, the system will apply advanced natural language processing algorithms. Relying on a deep understanding of the syntactic structure of sentences, it will first accurately identify the subject-predicate-object structure. This structure, like the skeleton of technical terms, forms the logical backbone, clearly presenting the core actions and objects of the terms. At the same time, the system will not miss any key information and will carefully extract conditional adverbials, which often limit the specific preconditions for the terms to come into effect or be executed. For example, conditions such as "if the annual precipitation in the project location exceeds 500 millimeters" have an important impact on the implementation of the technical solution; time adverbials will also be accurately captured, which clarify the applicability of the terms in the time dimension. Time limits like "within the first two years of project construction" are related to the time nodes of technical operations. In addition, cited documents, as an indispensable restrictive branch, will also be extracted. The standards, specifications, etc. contained in the cited documents provide detailed basis for the specific implementation of technical terms. Through such comprehensive analysis, technical term nodes are finally generated, and these nodes become the key units for constructing the technical logical framework of the bidding document.

[0043] When processing parameter table data, to establish a dynamic mathematical constraint network, the system demonstrates strong mathematical modeling capabilities. For linear relationships, difference equations will be constructed. For example, in a model describing the linear growth of project costs over time, the difference equation can accurately depict the changes in costs within each time interval, facilitating the dynamic monitoring and prediction of costs. For non-linear relationships, fuzzy equation systems with tolerance intervals will be established to handle the complex and variable parameter relationships in reality that are difficult to describe with simple linear relationships, such as the complex associations between material performance parameters and environmental factors. After generating parameter nodes, to provide more abundant information for parameter analysis, the system will add historical fluctuation range labels to each parameter node. This label is obtained based on the statistical analysis of a large amount of past data from similar projects and can intuitively reflect the possible value ranges of the parameter under different circumstances, providing a strong reference for subsequent parameter conflict detection and solution optimization.

[0044] When faced with qualification document images, the present invention adopts cutting-edge optical character recognition technology. First, the adaptive threshold segmentation technology is used. This technology can automatically adjust the threshold according to the local features of the image, accurately locate the key information area, and avoid information omission caused by complex image backgrounds or differences in character clarity. After positioning, perspective transformation is used to correct distorted text to ensure the accuracy of character recognition. Subsequently, the extracted information such as the issuing agency and validity period will establish a close dynamic association with the project location coordinates and time axis in the geographic information system. For example, the validity period of the qualification is compared with the time axis of project execution to determine whether the qualification is valid within the project cycle; the geographical information of the issuing agency is combined with the project location coordinates to analyze whether the qualification meets the specific requirements of the project location, thereby forming a verifiable spatio-temporal validity label and finally obtaining a qualification node. It is worth mentioning that a two-way verification channel is established between the parameter node and the technical clause node, which means that the rationality of the parameters can be verified through the logic of the technical clauses, and vice versa, the feasibility of the technical clauses can also be tested through the parameter data. The dynamic binding of the qualification node to the spatio-temporal coordinate axis further improves the entire analysis system, forming an all-round and three-dimensional network structure.

[0045] In another preferred embodiment of the present invention, the process of detecting document defects is as follows:

[0046] For technical clauses, the system will deeply analyze the interruption position of the logical backbone, which requires a detailed combing and backtracking of the logical link. Through the semantic trajectory tracking technology, this technology can accurately locate the associated nodes upstream and downstream of the missing point along the semantic context, just like finding the missing link on the key path in a complex map. At the same time, the system will quickly match the high-frequency completion pattern library in the knowledge base. This pattern library is summarized based on a large amount of historical tender document data and industry expert experience, covering various common technical clause missing situations and corresponding completion methods. During the matching process, the integrity threshold of the constraint branch will also be verified synchronously to ensure that all key elements of the technical clause are complete and meet the requirements.

[0047] In terms of detecting parameter data defects, the interval propagation algorithm is used to calculate the conflict value range. This algorithm can accurately find the range of parameter value domains with conflicts in a complex parameter network by continuously propagating and updating the value domains of parameters. To more intuitively display the associated influence range of conflicting parameters, the system constructs a parameter influence conduction path model and visualizes it in the form of a heat map. In the heat map, the depth of color represents the strength of the parameter influence degree, enabling users to clearly see which parameters are strongly associated and the possible diffusion range of conflicts. And by dynamically coloring and marking the key conflict areas, the parameter conflict points that need the most attention and processing are further highlighted, providing clear guidance for subsequent parameter adjustment and scheme optimization.

[0048] For qualification documents, the present invention conducts real-time online verification by connecting to the electronic certificate library to ensure the accuracy and timeliness of the verification results. During the verification process, the system will extract core information such as the issuing authority, validity period and scope of permission of the qualification documents. Then, the overlap ratio between the validity period of the qualification and the project cycle is calculated. This ratio directly reflects the effectiveness of the qualification during the project execution. If the overlap ratio is lower than the statutory requirements, the system will immediately generate a timeliness warning mark to remind relevant personnel that there are potential risks in their qualifications. At the same time, in the three-dimensional defect map, these problematic qualification documents will be highlighted in the form of flashing icons, allowing users to quickly discover and pay attention to qualification defects among many documents.

[0049] In another preferred embodiment of the present invention, the incremental transformation comprises:

[0050] When processing candidate clauses, the system will first use its powerful semantic analysis capabilities to accurately identify and retain the core terms and logical structures in the candidate clauses. Core terms are like the cornerstone of clauses, carrying key business information, while logical structures determine the rationality and coherence of clauses. On the basis of retaining these key elements, the system will deeply compare the specific needs and characteristics of the current project, and keenly identify redundant descriptions that are not related to the project. For example, some candidate clauses may contain broad industry introductions or general but non-essential technical descriptions that are applicable to other project scenarios but not applicable to the current project. The system will decisively delete these redundant parts to streamline the content of the clauses and improve their fit with the current project. At the same time, the system will intelligently adjust the order of expression of the clauses based on the complex semantic environment around the defective node. The semantic environment includes information such as the technical field where the defective node is located, the relevant business processes, and the logical relationship between the context. By comprehensively considering these factors, the system can reorganize the content of the clauses so that the expression of the clauses is more in line with the logical thinking of the current project and easier to understand and apply.

[0051] For the parameter part, the system will generate values with a reasonable tolerance interval closely based on the fluctuation range presented by the parameter constraint network. The setting of the tolerance interval is not arbitrary. Instead, it fully considers various uncertainty factors that may occur during the actual implementation of the project, as well as the generally recognized error range standards in the industry. For example, in a construction project, the actual consumption of materials may vary within a certain range due to factors such as minor differences in construction techniques and quality fluctuations of the materials themselves. The tolerance interval can reasonably reflect such fluctuations. At the same time, for key parameters, the system will carefully add dynamic calculation formula annotations. These annotations not only clearly show the calculation logic of the parameters but also can adjust the calculation process in real time according to changes in the actual situation of the project. For example, in a project involving cost accounting, key parameters such as labor costs and material costs will change with fluctuations in market prices. The dynamic calculation formula annotations can automatically update the calculation results of the parameters based on the latest market price data to ensure that the parameter values are always accurate and meet the requirements of technical specifications.

[0052] In addition, the system will actively retrieve the latest policy documents of the project location and deeply analyze the influence weight of regional characteristics on the technical solution. There may be significant differences in environmental protection policies, construction time restrictions, and regional preferences for material procurement in different regions. For example, in some ecologically sensitive areas, environmental protection requirements are extremely strict, and may set extremely high standards for dust control and waste treatment during the construction process; some cities will strictly limit the construction time to only allow construction operations within specific time periods to relieve traffic pressure; and in some regions, due to the support of local industries, there will be certain regional preferences in material procurement. The system will sort out and integrate these key information to generate constraint conditions including keywords such as environmental protection requirements, construction time restrictions, and regional preferences for material procurement. Subsequently, through semantic fusion technology, these newly added clauses will be seamlessly embedded into the original clause framework.

[0053] In another preferred embodiment of the present invention, the process of amending the bidding document based on the modified candidate clauses is as follows:

[0054] First, the system will ingeniously split the entire correction process into multiple independent verification units. Each independent verification unit is like a fully functional small module, which includes a technical feasibility evaluation module, a cost impact calculation module, and a legal risk analysis module. When constructing parallel test scenarios in a virtual environment, the system will carefully load complete and detailed project background data for each scenario. These data cover construction environment parameters, such as the topography and climate conditions of the construction site, which will directly affect the selection of construction techniques and the difficulty of implementation; the market price fluctuation curve, which reflects the price change trends of various cost elements such as materials and labor in real time and is crucial for cost control and budget formulation; and the latest version of legal regulations to ensure that the correction plan of the tender document fully complies with the requirements of current laws and regulations.

[0055] When multiple correction combinations are running synchronously, each module performs its own functions. The technical feasibility evaluation module will give full play to its simulation ability and meticulously simulate the impact of parameter adjustments during the construction process on the technological process. For example, in a mechanical manufacturing project, if the size parameters of a key component are adjusted, this module can simulate the possible changes brought about by this adjustment to the entire production and assembly process, including whether production equipment needs to be replaced and the order of the technological process adjusted. The cost impact calculation module will accurately track the chain fluctuations in materials, labor, and management costs caused by the correction. It will comprehensively consider factors such as market price fluctuations, changes in man-hours due to changes in construction techniques, and changes in management costs brought about by adjustments in the management process, and comprehensively calculate the specific impact of each correction combination on costs. The legal risk analysis module will rigorously scan the compatibility of clause changes with the latest regulations and carefully check whether there are any potential risks of violating laws and regulations in the corrected tender document, such as whether it complies with the regulations on fair competition and information disclosure in the bidding law and whether it follows the relevant local regulations on project construction.

[0056] After completing a comprehensive simulation and deduction, the system will generate an intuitive and highly valuable technical parameter fluctuation cloud map, a clear cost change curve, and an accurate risk probability distribution map. The technical parameter fluctuation cloud map can vividly display the fluctuation range and change trend of each technical parameter under different correction combinations; the cost change curve visually presents in a chart form the increase and decrease of the cost with the change of the correction plan; the risk probability distribution map can clearly mark the occurrence probability of various risks such as legal risks and technical risks that each correction combination may face. Finally, the system will use a carefully designed weighted scoring algorithm to comprehensively consider various factors such as technical feasibility, cost fluctuation, and risk probability, and calculate a comprehensive score for each plan. The weighted scoring algorithm will assign corresponding weights to different evaluation factors according to the actual needs and key focus directions of the project to ensure that the evaluation results more meet the actual expectations of the project. Ultimately, the system will screen out the correction combination with the highest score and quickly generate an execution instruction.

[0057] In another preferred embodiment of the present invention, when the user successfully completes the correction of the tender document according to the correction plan carefully selected by the system, the system will immediately start a series of important subsequent operations. First, the system will generate an irreversible correction trajectory record, which is like a detailed log, completely and accurately recording the whole process of the tender document from the initial state to the final state after a series of correction steps, covering key information such as the specific content of each modification, the modification time, and the corresponding correction basis, providing solid and reliable data support for subsequent review analysis and audit work. At the same time, the system will update the association weights of the crucial knowledge base to continuously optimize the performance of the knowledge base so that it can better serve the subsequent tender document processing tasks.

[0058] In a preferred case of this embodiment, the process of updating the association weights of the knowledge base includes:

[0059] The first to skillfully transform the amendment cases selected and successfully applied by users into a structured feature vector, which is constructed from four key dimensions. The project type dimension will clearly and explicitly mark the industry field, business scope and other information of the bidding project, such as construction engineering, software development or equipment procurement projects. Different project types often have different characteristics and requirements, which is crucial for the accurate matching of the knowledge base. The defect feature dimension will meticulously describe the specific manifestations, locations and technical terms, parameter data or qualification documents involved in the defects in the bidding documents, such as logical loopholes in technical terms, conflicting value ranges of parameter data and timeliness issues of qualification documents. Accurately grasping the defect characteristics is the basis for effective correction. The correction plan dimension will record in detail the specific correction methods adopted for the defect, including the candidate terms selected from the knowledge base, the details of the gradual transformation of the terms and the final correction operation steps. This information fully presents the correction process. The implementation effect dimension will objectively provide feedback on the actual results achieved by the revised bidding documents in terms of technical feasibility, cost control and risk avoidance, such as whether the original defects have been successfully solved, whether the costs have been effectively controlled within a reasonable range, and whether the risk probability has been significantly reduced.

[0060] After completing the construction of the structured feature vector, the system will use advanced similarity matching algorithms to carefully compare the newly generated case vector with the massive nodes already in the knowledge base. For those cases whose matching degree is higher than the pre-set threshold, the system will decisively take measures to enhance the associated weight. This means that in the subsequent bidding document completion recommendation process, the priority of these cases will be significantly improved, and they can be recommended to users more quickly and accurately by the system, providing users with more efficient and high-quality reference solutions. For those correction modes that have caused secondary problems in practice, in-depth feature analysis will be carried out on them, and through a series of complex data analysis methods and algorithm models, the key factors that cause secondary problems will be accurately extracted. For example, some correction solutions may cause compliance problems due to ignoring policy details in specific regions, or affect the stability of technical solutions due to excessive adjustment of parameters. Once the key factors are determined, the system will immediately reduce the associated weight of such correction modes in the knowledge base to avoid similar problems in subsequent recommendations.

[0061] In addition, to further improve the operating efficiency and response speed of the system, the system will also establish a fast response link for the high-frequency correction mode. Through in-depth mining and analysis of a large number of historical correction cases, the system can accurately identify those high-frequency correction modes that frequently appear in actual applications. When the system is in the subsequent detection process, once it finds that the defects in the tender documents are more similar to the high-frequency defects in the historical cases than the set value, it will automatically trigger the fast response mechanism and directly skip the conventional complex retrieval process. At this time, the system will quickly call the effective solutions that have been fully verified in past practices and perform personalized adaptation in combination with the unique characteristic parameters of the current project.

[0062] In another preferred embodiment of the present invention, if multiple users modify the same tender document simultaneously, the system will execute a series of operations in an orderly manner. First, the system uses marking technology to accurately record the correction operations of each user. The marked content covers the scope of operation influence, such as fine-tuning of technical terms, changes in cost accounting parameters, or modification of key information in qualification documents; at the same time, it records the modification time, the identity of the operator, and the changed content, providing data support for subsequent review.

[0063] Secondly, the system starts the conflict detection mechanism, focusing on checking the logical conflicts in the overlapping areas of the correction operations of different users. Deeply analyze the contradictions in the conflict terms in terms of technical parameters, cost calculation, and legal basis. For example, in a construction tender document, there may be conflicts between the concrete pouring speed and the mix ratio adjustment, or the material budget adjustment is not reflected in the cost summary, and inconsistent references to different versions of bidding and tendering regulations. Then, once a conflict is detected, the system quickly retrieves the historical arbitration case library, and through the similarity matching algorithm, accurately matches similar cases and generates a list of suggestions for conflict resolution paths. The suggestions include compromise solutions, reasonable parameter compromise values, and legal interpretation descriptions. Finally, after the user adopts the decision-making solution, the system records the content of the solution and updates the conflict resolution rule library. For high-frequency conflict modes, the system establishes a priority response strategy. When the same conflict type is detected later, it automatically recommends optimized solutions to improve the conflict resolution efficiency when multiple users modify tender documents.

[0064] The above has described in detail an embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equal changes and improvements made according to the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. An intelligent review method for tender documents based on blockchain, characterized in that, It includes the following steps: Obtain and analyze the tender documents uploaded by users. Use natural language processing technology to perform dependency parsing on technical terms, identify the logical dependencies between terms, and generate weighted semantic links; establish a mathematical constraint network for parameter data, and generate dynamic verification conditions by analyzing the functional relationships between values; extract spatio-temporal information from qualification documents, and dynamically bind the spatio-temporal information to the geographical information coordinates of the current project; Scan the breakpoints of the semantic links and match the missing pattern library; according to the established constraint equations, locate the range of conflicts in the parameter value domain; verify the validity of the qualification documents, and at the same time check the matching degree between the qualification documents and the project spatio-temporal coordinates to obtain file defects marked with risk levels; For the file defects, extract the semantic features within two levels around the defect nodes, including the topic keywords of related clauses, the fluctuation range of referenced parameters, and the regional policy constraint conditions; According to the semantic features, screen candidate clauses in the knowledge base with a semantic matching degree exceeding the set threshold, and progressively transform the candidate clauses. First, retain the basic framework of the candidate clauses, then inject quantization indicators matching the current parameters, and finally append restrictive clauses conforming to the regional characteristics of the project; Based on the transformed candidate clauses, correct the tender documents, synchronously simulate the correction combinations of different candidate clauses and different correction directions in a virtual test environment, and screen out the optimal correction plan by comprehensively evaluating the technical feasibility, cost fluctuation trend, and risk probability of different correction combinations.

2. The intelligent review method for tender documents based on blockchain according to claim 1, wherein The specific analysis processes for technical terms, parameter data, and qualification documents are as follows: When performing dependency syntax analysis on technical term texts, first identify the subject-predicate-object structure to form the logical backbone, extract conditional adverbials, time adverbials, and referenced documents as constraint branches to generate technical term nodes; when establishing a dynamic mathematical constraint network for parameter table data, construct difference equations for linear relationships and establish fuzzy equation systems with tolerance intervals for non-linear relationships to generate parameter nodes, and add historical fluctuation range labels to each parameter node; when performing optical character recognition on qualification document images, use adaptive threshold segmentation technology to locate key information areas, correct distorted texts through perspective transformation, and establish dynamic associations between the extracted issuing agency, validity period information and the project location coordinates and time axis in the geographical information system to form verifiable spatio-temporal validity labels to obtain qualification nodes; establish a two-way verification channel between the parameter nodes and the technical term nodes, and dynamically bind the qualification nodes to the spatio-temporal coordinate axes to form a three-dimensional network structure.

3. The intelligent review method for tender documents based on blockchain according to claim 1, wherein The detection process of the file defects is as follows: For technical terms, analyze the interruption position of the logical backbone, locate the associated nodes upstream and downstream of the missing points through semantic trajectory tracking technology, match the high-frequency completion pattern library in the knowledge base, and at the same time verify whether the integrity threshold of the constraint branches meets the standard; For parameter data, use the interval propagation algorithm to calculate the conflict value range, construct a parameter impact conduction path model, display the associated impact range of conflicting parameters in the form of a heat map, and mark the key conflict areas through dynamic coloring; For qualification documents, real-time online verification is carried out by connecting to the electronic certificate database, extracting the issuing authority, validity period and license scope of the qualification documents, calculating the overlap ratio between the validity period of the qualification and the project cycle, and generating timeliness warning marks for documents with an overlap ratio lower than the statutory requirements. These are highlighted with flashing icons in the three-dimensional defect map.

4. The intelligent review method for bidding documents based on blockchain according to claim 1, characterized in that, The progressive transformation specifically includes: Keep the core terms and logical structure of the candidate clauses, delete redundant descriptions that are not relevant to the current project, and adjust the expression order of the clauses according to the semantic environment of the defective node; Generate values with tolerance intervals based on the fluctuation range of the parameter constraint network, add dynamic calculation formula annotations to key parameters, and ensure that the adjustment range of key parameter values meets the requirements of technical specifications; Retrieve the latest policy documents of the project location, analyze the impact of regional characteristics on the technical solution, generate constraints containing keywords such as environmental protection requirements, construction time restrictions, and regional preferences for material procurement, and use semantic fusion technology to seamlessly embed new clauses into the original clause framework to maintain the logical coherence of the text.

5. The intelligent review method for tender documents based on blockchain according to claim 1, characterized in that, The process of revising the bidding documents based on the modified candidate clauses is as follows: The correction process is divided into independent verification units, which include a technical feasibility assessment module, a cost impact calculation module, and a legal risk analysis module. When building parallel test scenarios in a virtual environment, complete project background data is loaded for each scenario, including construction environment parameters, market price fluctuation curves, and legal specification versions. When running multiple correction combinations simultaneously, the technical feasibility assessment module simulates the impact of parameter adjustments on the process flow during the construction process, the cost impact calculation module tracks the chain fluctuations of materials, labor, and management costs, and the legal risk analysis module scans the compatibility of clause changes with the latest regulations. After the simulation and deduction generate the technical parameter fluctuation cloud map, cost change curve and risk probability distribution map, the comprehensive score of each plan is calculated through the weighted scoring algorithm, and finally the correction combination with the highest score is selected and the execution instructions are generated.

6. The intelligent review method for tender documents based on blockchain according to claim 1, characterized in that, When the user amends the bidding document according to the amendment plan, an irreversible amendment track record is generated and the associated weight of the knowledge base is updated.

7. The intelligent review method for tender documents based on blockchain according to claim 6, characterized in that The process of updating the association weight of the knowledge base is: The successful amendment cases selected by the user are converted into structured feature vectors, including four dimensions: project type, defect characteristics, amendment plan, and implementation effect. The new cases are compared with the knowledge base nodes through the similarity matching algorithm, and the cases with matching degrees higher than the threshold are enhanced in association weight, so as to increase their priority in subsequent completion recommendations. Perform feature analysis on the correction patterns that cause secondary problems, extract the key factors that cause the problems and reduce their associated weights; establish a rapid response link for high-frequency correction patterns. When a defect is detected whose similarity with historical cases exceeds a set value, the regular retrieval process is automatically skipped, and the verified effective solution is directly called and the current project feature parameters are injected.

8. The intelligent review method for tender documents based on blockchain according to claim 1, wherein When multiple users modify the same bidding document simultaneously, first mark the scope of influence of each correction operation, record the modification time, the identity of the operator, and the changed content; secondly, detect logical conflicts in the overlapping areas, and analyze the contradictions in technical parameters, cost calculation conflicts, and differences in legal bases of the conflicting clauses; Then retrieve the historical arbitration case library, match the conflict resolution path with the highest similarity to generate a list of suggestions, which include compromise solutions, parameter compromise values, and legal interpretation notes; finally, record the adopted decision-making plan and update the conflict resolution rule library, establish a priority response strategy for high-frequency conflict patterns, and automatically recommend optimized solutions when the same conflict type is detected in the subsequent detection.

Images