Tender opportunity prediction and collaborative decision-making method based on multi-source announcement data

Abstract

The application discloses a bidding opportunity prediction and collaborative decision-making method based on multi-source announcement data, relates to the technical field of intelligent bidding, and comprises the following steps: collecting bidding announcement data from no less than two information sources, extracting core information after format conversion, and generating a structured field in JSON format; carrying out integrity and consistency verification on the extracted structured field in sequence, and completing unit and format standardization modification of digital fields; based on the standardized field data, calculating project scale scores and market competition scores respectively, comprehensively obtaining opportunity scores, and dividing potential levels; collecting enterprise multidimensional internal resource data to calculate bearing capacity scores, combining the opportunity scores to calculate bidding scores, and making collaborative decisions.

Description

Technical Field

[0001] This invention relates to the field of intelligent bidding technology, specifically a method for predicting bidding opportunities and making collaborative decisions based on multi-source announcement data. Background Technology

[0002] With the deepening of digital transformation, bidding activities in fields such as government procurement, engineering construction, and IT services are becoming increasingly frequent, involving huge sums of money. However, traditional bidding opportunity acquisition and decision-making methods have many pain points: information acquisition is not timely or comprehensive, making it difficult for enterprises to quickly identify business opportunities that meet their own conditions; the decision-making process relies on experience-based judgment and lacks the support of scientific predictive models; resource allocation efficiency is low, and collaborative optimization among multiple projects cannot be achieved.

[0003] The Chinese invention patent application with publication number CN120725725A and publication date of September 30, 2025, automatically filters out vessels with high-potential ship repair opportunities by collecting multi-source data, including basic ship data, market dynamic data, and internal enterprise data, and setting multi-dimensional filtering rules such as special inspection window rules, geofencing rules, and tonnage restriction rules. It also dynamically calculates the total score of each vessel in the preliminary opportunity list by constructing data from multiple dimensions such as historical ship repair record scores, historical frequently used route record scores, and customer level scores, combined with static weights and dynamic adjustment mechanisms. Finally, it uses natural language processing technology combined with knowledge base and intent recognition conversion methods to automatically convert the user's natural language query request into structured query conditions to filter out ship matching results, thereby realizing the prediction of ship repair opportunities.

[0004] The above invention applications collect ship data, market dynamics data, and internal enterprise data, set rules, and construct historical individual indicator data of ships to calculate results. Finally, they use matching to predict ship repair business opportunities. However, the accuracy of multi-source heterogeneous data parsing is insufficient, there is a lack of a standardized quality assessment system, OCR recognition of complex PDF content is prone to missing information, and there is a delay in capturing change notices. The prediction model relies on a single algorithm and feature, fails to mine temporal change information, lacks a sample balancing strategy, has poor generalization and interpretability, lacks a cross-departmental resource collaboration mechanism, only covers a single link, lacks a data feedback loop, has weak cross-domain adaptability, is insufficient in response to dynamic policy adjustments, and lacks compliance risk control.

[0005] To this end, the present invention provides a method for predicting bidding opportunities and making collaborative decisions based on multi-source announcement data. Summary of the Invention

[0006] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a method for predicting bidding opportunities and collaborative decision-making based on multi-source announcement data. By constructing a standardized process for collecting, parsing, verifying, and correcting multi-source announcement data, it achieves high-quality structured processing of heterogeneous data, solving problems of information loss and format confusion, and ensuring the accuracy of basic data. By establishing a dual-dimensional quantitative scoring system based on project scale and competition, it achieves scientific hierarchical prediction of bidding opportunities, overcoming the limitations of experience-based judgment. By integrating multi-dimensional resource data on enterprise personnel, finance, projects, and region to construct a capacity model, and combining it with opportunity scoring to calculate a comprehensive bidding score, it achieves accurate matching of opportunities with enterprise capabilities and cross-departmental collaborative decision-making. Ultimately, it improves the efficiency of enterprise opportunity identification and the scientific nature of bidding decisions, optimizes resource allocation, and reduces bidding risks.

[0007] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: a method for predicting bidding opportunities and collaborative decision-making based on multi-source announcement data, comprising: S1. Collect bidding announcement data from no less than two information sources, extract core information after format conversion, and generate structured fields in JSON format; S2. Perform integrity and consistency checks on the extracted structured fields in sequence, and complete the standardization and correction of the units and formats of the numeric fields; S3. Based on standardized field data, calculate the project size score and market competition score respectively, and combine them to obtain the business opportunity score and classify the potential level. S4. Collect multi-dimensional internal resource data of enterprises to calculate the affordability score, combine it with the business opportunity score to calculate the bidding score and make collaborative decisions.

[0008] Preferably, the multi-source data announcement is obtained from no less than two information sources, the multi-source data announcement including source domain name, announcement title, publication time, announcement text, attachment list, and original link; the format of the multi-source data announcement is identified, the format including PDF, plain text file, HTML text and compressed file, and the first multi-source data announcement is converted into a Markdown document format first multi-source data announcement through Pandoc, PDF parsing component or OCR component.

[0009] Preferably, the core information is obtained by reading the announcement text in the first multi-dimensional data announcement. The core information includes the project name, purchaser, number of candidates, budget amount, bid opening time, registration deadline, region, and project duration. The Tabulate tool is used to convert the core information into the form of structured fields.

[0010] Structured fields are collections of key-value pairs output according to a preset set of structured fields, in JSON format, such as: Project Name: Intelligent Air Conditioning Equipment Purchaser: "Company X" Budget Amount: 5 million yuan Preferred structured field integrity validation: Obtain the number of filled fields and the preset number of fields. The number of filled fields refers to the number of structured fields extracted and successfully filled from the first multi-source data announcement. The preset number of fields is the total number of fields to be extracted according to the announcement format. Calculate the field completeness rate.

[0011] like A value of ≥0.85 indicates that the field extraction is complete and of high quality, and S3 and S4 can be performed.

[0012] like If the value is less than 0.85, it means that some fields are missing or have not been successfully extracted, and you need to return to the S1 field extraction stage to extract the fields again.

[0013] Preferably, when the same field is extracted from different data sources, multiple different values ​​may appear. The number of conflicting fields and the number of filled fields are obtained. The number of conflicting fields refers to the number of different values ​​extracted for the same field from different data sources, and the number of filled fields refers to the number of fields that have been successfully extracted and filled. The field conflict rate is calculated.

[0014] when A value > 0.05 indicates a significant conflict in the structured field, which may affect the accuracy of decision-making. Manual intervention is required, and the field extraction should be completed by dedicated personnel reviewing the first multi-source data announcement.

[0015] when If the value is ≤0.05, then the last extracted field from the conflicting fields will be used as the new field.

[0016] Preferably, for structured fields containing numerical values, the units are standardized, with monetary amounts standardized to ten thousand yuan and time units standardized to days, while the units for other numerical fields are adjusted according to the actual situation.

[0017] Preferably, the project budget and duration are obtained through structured fields, and the project size score is calculated. The project budget amount is obtained through structured fields and divided into 5 levels, corresponding to different scoring ranges: Budget amount < 1 million yuan, score 0-10 points; 1 million yuan ≤ budget amount < 5 million yuan, score 10-20 points; 5 million yuan ≤ budget amount < 1 million yuan, score 20-30 points; 10 million yuan ≤ budget amount < 50 million yuan, score 30-40 points; Budget amount ≥ 50 million yuan, score 40-50 points. The formula is: Budget amount score = Lowest score in the current range + (Current budget amount - Lowest score in the current range) / (Highest score in the current range - Lowest score in the current range) × Difference in scoring range. Project duration is obtained through structured fields and divided into 5 levels, corresponding to different scoring ranges: Project duration < 10 days, score 0-5; 10 days ≤ Project duration < 30 days, score 5-10; 30 days ≤ Project duration < 90 days, score 10-15; 90 days ≤ Project duration < 180 days, score 15-20; Project duration ≥ 180 days, score 20-25. The formula is: Project duration score = Lowest score in the current range + (Current project duration - Lowest level value) / (Highest level value - Lowest level value) × Difference in scoring range; Calculate the project size score: Project size score = ×Budget Amount Score+ ×Project schedule score; in, and These are the weighting coefficients.

[0018] Preferably, the number of candidates and the application deadline for a project are obtained through structured fields, and the project's competition score is calculated. The number of candidates set in the project bidding is obtained through structured fields. The fewer the candidates, the more moderate the competition and the higher the score. The number of candidates is divided into 5 levels, corresponding to different score ranges: ≤3 candidates, score 40-50 points; 3 < ≤5 candidates, score 30-40 points; 5 < ≤8 candidates, score 20-30 points; 8 < ≤10 candidates, score 10-20 points; >10 candidates, score 0-10 points. The candidate score = the lowest score in the current range + (current number of candidates - lowest score in the current level) / (highest score in the current level - lowest score in the current level) × the difference in the score range. The remaining registration time is calculated by obtaining the time difference between the project registration deadline and the current time using structured fields. A longer remaining registration time indicates that the company has more time to prepare its tender documents, resulting in a stronger competitive advantage and a higher score. The remaining registration time is divided into 5 levels, corresponding to different scoring ranges: ≥15 days remaining registration time, score 20-25 points; 10 days ≤ remaining registration time <15 days, score 15-20 points; 5 days ≤ remaining registration time <10 days, score 10-15 points; 3 days ≤ remaining registration time <5 days, score 5-10 points; <3 days remaining registration time, score 0-5 points. The registration time score is calculated as: (Lowest score in the current range) + (Current remaining registration time - Lowest level value) / (Highest level value - Lowest level value) × Difference in scoring range.

[0019] Calculate project competitiveness score: Project size score = ×Candidate Count Score+ ×Registration time slots; in, and These are the weighting coefficients.

[0020] Preferably, the business opportunity score is calculated by combining the project size score and the project competition score:

[0021] The business opportunity scoring formula is as described above.

[0022] Preferably, obtain internal company data and calculate the company's resilience: By collecting core data from four dimensions of internal enterprise resources—human resources, financial resources, project resources, and geographical resources—a score is calculated for each dimension, and then the weighted sum is used to obtain a comprehensive score of the company's resilience. The weights of the four dimensions are set according to the company's core competitiveness. Company resilience = Human resources score × 0.3 + Financial resources score × 0.3 + Project resources score × 0.2 + Geographical resources score × 0.2.

[0023] Human Resources Score (0-100 points): Collect data on the number of professional and technical personnel in the category of this project within the enterprise, the availability rate of personnel, and the matching degree of personnel qualifications. A full score of 100 points is obtained when the number of professional and technical personnel is greater than or equal to the number required by the project, the availability rate is greater than or equal to 80%, and the matching degree of qualifications is 100%. If any indicator is not met, points will be deducted according to the missing proportion.

[0024] Financial Resources Score (0-100 points): Data such as the company's available bidding budget, cash flow capacity, and project performance fund reserves are collected. A company can get a full score of 100 points when its available bidding budget is greater than or equal to the budget required for the project bidding, its cash flow capacity is greater than or equal to 80%, and its performance fund reserves are greater than or equal to 30% of the project budget. If any indicator is not met, points will be deducted according to the missing proportion.

[0025] Project Resource Rating (0-100 points): Data such as the number of projects being executed, remaining capacity, and equipment availability are collected from within the enterprise. A perfect score of 100 points is awarded when the number of projects being executed does not exceed the enterprise's capacity, remaining capacity is greater than or equal to the capacity required for project execution, and equipment availability is greater than or equal to 90%. If any indicator is not met, points will be deducted according to the proportion of the missing indicators.

[0026] Regional resource score (0-100 points): Based on the region where the project is located, data such as the company's offline service outlets, localized teams, and logistics and distribution capabilities in the region are collected. A full score of 100 points is awarded when the project has a complete set of offline service outlets, a fully equipped localized team, and logistics and distribution coverage throughout the region. If a certain indicator is not met, points will be deducted according to the proportion of the missing indicators.

[0027] After the scores for each dimension are calculated, a comprehensive score of the company's resilience is obtained according to a weighted formula. Based on the score results, the company's resilience is divided into three levels: high resilience (score ≥ 80 points), medium resilience (60 points ≤ score < 80 points), and low resilience (score < 60 points), which correspond to the company's ability to undertake the project as "fully capable", "basically capable, requiring a small amount of resource allocation", and "not capable, requiring a large amount of resource investment", respectively.

[0028] Preferably, a collaborative decision is made by combining the opportunity score and the company's affordability to calculate the bid score. The weight of both the opportunity score and the company's affordability is set to 0.5. The calculation formula is: Bid score = Opportunity score × 0.5 + Company affordability × 0.5.

[0029] Based on the comprehensive bid scoring results, and combined with the company's bidding strategy and risk appetite, standardized collaborative decision-making rules are formulated to output three types of decision recommendations: bid, bid cautiously, and abandon bid. The specific decision criteria are as follows: Bidding: A comprehensive bid score of ≥80 points indicates that the project is a high-potential business opportunity and the company has a high capacity to bear the risk. The project has high comprehensive value and low execution risk after winning the bid. It is recommended that the company prioritize resources and fully participate in the bidding. Bidding with caution: A score of 60 to 80 indicates that the project is a medium / high potential business opportunity but the company has medium risk tolerance, or a medium potential business opportunity but the company has high risk tolerance. The project has certain value but requires resource allocation or involves certain competitive risks. It is recommended that companies carefully participate in bidding after optimizing their resource allocation plan based on their own business layout, and at the same time formulate risk control measures. Abandoning the bid: If the overall bid score is less than 60 points, the corresponding project is a low-potential business opportunity or the company has low affordability. The project value is low or the company has no ability to undertake it. After winning the bid, the company is likely to encounter problems such as insufficient resources and difficulties in fulfilling the contract. It is recommended that the company abandon the project directly and invest its resources in higher-value business opportunities.

[0030] (III) Beneficial Effects This invention provides a method for predicting bidding opportunities and making collaborative decisions based on multi-source announcement data, which has the following beneficial effects: 1. Construct a business opportunity scoring system with two dimensions: project scale and project competitiveness. Transform bidding opportunities into quantitative indicators of 0-100 points and divide them into three potential levels: high, medium, and low. This enables standardized and visualized evaluation of business opportunities, allowing companies to quickly screen high-potential projects, significantly improving the efficiency and accuracy of business opportunity identification and avoiding missed or misjudgments.

[0031] 2. Achieve precise matching between project opportunities and corporate capabilities, and solve the problem of disconnect between decision-making and resources: By combining data from four dimensions of internal human resources, financial resources, project resources, and regional resources, a company's capacity assessment model is constructed. Then, through a comprehensive bidding scoring system, the opportunity score is combined with the company's capacity score to achieve two-way matching between the "project side" and the "corporate side". This avoids the waste of resources or contract performance risks caused by blind bidding, and improves the scientific nature and feasibility of bidding decisions.

[0032] 3. Provide standardized and customizable collaborative decision-making solutions to suit the needs of different enterprises: Based on comprehensive bidding scores, formulate three standardized decision suggestions: "bid, bid cautiously, and abandon bidding", and generate corresponding implementation suggestions such as resource allocation, bidding strategy, and risk control. At the same time, it supports enterprises to adjust the scoring weights and decision-making standards according to their own development strategies and risk preferences, adapting to the bidding decision-making needs of enterprises in different industries and at different stages of development, with strong cross-domain adaptability.

[0033] 4. Simple and efficient process, easy to implement in software and hardware: All steps of this method are based on standardized algorithms and formulas, without complex model design. It can be implemented through software, hardware, firmware or any combination thereof. It can be directly embedded into the enterprise's existing bidding management system and project management system to realize automatic data collection, automatic calculation and automatic output of decision suggestions, which greatly reduces the enterprise's manual operation costs and improves the efficiency of the entire bidding business process. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the bidding opportunity prediction and collaborative decision-making method based on multi-source announcement data according to the present invention. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] Please see Figure 1 This invention provides a method for predicting bidding opportunities and making collaborative decisions based on multi-source announcement data, comprising the following steps: S1. Collect bidding announcement data from no less than two information sources, extract core information after format conversion, and generate structured fields in JSON format.

[0037] S1.1 Obtain multi-source data announcements from no less than two information sources, wherein the multi-source data announcements include source domain name, announcement title, publication time, announcement body, attachment list, and original link; identify the format of the multi-source data announcements, wherein the format includes PDF, plain text file, HTML text, and compressed file; and convert the first multi-source data announcement into a Markdown document format first multi-source data announcement through Pandoc, PDF parsing component, or OCR component.

[0038] S1.2 Read the announcement text from the first multi-dimensional data announcement to obtain core information, including project name, purchaser, number of candidates, budget amount, bid opening time, registration deadline, region, and project duration. Use the Tabulate tool to convert the core information into structured fields.

[0039] Structured fields are collections of key-value pairs output according to a preset set of structured fields, in JSON format, such as: Project Name: Intelligent Air Conditioning Equipment Purchaser: "Company X" Budget Amount: 5 million yuan S2. Perform integrity and consistency checks on the extracted structured fields in sequence, and complete the standardization and correction of the units and formats of the numeric fields.

[0040] S201. Obtain the number of filled fields and the preset number of fields. The number of filled fields refers to the number of structured fields extracted and successfully filled from the first multi-dimensional data announcement. The preset number of fields is the total number of fields to be extracted according to the announcement format. Calculate the field completeness rate.

[0041] like A value of ≥0.85 indicates that the extracted fields are complete and of high quality, suitable for business opportunity screening and decision-making.

[0042] like If the value is less than 0.85, it means that some fields are missing or have not been successfully extracted. It is necessary to return to the S1 field extraction stage for supplementation or manual review.

[0043] In one possible implementation, the preset number of fields is 20.

[0044] S202. When the same field is extracted from different data sources, multiple different values ​​may appear. Obtain the number of conflicting fields and the number of filled fields. The number of conflicting fields refers to the number of different values ​​extracted from the same field in different data sources, and the number of filled fields refers to the number of fields that have been successfully extracted and filled. Calculate the field conflict rate.

[0045] when A value > 0.05 indicates a significant conflict in the structured field, which may affect the accuracy of decision-making. Manual intervention is required, and the field extraction should be completed by dedicated personnel reviewing the first multi-source data announcement.

[0046] when If the value is ≤0.05, then the last extracted field from the conflicting fields will be used as the new field.

[0047] S203. Identify structured fields containing numerical values, unify the units, unify the unit of amount to ten thousand yuan, unify the unit of time to day, and adjust the units of other numerical fields according to the actual situation.

[0048] S3. Based on standardized field data, calculate the project size score and market competition score respectively, and combine them to obtain the business opportunity score and classify the potential level.

[0049] S301. Obtain the project budget and project duration using structured fields, and calculate the project size score: The project budget amount is obtained through structured fields and divided into 5 levels, corresponding to different scoring ranges: Budget amount < 1 million yuan, score 0-10 points; 1 million yuan ≤ budget amount < 5 million yuan, score 10-20 points; 5 million yuan ≤ budget amount < 1 million yuan, score 20-30 points; 10 million yuan ≤ budget amount < 50 million yuan, score 30-40 points; Budget amount ≥ 50 million yuan, score 40-50 points. The formula is: Budget amount score = Lowest score in the current range + (Current budget amount - Lowest score in the current range) / (Highest score in the current range - Lowest score in the current range) × Difference in scoring range. Project duration is obtained through structured fields and divided into 5 levels, corresponding to different scoring ranges: Project duration < 10 days, score 0-5; 10 days ≤ Project duration < 30 days, score 5-10; 30 days ≤ Project duration < 90 days, score 10-15; 90 days ≤ Project duration < 180 days, score 15-20; Project duration ≥ 180 days, score 20-25. The formula is: Project duration score = Lowest score in the current range + (Current project duration - Lowest level value) / (Highest level value - Lowest level value) × Difference in scoring range; Calculate the project size score: Project size score = ×Budget Amount Score+ ×Project schedule score; in, and As a weighting coefficient, in one possible embodiment, =0.6, =0.4.

[0050] S302. Obtain the number of candidates and application deadline for the project using structured fields, and calculate the project's competition score: The number of candidates set in the project bidding is obtained through structured fields. The fewer the candidates, the more moderate the competition and the higher the score. The number of candidates is divided into 5 levels, corresponding to different score ranges: ≤3 candidates, score 40-50 points; 3 < ≤5 candidates, score 30-40 points; 5 < ≤8 candidates, score 20-30 points; 8 < ≤10 candidates, score 10-20 points; >10 candidates, score 0-10 points. The candidate score = the lowest score in the current range + (current number of candidates - lowest score in the current level) / (highest score in the current level - lowest score in the current level) × the difference in the score range. The remaining registration time is calculated by obtaining the time difference between the project registration deadline and the current time using structured fields. A longer remaining registration time indicates that the company has more time to prepare its tender documents, resulting in a stronger competitive advantage and a higher score. The remaining registration time is divided into 5 levels, corresponding to different scoring ranges: ≥15 days remaining registration time, score 20-25 points; 10 days ≤ remaining registration time <15 days, score 15-20 points; 5 days ≤ remaining registration time <10 days, score 10-15 points; 3 days ≤ remaining registration time <5 days, score 5-10 points; <3 days remaining registration time, score 0-5 points. The registration time score is calculated as: (Lowest score in the current range) + (Current remaining registration time - Lowest level value) / (Highest level value - Lowest level value) × Difference in scoring range.

[0051] Calculate project competitiveness score: Project size score = ×Candidate Count Score+ ×Registration time slots; in, and As a weighting coefficient, in one possible embodiment, =0.6, =0.4.

[0052] S303. Calculate the business opportunity score by combining the project size score and the project competition score:

[0053] The business opportunity scoring formula is as described above.

[0054] S4. Collect multi-dimensional internal resource data of enterprises to calculate the affordability score, combine it with the business opportunity score to calculate the bidding score and make collaborative decisions.

[0055] S401. Obtain internal company data and calculate the company's resilience: By collecting core data from four dimensions of internal enterprise resources—human resources, financial resources, project resources, and geographical resources—a score is calculated for each dimension, and then the weighted sum is used to obtain the comprehensive score of the company's resilience. The weights of the four dimensions are set according to the company's core competitiveness. Company resilience score = Human resources score × 0.3 + Financial resources score × 0.3 + Project resources score × 0.2 + Geographical resources score × 0.2.

[0056] Human Resources Score (0-100 points): Collect data on the number of professional and technical personnel in the category of this project within the enterprise, the availability rate of personnel, and the matching degree of personnel qualifications. A full score of 100 points is obtained when the number of professional and technical personnel is greater than or equal to the number required by the project, the availability rate is greater than or equal to 80%, and the matching degree of qualifications is 100%. If any indicator is not met, points will be deducted according to the missing proportion.

[0057] Financial Resources Score (0-100 points): Data such as the company's available bidding budget, cash flow capacity, and project performance fund reserves are collected. A company can get a full score of 100 points when its available bidding budget is greater than or equal to the budget required for the project bidding, its cash flow capacity is greater than or equal to 80%, and its performance fund reserves are greater than or equal to 30% of the project budget. If any indicator is not met, points will be deducted according to the missing proportion.

[0058] Project Resource Rating (0-100 points): Data such as the number of projects being executed, remaining capacity, and equipment availability are collected from within the enterprise. A perfect score of 100 points is awarded when the number of projects being executed does not exceed the enterprise's capacity, remaining capacity is greater than or equal to the capacity required for project execution, and equipment availability is greater than or equal to 90%. If any indicator is not met, points will be deducted according to the proportion of the missing indicators.

[0059] Regional resource score (0-100 points): Based on the region where the project is located, data such as the company's offline service outlets, localized teams, and logistics and distribution capabilities in the region are collected. A full score of 100 points is awarded when the project has a complete set of offline service outlets, a fully equipped localized team, and logistics and distribution coverage throughout the region. If a certain indicator is not met, points will be deducted according to the proportion of the missing indicators.

[0060] After the scores for each dimension are calculated, a comprehensive score of the company's resilience is obtained according to a weighted formula. Based on the score results, the company's resilience is divided into three levels: high resilience (score ≥ 80 points), medium resilience (60 points ≤ score < 80 points), and low resilience (score < 60 points), which correspond to the company's ability to undertake the project as "fully capable", "basically capable, requiring a small amount of resource allocation", and "not capable, requiring a large amount of resource investment", respectively.

[0061] S402. Combine opportunity score and company affordability score to calculate bid score for collaborative decision-making. Set the weight of both opportunity score and company affordability score to 0.5. The calculation formula is: Bid score = Opportunity score × 0.5 + Company affordability score × 0.5.

[0062] Based on the comprehensive bid scoring results, and combined with the company's bidding strategy and risk appetite, standardized collaborative decision-making rules are formulated to output three types of decision recommendations: bid, bid cautiously, and abandon bid. The specific decision criteria are as follows: Bidding: A comprehensive bid score of ≥80 points indicates that the project is a high-potential business opportunity and the company has a high capacity to bear the risk. The project has high comprehensive value and low execution risk after winning the bid. It is recommended that the company prioritize resources and fully participate in the bidding. Bidding with caution: A score of 60 to 80 indicates that the project is a medium / high potential business opportunity but the company has medium risk tolerance, or a medium potential business opportunity but the company has high risk tolerance. The project has certain value but requires resource allocation or involves certain competitive risks. It is recommended that companies carefully participate in bidding after optimizing their resource allocation plan based on their own business layout, and at the same time formulate risk control measures. Abandoning the bid: If the overall bid score is less than 60 points, the corresponding project is a low-potential business opportunity or the company has low affordability. The project value is low or the company has no ability to undertake it. After winning the bid, the company is likely to encounter problems such as insufficient resources and difficulties in fulfilling the contract. It is recommended that the company abandon the project directly and invest its resources in higher-value business opportunities.

[0063] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0064] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0065] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for predicting bidding opportunities and collaborative decision-making based on multi-source announcement data, characterized in that, include: S1. Collect bidding announcement data from no less than two information sources, extract core information after format conversion, and generate structured fields in JSON format; S2. Perform integrity and consistency checks on the extracted structured fields in sequence, and complete the standardization and correction of the units and formats of the numeric fields; S3. Based on standardized field data, calculate the project size score and market competition score respectively, and combine them to obtain the business opportunity score and classify the potential level. S4. Collect multi-dimensional internal resource data of enterprises to calculate the affordability score, combine it with the business opportunity score to calculate the bidding score and make collaborative decisions.

2. The bidding opportunity prediction and collaborative decision-making method based on multi-source announcement data according to claim 1, characterized in that: In S1, the collection of multi-source announcement data and extraction of structured fields are as follows: Obtain multi-source data announcements from no fewer than two information sources, wherein the multi-source data announcements include source domain name, announcement title, publication time, announcement body, attachment list, and original link; identify the format of the multi-source data announcements, wherein the formats include PDF, plain text file, HTML text, and compressed file; and convert the first multi-source data announcement into a Markdown document format first multi-source data announcement through Pandoc, PDF parsing component, or OCR component. The core information is obtained by reading the announcement text in the first multi-dimensional data announcement. The core information includes project name, purchaser, number of candidates, budget amount, bid opening time, registration deadline, region, and project duration. The Tabulate tool is used to convert the core information into the form of structured fields. Structured fields are collections of key-value pairs of fields output according to a preset set of structured fields, and the output format is JSON.

3. The bidding opportunity prediction and collaborative decision-making method based on multi-source announcement data according to claim 1, characterized in that: In S2, the structured field validation correction is as follows: Obtain the number of filled fields and the preset number of fields. The number of filled fields refers to the number of structured fields extracted and successfully filled from the first multi-source data announcement. The preset number of fields is the total number of fields to be extracted according to the announcement format. Calculate the field completeness rate.

4. If If the value is ≥0.85, S3 and S4 can be performed.

5. If If the value is less than 0.85, you need to return to the S1 field extraction stage and extract the data again. The number of conflicting fields refers to the number of different values ​​extracted from the same field in different data sources. The number of filled fields refers to the number of fields that have been successfully extracted and filled. The field conflict rate is calculated as follows:

6. When If the value is > 0.05, manual intervention is required, and the field extraction should be completed by designated personnel after reviewing the first multi-source data announcement. when If the value is ≤0.05, then the last extracted field from the conflicting fields will be used as the new field. Identify structured fields containing numerical values, standardize the units, unify the unit for monetary amounts to ten thousand yuan, unify the unit for time to day, and adjust the units for other numerical fields according to the actual situation.

7. The bidding opportunity prediction and collaborative decision-making method based on multi-source announcement data according to claim 1, characterized in that: In S3, the project size score is calculated as follows: Obtain the project's budget and duration using structured fields, and calculate the project size score: The project budget amount is obtained through structured fields and divided into 5 levels, corresponding to different scoring ranges: Budget amount < 1 million yuan, score 0-10 points; 1 million yuan ≤ budget amount < 5 million yuan, score 10-20 points; 5 million yuan ≤ budget amount < 1 million yuan, score 20-30 points; 10 million yuan ≤ budget amount < 50 million yuan, score 30-40 points; Budget amount ≥ 50 million yuan, score 40-50 points. The budget amount score = the lowest score in the current range + (current budget amount - lowest score in the current level) / (highest score in the current level - lowest score in the current level) × difference in the scoring range. Project duration is obtained through structured fields and divided into 5 levels, corresponding to different scoring ranges: Project duration < 10 days, score 0-5; 10 days ≤ Project duration < 30 days, score 5-10; 30 days ≤ Project duration < 90 days, score 10-15; 90 days ≤ Project duration < 180 days, score 15-20; Project duration ≥ 180 days, score 20-25. Project duration score = lowest score in the current range + (current project duration - lowest level value) / (highest level value - lowest level value) × difference in scoring range. Calculate the project size score: Project size score = α1 × budget amount score + α2 × project duration score; Where α1 and α2 are weighting coefficients.

8. The bidding opportunity prediction and collaborative decision-making method based on multi-source announcement data according to claim 1, characterized in that: In S3, the project competition score and business opportunity score are calculated as follows: The number of candidates and the application deadline for a project are obtained using structured fields, and the project's competitiveness score is calculated. The number of candidates set in the project bidding is obtained through structured fields. The fewer the candidates, the more moderate the competition and the higher the score. The number of candidates is divided into 5 levels, corresponding to different score ranges: ≤3 candidates, score 40-50 points; 3 < ≤5 candidates, score 30-40 points; 5 < ≤8 candidates, score 20-30 points; 8 < ≤10 candidates, score 10-20 points; >10 candidates, score 0-10 points. The candidate score = the lowest score in the current range + (current number of candidates - lowest score in the current level) / (highest score in the current level - lowest score in the current level) × the difference in the score range. The remaining registration time is calculated by obtaining the time difference between the project registration deadline and the current time using structured fields. A longer remaining registration time indicates more time for the company to prepare its tender documents, resulting in a stronger competitive advantage and a higher score. The remaining registration time is divided into five levels, corresponding to different scoring ranges: ≥15 days remaining registration time, score 20-25 points; 10 days ≤ remaining registration time <15 days, score 15-20 points; 5 days ≤ remaining registration time <10 days, score 10-15 points; 3 days ≤ remaining registration time <5 days, score 5-10 points; <3 days remaining registration time, score 0-5 points. The registration time score is calculated as: (Lowest score in the current range) + (Current remaining registration time - Lowest level value) / (Highest level value - Lowest level value) × Difference in scoring range. Calculate the project competition score: Project size score = β1 × number of candidates score + β2 × application time score; Where β1 and β2 are weighting coefficients; A business opportunity score is calculated by combining the project size score and the project competitiveness score.

9. The bidding opportunity prediction and collaborative decision-making method based on multi-source announcement data according to claim 1, characterized in that: In S4, internal company data is retrieved, and the company's resilience is calculated as follows: By collecting core data from four dimensions of internal enterprise resources—human resources, financial resources, project resources, and regional resources—scores are calculated for each dimension, and then a weighted sum is obtained to obtain a comprehensive company resilience score. The weights of the four dimensions are set according to the company's core competitiveness. Company resilience score = Human resources score × 0.3 + Financial resources score × 0.3 + Project resources score × 0.2 + Regional resources score × 0.

2. Human Resources Scoring: Data such as the number of professional and technical personnel in the category of this project, the availability rate of personnel, and the matching degree of personnel qualifications are collected within the enterprise. When the number of professional and technical personnel is greater than or equal to the number required by the project, the availability rate is greater than or equal to 80%, and the matching degree of qualifications is 100%, a full score of 100 points is obtained; if any indicator is not met, points are deducted according to the proportion of the missing indicators. Financial Resources Scoring: Data on the company's available bidding budget, cash flow capacity, and project performance funds reserves are collected. A company receives a full score of 100 points when its available bidding budget is greater than or equal to the budget required for the project, its cash flow capacity is greater than or equal to 80%, and its performance funds reserves are greater than or equal to 30% of the project budget. If any indicator is not met, points will be deducted according to the proportion of the missing indicators. Project resource scoring: Data such as the number of projects being executed, remaining capacity, and equipment availability are collected from within the enterprise. A perfect score of 100 points is awarded when the number of projects being executed does not exceed the enterprise's capacity, remaining capacity is greater than or equal to the capacity required for project execution, and equipment availability is greater than or equal to 90%. If any indicator is not met, points will be deducted according to the proportion of the missing indicators. Regional resource scoring: Based on the project location, data such as the company's offline service outlets, localized teams, and logistics and distribution capabilities in the region are collected. A full score of 100 points is awarded when the project location has a company with a complete offline service outlet, a well-equipped localized team, and logistics and distribution coverage throughout the region; points are deducted according to the proportion of missing indicators if any indicator is not met. After the scores for each dimension are calculated, a comprehensive score of the company's resilience is obtained according to the weighted formula. Based on the score results, the company's resilience is divided into three levels: high resilience, medium resilience, and low resilience.

10. The bidding opportunity prediction and collaborative decision-making method based on multi-source announcement data according to claim 1, characterized in that: In S4, a collaborative decision-making process is made by combining opportunity scores and company affordability to calculate bid scores, as follows: Bid score = Business opportunity score × 0.5 + Company affordability score × 0.5; Based on the comprehensive bid scoring results, standardized collaborative decision-making rules are formulated to output three types of decision recommendations: bid, bid cautiously, and abandon bid. The specific decision criteria are as follows: Bidding: Overall bid score ≥ 80 points; Bidding with caution: 60 points ≤ overall bid score < 80 points; Abandoning a bid: Overall bid score < 60 points.

Images