Engineering bidding and tendering platform management method and system

Abstract

The invention belongs to the field of platform management, and particularly discloses an engineering bidding and tendering platform management method and system, and the method comprises the steps: storing and obtaining a demand description text from a bidding project chain through a block chain node, extracting a keyword sequence through natural language processing, and obtaining a bidding demand feature vector; according to the bid invitation demand feature vector, performing vector conversion processing on qualification certificate data submitted by a bidding enterprise, judging the similarity between the converted vector and the demand feature vector, and if the similarity exceeds a preset threshold, determining preliminary qualification matching; extracting performance index data from historical performance records on the bidding enterprise chain, and fusing the performance index data with the preliminary qualification matching result by adopting a vector embedding method to obtain a comprehensive qualification vector; the objective of the invention is to solve the problems of low qualification examination efficiency, insufficient data credibility and non-transparent business scenes in the verification process in the traditional bidding process.

Description

Technical Field

[0001] This invention relates to the field of platform management technology, specifically to a method and system for managing an engineering bidding platform. Background Technology

[0002] In blockchain-based bidding scenarios, the core technical challenge lies in how to achieve dynamic and accurate matching between bidding company qualifications and bidding requirements in a decentralized environment, while ensuring the transparency and immutability of the verification process. This issue involves efficiently extracting the bidding requirement text from on-chain storage and converting it into feature vectors. However, natural language processing may face semantic ambiguity or inaccurate keyword extraction when processing complex bidding texts, resulting in feature vectors that cannot accurately reflect the characteristics of the requirements and affecting the accuracy of subsequent matching.

[0003] Furthermore, when converting the qualification certificates and historical performance data of bidding companies into vectors and performing similarity calculations with the demand vectors, heterogeneous data formats and inconsistent indicator systems may lead to vector space mismatch, reducing matching accuracy.

[0004] Furthermore, when calling external oracles to obtain real-time financial statements, there is a verification issue between the data update timestamp and the bid deadline. If the time window is too short or network delays cause the data to not be updated in time, the validity of the financial data may be misjudged, affecting the reliability of the comprehensive qualification assessment.

[0005] Ultimately, when executing dynamic verification score calculation and automatic approval logic through smart contracts, the weighted fusion algorithm may fail to accurately reflect the enterprise's capabilities due to unreasonable weight settings or data noise interference. Furthermore, when retrieving verification paths from audit logs stored on the chain, if node synchronization delays or on-chain data fragmentation occur, the paths may be incomplete, affecting the credibility of the review results.

[0006] This complex issue spans the entire process from data extraction and vector transformation to dynamic verification, involving complex interactions of semantic parsing, data fusion, real-time verification, and on-chain storage, directly impacting the fairness and efficiency of bidding and tendering. Summary of the Invention

[0007] This invention provides a management method and system for an engineering bidding platform, aiming to solve the problems of low efficiency in qualification review, insufficient data credibility, and opaque verification process in traditional bidding processes.

[0008] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for managing an engineering bidding platform includes: obtaining a requirement description text from the on-chain storage of a bidding project via a blockchain node; extracting keyword sequences using natural language processing to obtain a bidding requirement feature vector; performing vector transformation processing on the qualification certificate data submitted by the bidding company based on the bidding requirement feature vector; determining the similarity between the transformed vector and the requirement feature vector; if the similarity exceeds a preset threshold, determining a preliminary qualification match; extracting performance indicator data from the historical performance records on the bidding company's blockchain; fusing the performance indicator data with the preliminary qualification match result using a vector embedding method to obtain a comprehensive qualification vector; obtaining real-time financial statement updates of the bidding company through a smart contract by calling an external oracle interface; determining whether the financial statement update time is before the bidding deadline; if it is before the bidding deadline, determining the validity of the financial data; performing a weighted fusion calculation based on the comprehensive qualification vector and the validity result of the financial data to obtain a dynamic verification score for the bidding company; if the dynamic verification score is higher than a preset threshold, executing automatic approval logic through a smart contract; storing the verification process record on-chain using a hash function to obtain an immutable audit log; retrieving relevant verification paths from the blockchain network based on the immutable audit log; determining the completeness of the path; if complete, determining the final output of the review result.

[0009] In one aspect of this disclosure, the step of obtaining the requirement description text from the on-chain storage of the bidding project via a blockchain node, extracting keyword sequences using natural language processing, and obtaining a feature vector of the bidding requirements includes: The tender project requirement description text stored on the blockchain node is obtained, and the raw text data is obtained by using a distributed query protocol. The word segmentation technique in natural language processing is used to segment the original text data to obtain a word sequence; If the word sequence contains preset stop words, then the stop words are filtered out to obtain a filtered keyword sequence; By using a pre-trained word embedding model, the keyword sequence is transformed into a word vector sequence, generating an initial feature vector; Based on the initial feature vector, the principal component analysis algorithm is used to reduce the dimension of the vector to obtain the optimized feature vector of bidding requirements. If the similarity between the optimized feature vector and the preset business requirement template exceeds a preset threshold, then the feature vector is determined to meet the business requirements, and a structured requirement description is generated. By using structured requirement descriptions, clustering algorithms are employed to classify bidding projects and obtain classification results.

[0010] In one aspect of this disclosure, the step of performing vector transformation processing on the qualification certificate data submitted by the bidding enterprise based on the bidding requirement feature vector, determining the similarity between the transformed vector and the requirement feature vector, and determining preliminary qualification matching if the similarity exceeds a preset threshold includes: Obtain bidding requirement data and extract feature vectors from it. Use keyword extraction and semantic analysis methods to generate requirement feature vectors. The qualification certificate data submitted by bidding companies is structured, and text parsing technology is used to extract key fields and generate qualification data. The qualification data is converted into qualification feature vectors through a pre-trained word vector model. Calculate the cosine similarity between the qualification feature vector and the demand feature vector to obtain a similarity value; If the similarity value exceeds a preset threshold, a preliminary matching of qualifications is determined, and a matching result is generated; Based on the matching results, cluster analysis was used to group the bidding companies and obtain the grouping results. Based on the grouping results, extract the qualification feature vectors of each group of bidding companies and generate a sorted list.

[0011] In one aspect of this disclosure, the step of extracting performance indicator data from historical performance records on the bidding enterprise chain and fusing the performance indicator data with the preliminary qualification matching results using a vector embedding method to obtain a comprehensive qualification vector includes: Performance indicator data is extracted from the historical performance records of the bidding companies on the blockchain, and a structured performance indicator dataset is obtained by using preset indicator extraction rules. By using a structured performance indicator dataset, the weight values ​​of each indicator are calculated, and a weighted average method is used to generate a weighted set of performance indicators. Enterprise qualification data is obtained from the initial qualification matching results. The vector embedding method is used to merge the weighted set of performance indicators with the enterprise qualification data to generate an initial comprehensive qualification vector. If the dimension of the initial comprehensive qualification vector exceeds the preset threshold, principal component analysis is used to reduce the dimension and obtain the optimized comprehensive qualification vector. Based on the optimized comprehensive qualification vector, the similarity between the vectors is calculated, and the cosine similarity algorithm is used to determine the degree of qualification matching between enterprises. By assessing the degree of qualification matching between enterprises and combining it with preset classification rules, the qualification level of enterprises is determined, and the final qualification assessment result is generated. Based on the final qualification assessment results, a data correlation analysis method is used to link the assessment results with on-chain historical performance records to generate a verification dataset of comprehensive qualification vectors.

[0012] In one aspect of this disclosure, the step of obtaining real-time financial statement updates of bidding companies by calling an external oracle interface through a smart contract, determining whether the financial statement update time is before the bidding deadline, and determining the validity of the financial data if it is before the bidding deadline, includes: Obtain the update timestamps of the financial statements of bidding companies from external data sources through the oracle interface and store them as first-time data; Using time comparison logic, the data at the first time is compared with the preset bid deadline. If the data at the first time is earlier than the bid deadline, the update time of the financial statements is determined to be valid, and the validity judgment result is obtained. The validity determination result is obtained through the smart contract interface, and the validity status of the financial data is generated. Based on the validity status, call the oracle interface to obtain the complete dataset of the financial statements and store it as the first financial data; The integrity and consistency of the first financial data are verified through a data validation algorithm to obtain the second financial data that has passed the verification. A classification algorithm is used to assess the financial health of the second set of financial data and generate a financial health score. The final validity of the bidding company's financial data is determined based on the financial health score and validity assessment results.

[0013] In one aspect of this disclosure, the step of performing a weighted fusion calculation based on the comprehensive qualification vector and the validity results of the financial data to obtain the dynamic verification score of the bidding enterprise includes: The system obtains the validity of bidding company information and financial data, and uses preset data verification rules to determine the integrity and consistency of the data, thereby obtaining verified financial data quality. Based on the verified quality of financial data, qualification assessment indicators are extracted, and factor analysis is used to determine the comprehensive qualification vector. If at least one indicator in the comprehensive qualification vector exceeds the preset threshold, a preliminary fusion score is obtained by integrating the qualification assessment indicators and financial data quality through a linear weighting method. For the initial fusion score, a weighted fusion calculation is performed using a preset fusion weight allocation to obtain the adjusted fusion score; Based on the adjusted fusion score, a logistic regression model is applied to determine the enterprise's credit rating and output a dynamic validation score. If the dynamic verification score is lower than the preset threshold, the fusion weight allocation will be adjusted through the score calculation model, and the dynamic verification score will be recalculated. Obtain the recalculated dynamic verification score, and generate the comprehensive evaluation score for the bidding company based on the evaluation results output.

[0014] In one aspect of this disclosure, the step of executing automatic approval logic via a smart contract if the dynamic verification score exceeds a preset threshold, and using a hash function to store the verification process records on-chain to obtain an immutable audit log, includes: If the verification score is higher than the preset threshold, the data to be approved will be obtained through the triggering logic of the smart contract, the automatic approval operation will be executed, and the approval result will be obtained. The approval results and verification records are encrypted using a hash function to generate encrypted data; Encrypted data is written to on-chain storage via the blockchain network, generating a storage address; Retrieve on-chain data based on the storage address, verify data integrity, and obtain verification status; If the verification status is complete, an audit log is generated via a smart contract to record and store the approval information. Based on the audit log content, log data is distributed using a decentralized network to generate distribution confirmation; By distributing confirmations, we obtain consensus feedback from network nodes and ultimately achieve the final on-chain record state.

[0015] In one aspect of this disclosure, the step of retrieving relevant verification paths from the blockchain network based on the immutable audit log, determining the completeness of the paths, and if complete, determining the final output of the audit result includes: The audit logs are obtained from the blockchain network, and a distributed node query protocol is used to extract the log dataset containing timestamps and transaction hashes to obtain the raw audit logs. For the original audit logs, the log structure is parsed, the node sequence and hash chain in the verification path are extracted, and a path dataset is generated; If the node sequence in the path dataset is consistent with the preset blockchain topology, the Merkle tree verification algorithm is used to calculate the hash value of each node in the path, determine the path integrity, and obtain the integrity verification result. Based on the integrity verification results, if the integrity verification passes, the transaction timestamps and associated metadata are extracted from the path dataset to generate the review dataset; The SHA-256 algorithm is used to perform hash calculation on the review dataset to generate review hash values ​​and determine the uniqueness of the review results; By using network communication protocols, the audit hash value is compared with the preset audit standards to determine whether the audit result meets the standards and output the final audit result.

[0016] In another aspect, this disclosure also relates to an engineering bidding platform management system, comprising: The requirement feature extraction module is configured to obtain the requirement description text from the blockchain storage of the bidding project chain through a blockchain node, extract the keyword sequence using natural language processing, and obtain the bidding requirement feature vector. The qualification matching module is configured to perform vector transformation processing on the qualification certificate data submitted by the bidding company based on the bidding requirement feature vector, determine the similarity between the transformed vector and the requirement feature vector, and determine the preliminary qualification matching if the similarity exceeds a preset threshold. The performance fusion module is configured to extract performance indicator data from historical performance records on the bidding enterprise chain, and use a vector embedding method to fuse the performance indicator data with the preliminary qualification matching results to obtain a comprehensive qualification vector. The financial verification module is configured to obtain the real-time financial statement updates of the bidding company by calling the external oracle interface through a smart contract, determine whether the financial statement update time is before the bidding deadline, and determine the validity of the financial data if it is before the bidding deadline. The scoring calculation module is configured to perform a weighted fusion calculation based on the comprehensive qualification vector and the validity results of the financial data to obtain the dynamic verification score of the bidding enterprise. The contract execution module is configured to execute automatic approval logic through a smart contract if the dynamic verification score is higher than a preset threshold, and to use a hash function to store the verification process record on the chain to obtain an immutable audit log. The audit verification module is configured to retrieve relevant verification paths from the blockchain network based on the tamper-proof audit logs, determine the completeness of the paths, and if complete, determine the final output of the audit results.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This invention retrieves the requirement description text from the bidding project chain via blockchain nodes, extracts keywords using natural language processing to generate a requirement feature vector, and vectorizes the bidding company's qualification certificate data, calculating the similarity with the requirement feature vector for preliminary qualification matching. Simultaneously, it extracts historical performance indicators from the chain and uses vector embedding to fuse the preliminary matching results, generating a comprehensive qualification vector. This invention uses a smart contract to call an external oracle interface to verify the update time of financial statements, ensuring the validity of financial data. It then combines this with a weighted fusion calculation of the comprehensive qualification vector to derive a dynamic verification score. If the score exceeds a threshold, the smart contract automatically executes the approval logic. The verification process generates an immutable audit log using a hash function and verifies the integrity of the verification path based on the blockchain network, ensuring the traceability of the review results. This invention significantly improves the efficiency, transparency, and credibility of bid qualification verification, optimizing the bidding process. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart of the bidding platform management method in this invention.

[0020] Figure 2 This is one of the schematic diagrams of the bidding platform management method in this invention.

[0021] Figure 3 This is the second schematic diagram of the bidding platform management method in this invention. Detailed Implementation

[0022] The present invention will be further described below with reference to embodiments. These embodiments are merely some, not all, of the embodiments described. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the protection scope of the present invention.

[0023] Example 1 Please see Figures 1-3 As shown, this embodiment discloses a management method for an engineering bidding platform, including: obtaining the requirement description text from the blockchain storage of the bidding project through a blockchain node, extracting keyword sequences using natural language processing to obtain a bidding requirement feature vector; performing vector transformation processing on the qualification certificate data submitted by the bidding company based on the bidding requirement feature vector, judging the similarity between the transformed vector and the requirement feature vector, and determining preliminary qualification matching if the similarity exceeds a preset threshold; extracting performance indicator data from the historical performance records on the bidding company's blockchain, fusing the performance indicator data with the preliminary qualification matching results using a vector embedding method to obtain a comprehensive qualification vector; and adjusting the management method through a smart contract. The system uses an external oracle interface to obtain real-time updates of the bidding company's financial statements and determines whether the update time is before the bidding deadline. If it is, the validity of the financial data is confirmed. Based on the comprehensive qualification vector and the validity result of the financial data, a weighted fusion calculation is performed to obtain the dynamic verification score of the bidding company. If the dynamic verification score is higher than a preset threshold, an automatic approval logic is executed through a smart contract. The verification process record is stored on-chain using a hash function to obtain an immutable audit log. Based on the immutable audit log, the relevant verification path is retrieved from the blockchain network, and the completeness of the path is determined. If the path is complete, the final review result is determined and output.

[0024] This invention retrieves the requirement description text from the bidding project chain via blockchain nodes, extracts keywords using natural language processing to generate a requirement feature vector, and vectorizes the bidding company's qualification certificate data, calculating the similarity with the requirement feature vector for preliminary qualification matching. Simultaneously, it extracts historical performance indicators from the chain and uses vector embedding to fuse the preliminary matching results, generating a comprehensive qualification vector. This invention uses a smart contract to call an external oracle interface to verify the update time of financial statements, ensuring the validity of financial data. It then combines this with a weighted fusion calculation of the comprehensive qualification vector to derive a dynamic verification score. If the score exceeds a threshold, the smart contract automatically executes the approval logic. The verification process generates an immutable audit log using a hash function and verifies the integrity of the verification path based on the blockchain network, ensuring the traceability of the review results. This invention significantly improves the efficiency, transparency, and credibility of bid qualification verification, optimizing the bidding process.

[0025] Example 2 Please see Figures 1-3 As shown, this embodiment is a further optimization based on Embodiment 1. In this embodiment, the bidding platform management method may specifically include: S1 retrieves the requirement description text from the blockchain storage of the bidding project through blockchain nodes, extracts the keyword sequence using natural language processing, and obtains the feature vector of the bidding requirements.

[0026] The system retrieves the bidding project requirement description text stored on the blockchain node, uses a distributed query protocol to obtain the raw text data, and employs word segmentation technology from natural language processing to segment the raw text data to obtain a word sequence. If the word sequence contains preset stop words, the stop words are filtered to obtain a filtered keyword sequence. A pre-trained word embedding model is used to transform keyword sequences into word vector sequences, generating initial feature vectors. Based on these initial feature vectors, principal component analysis is employed to reduce the dimensionality of the vectors, resulting in optimized bidding requirement feature vectors. If the similarity between the optimized feature vectors and the preset business requirement template exceeds a preset threshold, the feature vectors are determined to meet the business requirements, and a structured requirement description is generated.

[0027] By using structured requirement descriptions, clustering algorithms are employed to classify bidding projects and obtain classification results.

[0028] For example, when retrieving the tender project requirement description text from a blockchain node, the text data stored on the chain can be called through a smart contract. Assuming the Ethereum blockchain is used, the node calls the ImmGetDescription function of the smart contract through the JSON-RPC interface to obtain the requirement description file stored on IPFS.

[0029] The node uses the IPFS get method to download the file, which takes about 0.8 seconds. The obtained text content is: This project requires the development of a supply chain management system that supports real-time inventory tracking, data analysis modules, compatibility with MySQL database, and a response time of less than 0.5 seconds.

[0030] Subsequently, natural language processing techniques were used to extract keyword sequences, and the BERT model was used to perform word segmentation and semantic analysis on the text, with the maximum sequence length set to 128 and the word embedding dimension set to 768.

[0031] After word segmentation, the word sequence obtained is: {"supply chain management", "real-time inventory tracking", "data analysis module", "MySQL database", "response time"}. The inverse document frequency of the term is calculated using the TF-IDF algorithm. Assuming that "supply chain management" appears 3 times in the word frequency statistics, the TF value is 0.15, the IDF value is 2.3, and the keyword weight is 0.345.

[0032] The keyword sequence was vectorized, and a 300-dimensional feature vector was generated using the Word2Vec model. Assuming the vector value of supply chain management is [0.12, -0.45, 0.67], the correlation between vectors was calculated using cosine similarity. The similarity between the real-time inventory tracking and data analysis modules was 0.82, indicating a strong correlation between demand.

[0033] Finally, all keyword feature vectors are integrated to form a 768-dimensional comprehensive feature vector of bidding requirements, which is stored in a local database for subsequent matching analysis. The entire process is implemented through automated scripts without manual intervention, and logically forms a closed loop from data acquisition to feature extraction.

[0034] S2. Based on the feature vector of the bidding requirements, the qualification certificate data submitted by the bidding companies is transformed into vectors. The similarity between the transformed vector and the feature vector of the requirements is judged. If the similarity exceeds the preset threshold, the qualification is initially matched.

[0035] The process involves acquiring bidding requirement data, extracting feature vectors from it, and generating requirement feature vectors using keyword extraction and semantic analysis. The qualification certificate data submitted by bidding companies is then structured using text parsing technology to extract key fields and generate qualification data. A pre-trained word vector model is used to convert the qualification data into qualification feature vectors. The cosine similarity between the qualification feature vectors and the requirement feature vectors is calculated to obtain a similarity value. If the similarity value exceeds a preset threshold, a preliminary qualification match is determined, and matching results are generated. Based on the matching results, cluster analysis is used to group the bidding companies, resulting in grouping results. According to the grouping results, the qualification feature vectors of each group of bidding companies are extracted, generating a ranked list.

[0036] For example, the process of vectorizing and judging the similarity of bidding companies' qualification certificate data based on the feature vector of bidding requirements can be achieved through the following information technology. Assume that the bidding requirements are for a construction project, requiring companies to have a Class A general contracting qualification, have completed 5 similar projects in the past three years, and have a registered capital of no less than 50 million yuan; the requirement feature vector can be defined as [1, 5, 5000], representing the qualification level (1 is Class A), the number of project experiences, and the registered capital, respectively.

[0037] The data submitted by bidding company A is as follows: Class A general contracting qualification, 4 similar projects completed in the past three years, and registered capital of 60 million yuan. It needs to be converted into a vector [1,4,6000]. In order to unify the dimensions, the vector is first normalized using the Min-Max normalization algorithm.

[0038] Assuming the maximum values ​​for each dimension are qualification level 2 (2 for level 2), number of projects 10, and registered capital 100 million yuan, and the minimum values ​​are 0, 0, and 0 respectively, then the demand vector is normalized to [(1-0) / (2-0), (5-0) / (10-0), (5000-0) / (10000-0)] = [0.5, 0.5, 0.5], and the vector for company A is normalized to [0.5, 0.4, 0.6]. Next, the similarity between the two vectors is calculated using the cosine similarity algorithm. The formula is: ; Calculate the dot product: 0.5×0.5+0.5×0.4+0.5×0.6=0.75; Length = ; = ; Cosine similarity is ≈0.987. If the preset threshold is 0.9, since 0.987>0.9, it is determined that Company A's qualifications are initially matched.

[0039] To ensure logical rigor, weight adjustments can be introduced, such as a qualification level weight of 0.5, project experience weight of 0.3, and capital weight of 0.2. The weighted similarity is calculated to further refine the matching accuracy. The system automatically outputs the matching results and records logs for subsequent auditing.

[0040] S3 extracts performance indicator data from the historical performance records of bidding companies on the blockchain, and uses vector embedding method to fuse the performance indicator data with the preliminary qualification matching results to obtain a comprehensive qualification vector.

[0041] S3-1: Extract performance indicator data from the bidding companies' historical on-chain performance records. Using pre-defined indicator extraction rules, obtain a structured performance indicator dataset. Calculate the weight of each indicator from this dataset and generate a weighted performance indicator set using a weighted average method. Obtain company qualification data from the initial qualification matching results. Use vector embedding to fuse the weighted performance indicator set with the company qualification data, generating an initial comprehensive qualification vector. If the dimensionality of the initial comprehensive qualification vector exceeds a pre-defined threshold, use principal component analysis for dimensionality reduction to obtain an optimized comprehensive qualification vector. Calculate the similarity between vectors based on the optimized comprehensive qualification vector, using a cosine similarity algorithm to determine the degree of qualification matching between companies. Based on the degree of qualification matching between companies and pre-defined classification rules, determine the company's qualification level and generate the final qualification assessment result. Based on the final qualification assessment result, use data association analysis to link the assessment result with historical on-chain performance records, generating a verification dataset for the comprehensive qualification vector.

[0042] For example, extracting performance indicator data from a bidding company's historical performance records on the blockchain can be achieved through blockchain smart contracts and data parsing technology. Assume the blockchain stores company A's project records for the past 5 years, including project amounts, completion times, customer ratings, and other information. The smart contract calls an API interface to extract data from the blockchain, obtaining company A's performance indicators: cumulative project amount of 500 million yuan, average project cycle of 6 months, and average customer rating of 4.8 (assuming a maximum score of 5).

[0043] S3-2, the data parsing module standardizes these indicators, for example, normalizing the amount to the range [0,1], with 500 million yuan corresponding to 0.85, assuming the maximum amount is 1 billion yuan, the period and score are normalized to 0.67 and 0.96 respectively.

[0044] Next, a vector embedding method was used to fuse performance indicator data with the preliminary qualification matching results. The qualification matching results came from an enterprise qualification database, including enterprise registered capital, qualification level, number of technical personnel, etc.

[0045] For example, Company A has a registered capital of 80 million yuan, which is 0.8 after normalization; its qualification level is Level 1, which is assigned a value of 1; and it has 50 technical personnel, which is 0.5 after normalization.

[0046] S3-3 uses the Word2Vec algorithm to convert the performance indicators (0.85, 0.67, 0.96) and qualification data (0.8, 1, 0.5) into high-dimensional vectors with a dimension of 128.

[0047] Word2Vec maps each indicator to a semantic vector using a pre-trained model, capturing potential correlations between indicators, such as the correlation between amount and qualification level. During fusion, a weighted average method is used, with the performance vector weight set to 0.6 and the qualification vector weight set to 0.4, calculating the comprehensive vector V = 0.6 × V. 业绩 + 0.4 × V 资质 A 128-dimensional comprehensive qualification vector was obtained.

[0048] The analysis process ensures semantic consistency between vectors, and the fusion effect is verified by cosine similarity. Assuming that the similarity between V and the ideal qualification vector is 0.92, it indicates that the fusion result is reliable.

[0049] On the logical chain, on-chain data extraction ensures data authenticity, standardized processing unifies dimensions, vector embedding captures semantics, and fusion algorithms balance performance and qualification weights, ultimately outputting a comprehensive qualification vector for subsequent bidding evaluation.

[0050] Steps S3-4: If the dimensionality of the initial comprehensive qualification vector exceeds a preset threshold, principal component analysis is used for dimensionality reduction to obtain an optimized comprehensive qualification vector. The initial comprehensive qualification vector typically contains multiple dimensions of qualification data and performance indicators, such as 128 dimensions or higher; using it directly may increase computational complexity and storage requirements.

[0051] Therefore, principal component analysis (PCA) projects high-dimensional vectors into a low-dimensional space, preserving key information. For example, assuming an initial comprehensive qualification vector of 128 dimensions, PCA retains the first 32 dimensions as the optimized comprehensive qualification vector. These dimensions typically correspond to core characteristics of qualifications and performance, such as technical certification levels or project delivery efficiency.

[0052] Specifically, principal component analysis first calculates the covariance matrix of the initial comprehensive qualification vector to determine the main direction of data variation. Then, through eigenvalue decomposition, it selects the principal components with larger eigenvalues ​​as projection basis vectors. The initial comprehensive qualification vector is projected onto these basis vectors to generate a low-dimensional optimized vector. This method effectively reduces redundant information while retaining the core features of a company's qualifications and performance. For example, in an IoT device procurement scenario, the optimized comprehensive qualification vector may highlight IoT certification and delivery time, weaken the influence of secondary information, and improve the efficiency of subsequent matching calculations.

[0053] Steps S3-5 involve calculating the similarity between the optimized comprehensive qualification vectors and using a cosine similarity algorithm to determine the degree of qualification matching between companies. The optimized comprehensive qualification vectors reflect the overall capabilities of bidding companies. By comparing the similarity between the vectors of different companies, the degree of qualification matching is determined. For example, suppose the demand feature vector of a bidding project emphasizes 5G technology and fast delivery, while a company's comprehensive qualification vector highlights 5G certification and a 6-month delivery record. The angle between the two in the vector space is small, and the similarity value may be 0.88, indicating a high degree of matching.

[0054] In one possible implementation, cosine similarity is calculated by the ratio of the vector dot product to the magnitude, focusing on the similarity of vector directions; the result is between 0 and 1, and the closer the value is to 1, the higher the degree of matching between the enterprise's qualifications and the bidding requirements; It should be noted that similarity calculation can be combined with other methods, such as Euclidean distance, as supplementary verification to ensure the robustness of the results. The similarity value serves as the basis for subsequent qualification level determination and directly affects the company's assessment results.

[0055] Steps S3-6: Based on the degree of qualification matching between enterprises and the preset classification rules, determine the qualification level of the enterprises and generate the final qualification assessment result. The classification rules are customized based on the business needs of the bidding project. For example, in the IoT device procurement scenario, the rules may stipulate that a similarity greater than 0.85 is a superior qualification and 0.7 to 0.85 is a qualified qualification.

[0056] For example, if the similarity between a company's comprehensive qualification vector and its demand feature vector is 0.88, then according to the classification rules, it is determined to be of superior quality. The qualification assessment results are recorded in a structured form, for example, Company A: superior quality; similarity: 0.88.

[0057] In one embodiment, the classification rules can also incorporate other business factors, such as the company's registered capital or industry experience. For example, if a company's registered capital exceeds a certain threshold, it may receive a higher qualification level under the same similarity criteria. This approach improves the comprehensiveness of the assessment and adapts to the needs of different bidding scenarios. The generated qualification assessment results provide an important basis for subsequent comprehensive assessments.

[0058] Steps S3-7 involve using data correlation analysis to link the final qualification assessment results with historical on-chain performance records, generating a verification dataset for the comprehensive qualification vector. This data correlation analysis verifies the accuracy of the assessment by comparing the results with historical on-chain performance records.

[0059] For example, if a company's qualification assessment result is excellent, the correlation analysis will check whether its historical performance records include high-quality IoT project delivery records, such as delivery time: 6 months; satisfaction rate: 95%. If the records are consistent with the assessment results, a verification dataset will be generated to record the correlation information.

[0060] Specifically, data correlation analysis is achieved by matching key fields, such as comparing the IoT certification in the evaluation results with the IoT project field in the historical performance records; verifying that the dataset is stored in a structured form, such as Company A: excellent qualification; correlation performance: IoT project, satisfaction rate 95%.

[0061] This approach ensures the credibility of the assessment results and provides data support for subsequent audits.

[0062] Step S4: The smart contract calls an external oracle interface to obtain the real-time financial statement updates of the bidding company. It then determines whether the financial statement update time is before the bidding deadline. If it is, the validity of the financial data is confirmed. The smart contract interacts with the external data source through the oracle interface to obtain the latest financial statement timestamp of the bidding company. For example, if a company's financial statement update time is [Date], and the bidding deadline is [Date], the timestamp meets the requirements, indicating that the financial data is valid.

[0063] Step S4-1: Obtain the updated timestamps of the financial statements of the bidding companies from an external data source through the oracle interface and store them as first-time data. The oracle acts as a bridge between the blockchain and the outside world, obtaining the timestamp information of the financial statements from trusted data sources.

[0064] For example, an oracle may obtain a timestamp, such as 20XX-0X-0X, XX:XX:XX, from a company's financial system or a third-party data platform, and store the data in a temporary storage area on the blockchain as soon as it is available, ensuring the traceability of the data.

[0065] In one possible implementation, the oracle interface employs a multi-source verification mechanism, obtaining timestamps from multiple data sources and determining the final value through a consensus algorithm. For example, if three data sources provide timestamps such as 20XX-0X-0X,XX:XX:XX, 20XX-0X-0X,XX:XX:XX, and 20XX-0X-0X,XX:XX:XX, the timestamp with the majority agreement is taken as the first-time data. This method improves the reliability of data acquisition and reduces the potential risks associated with a single data source.

[0066] Step S4-2 employs time comparison logic to compare the first-time data with the preset bid deadline. If the first-time data is earlier than the bid deadline, the financial statement update time is determined to be valid, resulting in a validity assessment. The time comparison logic determines the validity of the financial statements by comparing the order of timestamps. For example, if the first-time data is "20XX-0X-0X XX:XX:XX" and the bid deadline is "20XX-0X-0X XX:XX:XX", then the timestamps meet the requirements, and the statement is deemed valid.

[0067] Specifically, time comparison logic can be implemented through simple date comparison, such as converting the timestamp to a standard format and comparing the year, month, day, hour, minute, and second digit by digit.

[0068] The validity determination result is recorded in Boolean form, for example, valid: yes.

[0069] It should be noted that if the timestamp is later than the bid deadline, the result will be deemed invalid, and the company may need to resubmit its financial data. This rigorous time verification ensures the fairness of the evaluation process.

[0070] Step S4-3: Obtain the validity determination result through the smart contract interface and generate the financial data validity status. The smart contract updates the company's financial data status based on the validity determination result.

[0071] For example, if the validity determination result is valid: Yes, the smart contract generates valid state financial data and records it on the blockchain. The state information is stored in a structured form, such as: Company A: Financial data is valid; Timestamp: 2025-0X-0X.

[0072] In one embodiment, the smart contract can also trigger a notification mechanism to send validity status feedback to bidding companies. For example, if the financial data is valid, the smart contract notifies the company to continue the subsequent evaluation process via an on-chain messaging system. This automated feedback improves the transparency and efficiency of the evaluation.

[0073] Step S4-4: Based on the validity status, call the oracle interface to obtain the complete dataset of financial statements and store it as the first financial data. If the financial data is valid, the oracle interface obtains the complete financial statement data from an external data source, such as the balance sheet and income statement.

[0074] For example, a company's initial financial data might include total assets: 50 million yuan; net profit: 5 million yuan; and debt ratio: 30%.

[0075] These data are stored in the distributed storage of the blockchain, ensuring they cannot be tampered with.

[0076] Specifically, the oracle interface obtains data through a secure communication protocol and performs preliminary formatting, such as converting the data into JSON format.

[0077] It should be noted that the data acquisition process may involve multiple verifications, such as confirming the integrity of the data through multiple oracle nodes. The primary financial data provides a comprehensive foundation for subsequent financial health assessments.

[0078] Steps S4-5 involve verifying the completeness and consistency of the first financial data using a data validation algorithm to obtain the validated second financial data. The data validation algorithm checks the format, field completeness, and logical consistency of the first financial data. For example, the algorithm might check whether total assets in the balance sheet equal total liabilities plus owner's equity, or whether net profit in the income statement is consistent with revenue and expense data.

[0079] If the verification passes, the second set of financial data will be generated.

[0080] In one possible implementation, the data validation algorithm combines rule-based validation and statistical analysis. For example, rule-based validation ensures that field formats conform to standards, while statistical analysis detects outliers in the data, such as abnormally high debt ratios. The validated second set of financial data is stored in a structured format, for example, Company A: Total assets of 50 million yuan, net profit of 5 million yuan.

[0081] This approach ensures the reliability of financial data and provides high-quality input for subsequent evaluations.

[0082] Steps S4-6 involve using a classification algorithm to assess the financial health of the second set of financial data and generate a financial health score. The classification algorithm evaluates the company's financial health based on key indicators of the financial data.

[0083] For example, the algorithm may classify companies into three levels: healthy, average, and risky, based on indicators such as net profit, debt ratio, and cash flow.

[0084] Suppose a company has a net profit of 5 million yuan, a debt ratio of 30%, and stable cash flow. The classification algorithm may generate a financial health score of 0.9, indicating that the company's financial condition is good.

[0085] Specifically, classification algorithms can be implemented using decision trees or logistic regression. For example, a decision tree judges each level according to a preset threshold, such as a debt ratio below 50%, and finally generates a score.

[0086] It should be noted that the scoring can be further refined into multiple dimensions, such as profitability: 0.85; solvency: 0.9; the financial health score directly affects the company's overall assessment result.

[0087] Step S4-7: Based on the financial health score and validity assessment results, determine the final validity of the bidding company's financial data. If the financial health score is higher than a preset threshold, such as 0.8, and the validity assessment is "yes," then the financial data is determined to be ultimately valid.

[0088] For example, if a company's financial health score is 0.9 and its validity is determined to be yes, then its final validity is passed.

[0089] The results are recorded in a structured format, for example, Company A: Financial data passed; Score: 0.9.

[0090] In one embodiment, the final validity can also be combined with other factors, such as the company's industry ranking or credit record. For example, if a company is an industry leader, it may be deemed valid even if its score is slightly below a threshold, by adjusting the weights. This flexible determination method adapts to the needs of different bidding scenarios.

[0091] Step S5 involves performing a weighted fusion calculation based on the comprehensive qualification vector and the financial data validity results to obtain the dynamic verification score for the bidding company. The comprehensive qualification vector includes the company's technical capabilities and historical performance information, while the financial data validity results reflect its financial health. The two are weighted and fused to generate a comprehensive score. For example, a company's comprehensive qualification vector highlights 5G certification, and its financial health score is 0.9. Through fusion calculation, a dynamic verification score of 0.87 is generated.

[0092] Step S5-1: Obtain the validity of the bidding company's information and financial data. Using pre-defined data validation rules, assess the data's completeness and consistency to obtain verified financial data quality. The data validation rules ensure data reliability by checking the completeness of the company information and financial data.

[0093] For example, if a company's registered capital is 50 million yuan, its financial statements show total assets of 50 million yuan, and all required fields are complete and the verification passes, the generated financial data is of high quality.

[0094] The verification process includes comparing the company name, registered capital, and key fields in the financial statements to ensure consistency.

[0095] Specifically, data validation rules first check the completeness of fields, such as confirming whether the financial statements include required fields like assets, liabilities, and revenue. Secondly, they ensure data validity through logical checks, such as ensuring that total assets equal the sum of liabilities and owner's equity.

[0096] If missing fields or contradictory data are found, the system will mark it as low quality and require the company to supplement the data. Verified financial data quality provides a reliable foundation for subsequent assessments, reducing assessment bias caused by data errors.

[0097] Step S5-2: Based on the verified quality of the financial data, extract qualification assessment indicators and use factor analysis to determine the comprehensive qualification vector. Factor analysis analyzes the correlation between financial data and qualification data to extract key indicators and form a comprehensive qualification vector. For example, if a company's net profit indicator is prominent in its financial data and its technical certification level is high in its qualification data, factor analysis will extract both as the main factors to generate a comprehensive qualification vector.

[0098] This vector represents a company's overall capabilities in a multidimensional form, such as technology certification: 0.9; profitability: 0.8.

[0099] In one possible implementation, factor analysis first calculates the correlation matrix of each indicator to determine which indicators have a greater impact on the evaluation results.

[0100] For example, technical certification and net profit may have a high correlation and be selected as principal factors. Then, through factor loading analysis, qualification and financial data are projected onto the factor space to generate a comprehensive qualification vector. This method can effectively integrate multi-source data, improving the comprehensiveness and accuracy of the assessment; the generated comprehensive qualification vector provides the core input for subsequent fusion calculations.

[0101] Step S5-3: If at least one indicator in the comprehensive qualification vector exceeds the preset threshold, the qualification assessment indicators and financial data quality are integrated using a linear weighting method to obtain a preliminary integrated score.

[0102] The linear weighted method sums the qualification and financial indicators according to preset weights.

[0103] For example, if the technical certification index value in the comprehensive qualification vector is 0.9, which exceeds the threshold of 0.8, and the financial data quality is high, then through weight allocation, the technical certification is 0.6, the financial data quality is 0.4, and the preliminary integration score is calculated to be 0.85.

[0104] Specifically, the linear weighting method calculates a weighted sum by assigning weights to each indicator.

[0105] For example, if the weight of technical certification is 0.6 and the weight of financial data quality is 0.4, the fusion formula is: Score = 0.6 × Technical Certification Value + 0.4 × Financial Data Quality Value.

[0106] It should be noted that the weighting can be dynamically adjusted based on the priority of the bidding project. For example, in IoT projects with high technical requirements, the weight of technical certification may be increased to 0.7.

[0107] This flexible weighting improves the adaptability of the assessment and ensures that scores reflect project needs.

[0108] Step S5-4: Based on the initial fusion score, a weighted fusion calculation is performed using a preset fusion weight allocation to obtain the adjusted fusion score. The fusion weight allocation is further optimized based on the business scenario to ensure that the score more accurately reflects the company's comprehensive capabilities. For example, an initial fusion score of 0.85 is adjusted by weighting technical certification (0.7) and financial data quality (0.3) to generate an adjusted fusion score of 0.87. The adjusted score better aligns with the technical priority requirements of the IoT device procurement project.

[0109] In one embodiment, the weighting of the fusion score is determined by analyzing successful cases from historical bidding projects. For example, if historical data indicates that technical capabilities have a greater impact on the success rate, the weight of technical certification is increased. The adjusted fusion score is generated through a weighted calculation, for example: Score = 0.7 × Technical Certification Value + 0.3 × Financial Data Quality Value. This method ensures the score's relevance while retaining the supplementary role of financial data. The adjusted fusion score provides a more accurate input for subsequent credit rating assessments.

[0110] Step S5-5: Based on the adjusted fusion score, apply a logistic regression model to determine the company's credit rating and output a dynamic validation score. The logistic regression model predicts the company's credit rating, such as high, medium, or low, by analyzing the fusion score.

[0111] For example, if the adjusted fusion score is 0.87, the logistic regression model predicts a high credit rating and outputs a dynamic validation score of 0.87, which comprehensively reflects the company's qualifications and financial capabilities.

[0112] Specifically, the logistic regression model learns the mapping relationship between the fusion score and the credit rating by training on historical data.

[0113] For example, the model might determine that a score of 0.8 or higher corresponds to a high credit rating based on a large amount of historical bidding data. It should be noted that the model can also incorporate other features, such as the company's industry experience or registered capital, to further improve the accuracy of the prediction. The dynamic validation score, as the core output of the assessment, provides a direct basis for the subsequent approval process.

[0114] In steps S5-6, if the dynamic verification score is lower than a preset threshold, the fusion weight allocation is adjusted using the score calculation model, and the dynamic verification score is recalculated. The score calculation model maximizes the stability of the score by iteratively optimizing the weights. For example, if a company's dynamic verification score is 0.6, which is lower than the threshold of 0.8, the weights are adjusted, for example, the weight of financial data quality is increased from 0.3 to 0.4, and the score is recalculated to 0.82.

[0115] In one possible implementation, the score calculation model optimizes the weights using gradient descent to find the weight combination that maximizes the score. If the recalculated score is still below a threshold, the system may require the company to provide supplementary data, such as updated financial statements or additional qualification documentation. This approach ensures the fairness of the assessment while providing companies with opportunities for improvement. The dynamically validated recalculated score enhances the reliability of the assessment results.

[0116] Step S57: Obtain the recalculated dynamic verification score and generate a comprehensive evaluation score for the bidding company based on the evaluation results output. The comprehensive evaluation score integrates the results of qualification, performance, and financial data, and is output in a structured form. For example, a company's comprehensive evaluation score is 0.82, recorded as Company A: Comprehensive Evaluation Score 0.82; Technical Capability: 0.9; Financial Health: 0.8; This score provides the bidding party with a comprehensive evaluation basis.

[0117] Specifically, the comprehensive evaluation score output includes a score and evaluation details, such as specific values ​​for dimensions like technical capabilities and financial health. The output is stored via blockchain to ensure transparency and immutability. It's worth noting that the evaluation results can also generate visual reports, such as bar charts displaying the scores for each dimension, facilitating quick review by the bidding party. This structured output enhances the practicality and operability of the evaluation.

[0118] Step S6: If the dynamic verification score is higher than a preset threshold, the automatic approval logic is executed through the smart contract. A hash function is used to store the verification process records on-chain, resulting in an immutable audit log. The dynamic verification score reflects the company's overall competitiveness; if it exceeds a threshold, such as 0.8, the automatic approval logic of the smart contract is triggered.

[0119] For example, if a company has a dynamic verification score of 0.82, the smart contract will automatically approve its bidding eligibility and record the verification process.

[0120] Step S6-1: If the verification score is higher than a preset threshold, the data to be approved is obtained through the triggering logic of the smart contract, an automatic approval operation is executed, and an approval result is generated. The smart contract extracts the data of the enterprise to be approved, such as qualification vectors and financial data, based on the score threshold, and generates the approval result. For example, if an enterprise's score is 0.82, the smart contract generates the result: Enterprise A: Approved.

[0121] In one embodiment, the smart contract uses a condition-triggered mechanism, such as approving if the score is greater than 0.8. The triggering logic checks whether the score meets the condition; if so, it extracts the relevant data and generates an approval result. The approval result is stored on-chain to ensure transparency and traceability. This automated approval process improves efficiency and reduces human intervention.

[0122] Step S6-2: Encrypt the approval result and verification record using a hash function to generate encrypted data. The hash function converts the approval result and verification record into a fixed-length hash value, ensuring the uniqueness and security of the data. For example, the approval result for company A: The approval and verification record generate the hash value abc123.

[0123] Specifically, the hash function uses the SHA-256 algorithm, and the input data includes approval results, verification scores, and timestamps. The generated hash value has a fixed length, such as 256 bits, which is suitable for on-chain storage.

[0124] It should be noted that the one-way nature of hash functions ensures data immutability, providing a reliable guarantee for subsequent auditing. Encrypted data provides secure input for on-chain storage.

[0125] Step S6-3 involves writing encrypted data to on-chain storage via the blockchain network, generating a storage address. The encrypted data is written to the blockchain through a distributed ledger, generating a unique storage address, such as 0x1234. The writing process is completed through a consensus mechanism to ensure the data is recognized by all nodes.

[0126] In one possible implementation, the write process employs a proof-of-work or proof-of-stake mechanism to ensure data consistency. The storage address is recorded in hexadecimal format for easy subsequent retrieval. For example, after a company's encrypted data is written, address 0x1234 is generated and recorded in the smart contract's log.

[0127] This approach ensures distributed storage and traceability of data.

[0128] Step S6-4: Retrieve on-chain data based on the storage address, verify data integrity, and obtain the verification status. The verification process compares the hash value of the on-chain data with the original encrypted data to determine if the data is consistent. For example, if the hash value of the data in storage address 0x1234 matches the original hash value abc123, the verification status is complete.

[0129] Specifically, the verification process is implemented through the query interface of blockchain nodes, retrieving data from the storage address and calculating its hash value. If the hash values ​​match, it indicates that the data has not been tampered with. It should be noted that verification can also be combined with timestamp checks to ensure the time validity of the data. The verification status provides a reliable basis for the generation of subsequent audit logs.

[0130] In step S6-5, if the verification status is complete, an audit log is generated via a smart contract, recording approval and storage information. The audit log is stored in a structured format, containing information such as approval result, timestamp, and storage address. For example, the audit log records: Company A: Approved; Time: 20XX-0X-0X; Storage Address: 0x1234.

[0131] In one embodiment, audit logs are generated via the logging function of a smart contract and stored in the public ledger of a blockchain. The log content supports multi-party queries; for example, the bidding party can retrieve approval records by accessing the storage address. This approach improves the transparency of the evaluation process and facilitates oversight and review.

[0132] Step S6-6: Based on the audit log content, distribute log data using a decentralized network and generate a distribution confirmation. The log data is distributed to multiple nodes through the blockchain network to ensure widespread data dissemination. For example, after distribution is complete, a confirmation message is generated indicating distribution completed on 10 nodes.

[0133] Specifically, the distribution process is implemented through a peer-to-peer network, with log data transmitted to each node in encrypted form. It's important to note that distribution confirmation is verified through node signatures, ensuring that all nodes receive consistent data. Distribution confirmation enhances the credibility and accessibility of audit logs, providing a foundation for subsequent consensus feedback.

[0134] Steps S6-7 involve distributing confirmations to obtain consensus feedback from network nodes and determining the final on-chain record status. The consensus feedback is verified by a majority of nodes to confirm the consistency of the log data. For example, if 8 out of 10 nodes confirm data consistency, the record status is "confirmed."

[0135] In one possible implementation, consensus feedback employs a Byzantine fault-tolerant algorithm, tolerating partial node failures or malicious behavior. The final recorded state is stored in a structured format, for example, Company A: Recorded state confirmed; Number of nodes: 8.

[0136] This method ensures the reliability and stability of on-chain records.

[0137] Step S7: Based on the tamper-proof audit log, retrieve the relevant verification path from the blockchain network, determine the path integrity, and if complete, determine the final output of the audit result.

[0138] Audit logs are stored on the blockchain and contain a complete record of approvals and verifications, providing a reliable basis for audits. For example, retrieving an enterprise's audit logs reveals an approval time: 20XX-0X-0X; hash value: abc123; storage address: 0x1234.

[0139] Step S7-1: Obtain audit logs from the blockchain network. Employ a distributed node query protocol to extract a log dataset containing timestamps and transaction hashes, obtaining the original audit logs. The distributed query protocol ensures the integrity and consistency of log data through multi-node collaboration. For example, the query protocol extracts logs from 10 nodes to generate a dataset: Enterprise A: Approved; Time: 20XX-0X-0X; Hash: abc123.

[0140] Specifically, the query protocol verifies the authenticity of the log data source by verifying the digital signatures of the nodes. The extracted log dataset contains fields such as timestamps, transaction hashes, and approval results, and is stored in a structured format. It should be noted that the query process may involve multiple rounds of communication to ensure that all nodes return consistent data. The raw audit logs provide complete foundational data for subsequent path analysis.

[0141] Step S7-2: For the original audit log, parse the log structure, extract the node sequence and hash chain in the verification path, and generate a path dataset. The log structure typically contains multiple records, each corresponding to a verification step. For example, after log parsing, the path node A-node B-node C is extracted, and the hash chain is abc123-def456-ghi789.

[0142] In one embodiment, the parsing process extracts key fields, such as timestamps, hash values, and node identifiers, using regular expressions.

[0143] The generated path dataset is stored in list form, for example, path: [node A, abc123]-[node B, def456]; this structured parsing improves data processing efficiency and provides clear input for path integrity verification.

[0144] Step S7-3: If the node sequence in the path dataset matches the preset blockchain topology, the Merkle tree verification algorithm is used to calculate the hash value of each node in the path, determine the path integrity, and obtain the integrity verification result. The Merkle tree algorithm verifies the path integrity by calculating hash values ​​layer by layer.

[0145] For example, if the hash chain of path node A-node B-node C matches the root hash stored on the chain, the verification result is complete.

[0146] Specifically, the Merkle tree algorithm first pairs the hash values ​​in the path together, calculates the hash of the next level, and so on until the root hash is generated.

[0147] Then, the root hash is compared with the root hash stored on the chain; if they match, the path is complete.

[0148] It should be noted that the algorithm can also detect missing or tampered information in the path. For example, if the hash value of a node does not match, the verification result is "incomplete". The integrity verification result ensures the credibility of the log data.

[0149] Step S7-4: Based on the integrity verification result, if the integrity verification passes, extract the transaction timestamp and associated metadata from the path dataset to generate a review dataset. The review dataset integrates key information from the path, such as timestamps, hash values, and approval results. For example, the review dataset records: Company A: Time 20XX-0X-0X; Hash: abc123; Approval.

[0150] In one possible implementation, the review dataset is stored in a structured format, such as JSON, to facilitate subsequent analysis and comparison. The extraction process ensures that all fields are complete, for example, the timestamp format is "YYYY-MM-DD HH:MM:SS"; the review dataset provides direct input for generating the final review results.

[0151] Step S7-5: Use the SHA-256 algorithm to perform hash calculation on the review dataset to generate review hash values ​​and determine the uniqueness of the review results.

[0152] The SHA-256 algorithm converts the review dataset into a fixed-length hash value, ensuring the uniqueness and irreversibility of the result. For example, the review dataset generates the hash value xyz789.

[0153] Specifically, the SHA-256 algorithm takes all fields of the review dataset as input, such as timestamps, approval results, and path information, and generates a 256-bit hash value.

[0154] It should be noted that the uniqueness of the hash value ensures the verifiability of the review results; any data tampering will cause the hash value to change. The review hash value provides a reliable identifier for the final review.

[0155] Step S7-6: Compare the review hash value with the preset audit standards through the network communication protocol to determine whether the review result meets the standards, and output the final review result.

[0156] Audit criteria include requirements such as hash value consistency and timestamp validity. For example, if the audited hash value xyz789 matches the hash value stored on the chain and the timestamp is within a valid range, the audit is considered passed.

[0157] In one embodiment, the network communication protocol transmits hash values ​​through an encrypted channel to ensure the security of the comparison process. The comparison process checks whether the hash values ​​match and whether the timestamps meet the bid deadline requirements. The final review result is output in a structured form, for example, Company A: Review Passed; Hash: xyz789.

[0158] This approach ensures the transparency and credibility of the review process, providing a final conclusion for the tender evaluation.

[0159] In one possible implementation, the review results can be automatically notified to relevant parties, such as the tendering party and bidding companies, via smart contracts.

[0160] The notification includes the review result, hash value, and timestamp, stored on-chain for easy verification by multiple parties. For example, after a company's review is approved, the smart contract generates a notification to Company A: Review Approved; Time: 20XX-0X-0X, which is sent via the on-chain messaging system.

[0161] This automated notification improves the efficiency and transparency of the evaluation process.

[0162] Specifically, the output of the review results can also be combined with visualization tools, such as generating review reports that display hash values, timestamps, and approval details; the reports are presented in the form of charts, such as timelines that show the node sequence of the verification path, making it easier for the tendering party to quickly understand the review process.

[0163] It should be noted that the on-chain storage of the review results ensures their immutability, and any party can retrieve the original data for verification via the storage address. This method significantly improves the credibility and operability of the assessment.

[0164] In one embodiment, the review results can also be correlated with historical bidding data to generate a statistical analysis report. For example, analyzing a company's review pass rate in multiple bidding processes can assess its stability and competitiveness. The report might show Company A: review pass rate 90%; average score 0.85, providing decision-making reference for the bidding party. This correlation analysis further enhances the comprehensiveness of the evaluation and optimizes the bidding process.

[0165] In the description of this invention, it should be understood that the terms "coaxial," "bottom," "one end," "top," "middle," "other end," "upper," "side," "top," "inner," "front," "center," "both ends," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0166] Furthermore, the terms “first,” “second,” “third,” and “fourth” are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as “first,” “second,” “third,” or “fourth” may explicitly or implicitly include at least one of those features.

[0167] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "setting," "connection," "fixing," "screw connection," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal connection of two components or the interaction between two components. Unless otherwise explicitly limited, those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0168] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A management method for an engineering bidding platform, characterized in that, include: The requirement description text is obtained from the on-chain storage of the bidding project through blockchain nodes, and the keyword sequence is extracted by natural language processing to obtain the feature vector of the bidding requirements. Based on the bidding requirement feature vector, the qualification certificate data submitted by the bidding companies are vectorized, and the similarity between the transformed vector and the requirement feature vector is judged. If the similarity exceeds a preset threshold, the qualification is initially matched. Performance indicator data is extracted from the historical performance records of bidding companies, and the performance indicator data is fused with the preliminary qualification matching results using the vector embedding method to obtain a comprehensive qualification vector. The system uses a smart contract to call an external oracle interface to obtain real-time updates of the bidding companies' financial statements, and determines whether the financial statement update time is before the bidding deadline. If it is before the bidding deadline, the validity of the financial data is confirmed. Based on the comprehensive qualification vector and the validity results of the financial data, a weighted fusion calculation is performed to obtain the dynamic verification score of the bidding enterprise; If the dynamic verification score is higher than the preset threshold, the automatic approval logic is executed through the smart contract, and the verification process record is stored on the chain using a hash function to obtain an immutable audit log. Based on the immutable audit log, relevant verification paths are retrieved from the blockchain network to determine the completeness of the paths. If the paths are complete, the final audit results are determined and output.

2. The engineering bidding platform management method according to claim 1, characterized in that: The steps of obtaining the requirement description text from the blockchain storage of the bidding project through blockchain nodes, extracting keyword sequences using natural language processing, and obtaining the bidding requirement feature vector include: The tender project requirement description text stored on the blockchain node is obtained, and the raw text data is obtained by using a distributed query protocol. The word segmentation technique in natural language processing is used to segment the original text data to obtain a word sequence; If the word sequence contains preset stop words, then the stop words are filtered out to obtain a filtered keyword sequence; By using a pre-trained word embedding model, the keyword sequence is transformed into a word vector sequence, generating an initial feature vector; Based on the initial feature vector, the principal component analysis algorithm is used to reduce the dimension of the vector to obtain the optimized feature vector of bidding requirements. If the similarity between the optimized feature vector and the preset business requirement template exceeds a preset threshold, then the feature vector is determined to meet the business requirements, and a structured requirement description is generated. By using structured requirement descriptions, clustering algorithms are employed to classify bidding projects and obtain classification results.

3. The engineering bidding platform management method according to claim 1, characterized in that: The step of performing vector transformation on the qualification certificate data submitted by the bidding companies based on the bidding requirement feature vector, determining the similarity between the transformed vector and the requirement feature vector, and determining preliminary qualification matching if the similarity exceeds a preset threshold includes: Obtain bidding requirement data and extract feature vectors from it. Use keyword extraction and semantic analysis methods to generate requirement feature vectors. The qualification certificate data submitted by bidding companies is structured, and text parsing technology is used to extract key fields and generate qualification data. The qualification data is converted into qualification feature vectors through a pre-trained word vector model. Calculate the cosine similarity between the qualification feature vector and the demand feature vector to obtain a similarity value; If the similarity value exceeds a preset threshold, a preliminary matching of qualifications is determined, and a matching result is generated; Based on the matching results, cluster analysis was used to group the bidding companies and obtain the grouping results. Based on the grouping results, extract the qualification feature vectors of each group of bidding companies and generate a sorted list.

4. The engineering bidding platform management method according to claim 1, characterized in that: The steps of extracting performance indicator data from the historical performance records of bidding companies and fusing the performance indicator data with the preliminary qualification matching results using a vector embedding method to obtain a comprehensive qualification vector include: Performance indicator data is extracted from the historical performance records of the bidding companies on the blockchain, and a structured performance indicator dataset is obtained by using preset indicator extraction rules. By using a structured performance indicator dataset, the weight values ​​of each indicator are calculated, and a weighted average method is used to generate a weighted set of performance indicators. Enterprise qualification data is obtained from the initial qualification matching results. The vector embedding method is used to merge the weighted set of performance indicators with the enterprise qualification data to generate an initial comprehensive qualification vector. If the dimension of the initial comprehensive qualification vector exceeds the preset threshold, principal component analysis is used to reduce the dimension and obtain the optimized comprehensive qualification vector. Based on the optimized comprehensive qualification vector, the similarity between the vectors is calculated, and the cosine similarity algorithm is used to determine the degree of qualification matching between enterprises. By assessing the degree of qualification matching between enterprises and combining it with preset classification rules, the qualification level of enterprises is determined, and the final qualification assessment result is generated. Based on the final qualification assessment results, a data correlation analysis method is used to link the assessment results with on-chain historical performance records to generate a verification dataset of comprehensive qualification vectors.

5. The engineering bidding platform management method according to claim 1, characterized in that: The step of obtaining real-time financial statement updates of bidding companies by calling an external oracle interface through a smart contract, determining whether the financial statement update time is before the bidding deadline, and confirming the validity of the financial data if it is before the bidding deadline includes: Obtain the update timestamps of the financial statements of bidding companies from external data sources through the oracle interface and store them as first-time data; Using time comparison logic, the data at the first time is compared with the preset bid deadline. If the data at the first time is earlier than the bid deadline, the update time of the financial statements is determined to be valid, and the validity judgment result is obtained. The validity determination result is obtained through the smart contract interface, and the validity status of the financial data is generated. Based on the validity status, call the oracle interface to obtain the complete dataset of the financial statements and store it as the first financial data; The integrity and consistency of the first financial data are verified through a data validation algorithm to obtain the second financial data that has passed the verification. A classification algorithm is used to assess the financial health of the second set of financial data and generate a financial health score. The final validity of the bidding company's financial data is determined based on the financial health score and validity assessment results.

6. The engineering bidding platform management method according to claim 1, characterized in that: The step of performing a weighted fusion calculation based on the comprehensive qualification vector and the validity results of the financial data to obtain the dynamic verification score of the bidding enterprise includes: The system obtains the validity of bidding company information and financial data, and uses preset data verification rules to determine the integrity and consistency of the data, thereby obtaining verified financial data quality. Based on the verified quality of financial data, qualification assessment indicators are extracted, and factor analysis is used to determine the comprehensive qualification vector. If at least one indicator in the comprehensive qualification vector exceeds the preset threshold, a preliminary integrated score is obtained by integrating the qualification assessment indicators and financial data quality through a linear weighting method. For the initial fusion score, a weighted fusion calculation is performed using a preset fusion weight allocation to obtain the adjusted fusion score; Based on the adjusted fusion score, a logistic regression model is applied to determine the enterprise's credit rating and output a dynamic validation score. If the dynamic verification score is lower than the preset threshold, the fusion weight allocation will be adjusted through the score calculation model, and the dynamic verification score will be recalculated. Obtain the recalculated dynamic verification score, and generate the comprehensive evaluation score for the bidding company based on the evaluation results output.

7. The engineering bidding platform management method according to claim 1, characterized in that: The step of executing automatic approval logic through a smart contract if the dynamic verification score is higher than a preset threshold, and using a hash function to store the verification process records on-chain to obtain an immutable audit log, includes: If the verification score is higher than the preset threshold, the data to be approved will be obtained through the triggering logic of the smart contract, the automatic approval operation will be executed, and the approval result will be obtained. The approval results and verification records are encrypted using a hash function to generate encrypted data; Encrypted data is written to on-chain storage via the blockchain network, generating a storage address; Retrieve on-chain data based on the storage address, verify data integrity, and obtain verification status; If the verification status is complete, an audit log is generated through a smart contract to record approval and storage information; Based on the audit log content, log data is distributed using a decentralized network to generate distribution confirmation; By distributing confirmations, we obtain consensus feedback from network nodes and ultimately obtain the on-chain recorded state.

8. The engineering bidding platform management method according to claim 1, characterized in that: The step of retrieving relevant verification paths from the blockchain network based on the immutable audit log, determining the completeness of the paths, and if complete, determining the final output of the audit results includes: The audit logs are obtained from the blockchain network, and a distributed node query protocol is used to extract the log dataset containing timestamps and transaction hashes to obtain the original audit logs. For the original audit logs, the log structure is parsed, the node sequence and hash chain in the verification path are extracted, and a path dataset is generated; If the node sequence in the path dataset is consistent with the preset blockchain topology, the Merkle tree verification algorithm is used to calculate the hash value of each node in the path, determine the path integrity, and obtain the integrity verification result. Based on the integrity verification results, if the integrity verification passes, the transaction timestamps and associated metadata are extracted from the path dataset to generate the review dataset; The SHA-256 algorithm is used to perform hash calculation on the review dataset to generate review hash values ​​and determine the uniqueness of the review results; By using network communication protocols, the audit hash value is compared with the preset audit standards to determine whether the audit result meets the standards and output the final audit result.

9. A project bidding platform management system, characterized in that, include: The requirement feature extraction module is configured to obtain the requirement description text from the blockchain storage of the bidding project chain through a blockchain node, extract the keyword sequence using natural language processing, and obtain the bidding requirement feature vector. The qualification matching module is configured to perform vector transformation processing on the qualification certificate data submitted by the bidding company based on the bidding requirement feature vector, determine the similarity between the transformed vector and the requirement feature vector, and determine the preliminary qualification matching if the similarity exceeds a preset threshold. The performance fusion module is configured to extract performance indicator data from historical performance records on the bidding enterprise chain, and use a vector embedding method to fuse the performance indicator data with the preliminary qualification matching results to obtain a comprehensive qualification vector. The financial verification module is configured to obtain the real-time financial statement updates of the bidding company by calling the external oracle interface through a smart contract, determine whether the financial statement update time is before the bidding deadline, and determine the validity of the financial data if it is before the bidding deadline. The scoring calculation module is configured to perform a weighted fusion calculation based on the comprehensive qualification vector and the validity results of the financial data to obtain the dynamic verification score of the bidding enterprise. The contract execution module is configured to execute automatic approval logic through a smart contract if the dynamic verification score is higher than a preset threshold, and to use a hash function to store the verification process record on the chain to obtain an immutable audit log. The audit verification module is configured to retrieve relevant verification paths from the blockchain network based on the tamper-proof audit logs, determine the completeness of the paths, and if complete, determine the final output of the audit results.

Images