An intelligent auxiliary bid evaluation method configured by rules

Abstract

The present invention discloses an intelligent assisted bid evaluation method configured by rules, including: determining the evaluation criteria for the bid evaluation task based on rule configuration; compiling bid documents, extracting bid response contents from the bid documents, and converting the bid response contents into structured data through OCR recognition; comparing the bid response contents converted into structured data item by item according to the evaluation criteria, determining whether they are responsive and giving the evaluation results; sending the evaluation results to the bid evaluation expert terminal, and receiving the result confirmation information returned by the bid evaluation expert terminal to generate bid evaluation information. It makes the evaluation process more standardized, the evaluation results more correct, and the evaluation efficiency more efficient.

Description

Technical Field

[0001] The present invention relates to the technical field of bid evaluation, and particularly to an intelligent assisted bid evaluation method configured by rules. Background Art

[0002] In the traditional mode, the bid evaluation work is carried out by bid evaluation experts manually for comparison, analysis and judgment, and quantitative scoring is performed. In this mode, the following technical problems exist: there is a serious conflict between quality and efficiency. With the improvement of the business environment, the number of bidders in a single tender project has gradually increased in recent years, and projects with hundreds or thousands of people are not uncommon. In a limited time, in the face of cumbersome evaluation terms, it is extremely time-consuming for bid evaluation experts to carefully read each bid document, and a large amount of time and energy are consumed in the confirmation of qualification conditions, compliance verification and quotation analysis, resulting in less time for the evaluation of technical bids and low evaluation efficiency. Summary of the Invention

[0003] The present invention aims to solve at least one of the technical problems in the above technologies to some extent. For this purpose, the object of the present invention is to propose an intelligent assisted bid evaluation method configured by rules. Through intelligent technologies such as OCR and NLP technologies, relying on the practical experience accumulated over the years in bid evaluation work, rules that meet the needs of bid evaluation business are sorted out, a rule library is established, and the rule library is dynamically managed, and the evaluation rules are dynamically configured according to business changes. With the help of information technology means, relevant data in the credit part and business part are extracted from the bid documents submitted by bidders, the data is processed to output structured data recognizable by a computer, and the structured data of the bid documents is compared and verified with the evaluation criteria according to the rule requirements, so as to complete the evaluation and scoring of the bid documents. Bid evaluation experts do not need to make manual comparisons and scores, but only need to check and confirm the automatic scoring results of the computer. This method can make full use of the characteristics of fast computer calculation speed, accurate calculation and unified judgment criteria, making the evaluation process more standardized, the evaluation results more correct, and the evaluation efficiency more efficient.

[0004] To achieve the above object, an embodiment of the present invention proposes an intelligent assisted bid evaluation method configured by rules, including:

[0005] Determine the evaluation criteria for the bid evaluation task based on rule configuration;

[0006] Prepare a bid document, extract the bid response content from the bid document, and convert the bid response content into structured data through OCR recognition;

[0007] According to the evaluation criteria, compare the bid response content converted into structured data item by item, determine whether it responds and give an evaluation result;

[0008] Send the evaluation result to the bid evaluation expert terminal, and receive the result confirmation information returned by the bid evaluation expert terminal to generate bid evaluation information.

[0009] According to some embodiments of the present invention, determining the evaluation criteria for bid evaluation tasks based on rule configurations includes:

[0010] Defining evaluation rule configuration information and establishing an evaluation rule library;

[0011] Compiling a tender document and performing identification to extract the bid evaluation method content in the tender document;

[0012] Determining the evaluation criteria for the bid evaluation task according to the evaluation rule library and the bid evaluation method content.

[0013] According to some embodiments of the present invention, compiling a tender document includes: setting requirement standards, evaluation steps, and scoring mechanisms; wherein,

[0014] The requirement standards include qualification requirements, reputation requirements, and qualifications of key personnel;

[0015] The qualification requirements include work safety license, business license, and professional qualification requirements.

[0016] According to some embodiments of the present invention, compiling a bid document includes:

[0017] Compiling a response document directory, exporting a file template under the corresponding chapter, filling in the file template, importing after compilation, signing the imported file, and generating a response document as the bid document.

[0018] According to some embodiments of the present invention, based on the evaluation criteria, comparing the bid response content converted into structured data item by item, determining whether it responds and giving the evaluation result, including:

[0019] Sequentially performing hardware information evaluation, price correction, preliminary evaluation, commercial evaluation, technical evaluation, first price scoring, second price scoring, and comprehensive scoring on the bid response content converted into structured data; wherein the evaluation items for hardware information evaluation include MAC address and IP address.

[0020] According to some embodiments of the present invention, the evaluation result includes:

[0021] The compliance analysis result, including inspection of fees and taxes, inspection of non-competitive fees, and inspection of list consistency, is the inspection of the consistency between the bid lists of each bidder and the tender list;

[0022] The computational analysis result is the inspection of calculation errors in the bid lists of each bidder;

[0023] The rationality analysis result is the unbalanced analysis of the bid lists of each bidder;

[0024] The similarity analysis result is the analysis of the similarity of the bid list quotes of each bidder.

[0025] According to some embodiments of the present invention, before generating bid evaluation information, it further includes:

[0026] Summarize the scores of each bid document in the evaluation results, rank them, and obtain the first ranking;

[0027] Determine the ranking of each bid document according to the result confirmation information returned by the bid evaluation expert terminal, and obtain the second ranking;

[0028] Compare the first ranking with the second ranking. When it is determined that the first ranking is inconsistent with the second ranking, generate a prompt message for modifying the basis of the remarks.

[0029] According to some embodiments of the present invention, extracting bid response content from the bid document includes:

[0030] Analyze the bid document to determine feature information; the feature information includes the extension name of the bid document, the file header feature, and the file tail feature; calculate the corresponding feature value according to the feature information of the bid document, compare the feature value with the preset feature value range corresponding to the preset set, determine the corresponding target preset set according to the comparison result, and cluster the bid documents to obtain several classification sets;

[0031] Perform part-of-speech tagging on any file in the classification set, perform dependency syntactic analysis according to the result of the part-of-speech tagging, obtain a dependency syntactic tree corresponding to the file, and use the root of the dependency syntactic tree as the keyword of the file; determine the corresponding preset domain dictionary according to the keyword; perform in-domain dependency syntactic analysis on the corresponding classification set according to the preset domain dictionary to obtain the first semantic analysis result;

[0032] Determine the preset latent semantic dictionary associated with the preset domain dictionary, perform in-domain dependency syntactic analysis on the corresponding classification set according to the preset latent semantic dictionary to obtain the second semantic analysis result; the preset latent semantic dictionary includes a part-of-speech aggregation dictionary and an event recognition dictionary;

[0033] Extract bid response content from the bid document according to the first semantic analysis result and the second semantic analysis result.

[0034] According to some embodiments of the present invention, before compiling the bid document and extracting bid response content from the bid document, it further includes:

[0035] Collect the risk information of the bidding enterprise and its associated enterprises in the bid document, and determine the enterprise risk value. When it is determined that the enterprise risk value is greater than the preset risk threshold, generate a prompt message for unqualified enterprises.

[0036] According to some embodiments of the present invention, extracting bid response content from bid documents according to the first semantic analysis result and the second semantic analysis result includes:

[0037] According to the first semantic analysis result and the second semantic analysis result, a number of semantic information is obtained. According to the quantity of the semantic information, the corresponding analysis function is obtained in the preset quantity - analysis function relationship;

[0038] Calculating the matching degree between each semantic information and the target semantic information corresponding to the bid response content according to the analysis function;

[0039]

[0040] Wherein, is the semantic feature value of the P-th semantic information; is the number of keywords in the P-th semantic information; is the character feature value of the i-th keyword in the P-th semantic information; is the fusion coefficient of each keyword in the P-th semantic information, and the value range is (0, 2);

[0041]

[0042] Wherein, is the matching degree between the P-th semantic information and the target semantic information corresponding to the bid response content; is the semantic feature value of the target semantic information; is the first coefficient of the analysis function Act, and the value range is (0, 1); is the second coefficient of the analysis function Act, which is proportional to the size of the semantic information, and the value range is (0, 1), and + = 1; is the average value of the semantic feature values of all semantic information; the analysis function means expanding by 10𝛾2 times;

[0043] Determine the semantic information with the highest matching degree, and use the content corresponding to the semantic information with the highest matching degree as the bid response content.

[0044] The present invention provides an intelligent auxiliary bid evaluation method configured by rules. Through intelligent technologies such as OCR and NLP technologies, and relying on the practical experience accumulated over the years in bid evaluation work, rules that meet the requirements of bid evaluation business are sorted out, a rule library is established, and the rule library is dynamically managed, and the evaluation rules are dynamically configured according to business changes. By means of information technology, relevant data in the credit part and business part are extracted from the bid documents submitted by bidders, the data is processed to output structured data recognizable by a computer, and the structured data of the bid documents is compared and verified with the evaluation criteria according to the rule requirements, so as to complete the evaluation and scoring of the bid documents. The bid evaluation experts do not need to make manual comparisons and scores, but only need to check and confirm the automatic scoring results of the computer. This method can make full use of the characteristics of fast computer calculation speed, accurate calculation, and unified judgment criteria, making the evaluation process more standardized, the evaluation results more correct, and the evaluation efficiency more efficient.

[0045] Other features and advantages of the present invention will be described in the following description, and in part will be obvious from the description, or can be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained by the structures specifically pointed out in the written description and the drawings.

[0046] The technical solutions of the present invention will be further described in detail below through the drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The drawings are used to provide a further understanding of the present invention, and constitute a part of the description. They are used to explain the present invention together with the embodiments of the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0048] Figure 1 is a flowchart of an intelligent auxiliary bid evaluation method configured by rules according to an embodiment of the present invention;

[0049] Figure 2 is a schematic diagram of an intelligent auxiliary bid evaluation method configured by rules according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] The following describes the preferred embodiments of the present invention with reference to the drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0051] As Figure 1 - Figure 2 shown, an embodiment of the present invention provides an intelligent auxiliary bid evaluation method configured by rules, including steps S1-S4:

[0052] S1. Determine the evaluation criteria for the bid evaluation task based on rule configuration;

[0053] S2. Prepare the tender documents, extract the tender response content from the tender documents, and convert the tender response content into structured data through OCR recognition.

[0054] S3. According to the evaluation criteria, compare the tender response content converted into structured data item by item, determine whether it responds, and give the evaluation result.

[0055] S4. Send the evaluation result to the bid evaluation expert terminal, receive the result confirmation information returned by the bid evaluation expert terminal, and generate the bid evaluation information.

[0056] The working principle of the above technical solution: Through OCR recognition, the tender response content is converted into structured data, including converting the image files and attachments in the tender documents into structured data, which is convenient for the computer to identify. The evaluation result is sent to the bid evaluation expert terminal, and the bid evaluation expert verifies the automatically judged result of the system. If the result is considered correct, it is confirmed; if the result is considered incorrect, the relevant instructions are given and corrected. Receive the result confirmation information returned by the bid evaluation expert terminal and generate the bid evaluation information.

[0057] The beneficial effects of the above technical solution: Through intelligent technologies such as OCR and NLP technologies, relying on the practical experience accumulated in the bid evaluation work over the years, the rules that meet the bid evaluation business requirements are sorted out, a rule library is established, and the rule library is dynamically managed, and the evaluation rules are dynamically configured according to business changes. With the help of information technology means, relevant data in the credit part and business part are extracted from the tender documents submitted by the bidders, and the data is processed and output into structured data that can be recognized by the computer. The structured data of the tender documents is compared and verified with the evaluation criteria according to the rule requirements, so as to complete the evaluation and scoring of the tender documents. The bid evaluation experts do not need to make manual comparisons and scores, but only need to check and confirm the automatically scored results of the computer. This method can make full use of the characteristics of the computer's fast calculation speed, accurate calculation, and unified judgment criteria, making the evaluation process more standardized, the evaluation results more correct, and the evaluation efficiency more efficient. Unified evaluation criteria: Adopt an intelligent evaluation method, and the computer uses unified standards and rules to automatically identify the evaluation criteria and methods in the tender documents to evaluate each tender document. The standards are the same and the rules are the same, improving the objectivity and fairness of the evaluation results. High-quality evaluation: Rely on intelligent technologies to analyze and compare each tender document, replace experts to complete repetitive and mechanical work, so that the bid evaluation experts can focus on the technical bid evaluation that can better reflect the experts' abilities within a limited time, ensuring more sufficient evaluation time and higher-quality evaluation. Accurate evaluation results: Give full play to the supercomputing ability and accuracy characteristics of the computer, combined with a perfect rule system, follow the evaluation criteria and requirements, and judge and compare the tender documents faster and more accurately.

[0058] According to some embodiments of the present invention, the evaluation criteria for the bid evaluation task are determined based on rule configuration, including:

[0059] Define the review rule configuration information and establish a review rule library;

[0060] Prepare the tender documents, identify them, and extract the content of the evaluation method in the tender documents;

[0061] Determine the review criteria for the bid evaluation task according to the review rule library and the content of the evaluation method.

[0062] The working principle and beneficial effects of the above technical solution: Provide a rule configuration function, which can configure review rules according to business requirements and establish a review rule library. Prepare the tender documents, identify them, extract the content of the evaluation method in the tender documents, and determine the review criteria for the bid evaluation task according to the review rule library and the content of the evaluation method. It is convenient to accurately determine the review criteria for the bid evaluation task.

[0063] According to some embodiments of the present invention, prepare the tender documents, including: setting requirement standards, review steps, and scoring mechanisms; among them,

[0064] The requirement standards include qualification requirements, reputation requirements, and qualifications of key personnel;

[0065] The qualification requirements include a work safety license, a business license, and professional qualification requirements; among them, the professional qualification requirements include the old version / construction / general contracting for construction / municipal public works construction general contracting / level 3 or above, or the new version / construction / special contracting for construction / crane equipment installation engineering special contracting / grade B or above.

[0066] In one embodiment, set the highest score. The total score of all review processes is 100 points, and the system displays the total score of the current step.

[0067] According to some embodiments of the present invention, prepare the bid documents, including:

[0068] Prepare the response file directory, export the file template under the corresponding chapter, fill in the file template, import it after compilation, sign the imported file, and generate the response file as the bid document.

[0069] In one embodiment, the bidder can import other files through "adding chapters".

[0070] According to some embodiments of the present invention, compare the bid response content converted into structured data item by item according to the review criteria, determine whether it responds, and give the review result, including:

[0071] Sequentially conduct hardware information review, price correction, preliminary review, business review, technical review, first price scoring, second price scoring, and comprehensive scoring on the bid response content converted into structured data; among them, the review items for the hardware information review include MAC address and IP address.

[0072] Advantages of the above technical solution: In the hardware information review process, the system provides a one-key analysis function to determine whether there is duplicate hardware information, and the judge makes a confirmation after viewing the analysis result. It also determines whether the MAC address and IP address of the enterprise lessee organization are correct.

[0073] According to some embodiments of the present invention, the review results include:

[0074] The compliance analysis results, including inspection of fees, taxes, non-competitive fees, and list consistency, are for checking the consistency between the tender lists of each tendering unit and the tender invitation list;

[0075] The computational analysis results are for checking calculation errors in the tender lists of each tendering unit;

[0076] The rationality analysis results are for unbalanced analysis of the tender lists of each tendering unit;

[0077] The similarity analysis results are for analyzing the similarity of tender list quotations of each tendering unit.

[0078] Advantages of the above technical solution: Automatically review the strength, qualifications, and performance of the enterprise. The system provides a one-key review function, and the judge makes a confirmation after viewing the review result. The system calculates scores according to the set price review rules, and the final score is confirmed and submitted by the judge team leader. It is convenient to comprehensively analyze the tender lists of each tendering unit, conduct comprehensive consideration, and improve the fairness and accuracy of bid evaluation.

[0079] According to some embodiments of the present invention, before generating bid evaluation information, it further includes:

[0080] Summarize the scores of each tender document in the review results, rank them, and obtain the first ranking;

[0081] Determine the ranking of each tender document according to the result confirmation information returned by the bid evaluation expert terminal to obtain the second ranking;

[0082] Compare the first ranking with the second ranking. When it is determined that the first ranking is inconsistent with the second ranking, generate a prompt message for modifying the basis of the remarks.

[0083] Working principle and advantages of the above technical solution: After all reviews are completed, the system automatically summarizes the scores and ranks them, that is, the first ranking; the judge team leader confirms the final ranking according to the opinions of the judge team, that is, the second ranking. Compare the first ranking with the second ranking. When it is determined that the first ranking is inconsistent with the second ranking, generate a prompt message for modifying the basis of the remarks. This avoids randomly interfering with the bid evaluation results of the system and ensures the fairness and accuracy of the bid evaluation work.

[0084] According to some embodiments of the present invention, extracting tender response content from the tender documents includes:

[0085] Analyze the tender documents to determine the characteristic information; the characteristic information includes the extension name of the tender document, the file header characteristics, and the file tail characteristics; calculate the corresponding characteristic values according to the characteristic information of the tender document, compare the characteristic values with the preset characteristic value ranges corresponding to the preset sets, determine the corresponding target preset set according to the comparison results, and cluster the tender documents to obtain several classification sets;

[0086] Perform part-of-speech tagging on any file in the classification set, perform dependency syntactic analysis according to the results of the part-of-speech tagging, obtain the dependency syntactic tree corresponding to the file, and use the root of the dependency syntactic tree as the keyword of the file; determine the corresponding preset domain dictionary according to the keyword; perform dependency syntactic analysis within the domain on the corresponding classification set according to the preset domain dictionary to obtain the first semantic analysis result;

[0087] Determine the preset latent semantic dictionary associated with the preset domain dictionary, perform dependency syntactic analysis within the domain on the corresponding classification set according to the preset latent semantic dictionary to obtain the second semantic analysis result; the preset latent semantic dictionary includes a part-of-speech aggregation dictionary and an event recognition dictionary;

[0088] Extract the tender response content from the tender documents according to the first semantic analysis result and the second semantic analysis result.

[0089] Working principle of the above technical solution: In this embodiment, read and analyze the content of the tender documents to determine their basic characteristic information, such as the extension name of the file (such as.docx,.pdf), the file header characteristics (such as metadata or header descriptions in a specific format), and the file tail characteristics (such as signatures, dates, etc.). Calculate the characteristic values: Based on the extracted characteristic information, calculate the corresponding characteristic values. Set corresponding weights for the extension names, file header characteristics, and file tail characteristics of different tender documents, and perform weighted calculations based on the weights and the numerically processed data to calculate the corresponding characteristic values for subsequent classification or recognition processes. Compare the calculated characteristic values with the characteristic value ranges in the preset sets to determine the target preset set to which each file belongs. Subsequently, cluster the tender documents according to these sets to form several classification sets. Each classification set represents different types of tender documents or different response contents. Realize the classification and sorting of tender documents, which is convenient for improving the recognition efficiency and accuracy of data.

[0090] In this embodiment, text processing is performed on each file in the classification set. First, part-of-speech tagging is carried out. Each word or phrase in the text is tagged with its corresponding part of speech (such as noun, verb, adjective, etc.), which is the basis for understanding the semantic meaning of the text. Based on the results of part-of-speech tagging, dependency syntactic analysis is performed to generate a dependency syntactic tree for each file. The dependency syntactic tree shows the dependency relationships between the various components in the sentence, where the root of the tree usually represents the core information or main action of the sentence. The root of the dependency syntactic tree is used as the keyword of the file, and the keyword can highly summarize the core content of the file. Based on the keyword, a preset keyword-preset domain dictionary data table is queried to determine the corresponding preset domain dictionary. The domain dictionary is a collection of commonly used words and expressions in a specific domain, which helps to more accurately understand the text within the domain. The classification set is subjected to dependency syntactic analysis within the domain using the preset domain dictionary to obtain the first semantic analysis result. The semantic information of the file in the specific domain is further mined, and semantic analysis is performed based on the corresponding domain dictionary, improving the accuracy of the semantic analysis result.

[0091] In this embodiment, a preset implicit semantic dictionary associated with the preset domain dictionary is determined. This dictionary includes a part-of-speech aggregation dictionary and an event recognition dictionary. The part-of-speech aggregation dictionary is used to classify similar parts of speech for a more abstract understanding of the text; the event recognition dictionary is used to identify the events described in the text and their related elements. The preset domain dictionary represents a direct semantic analysis dictionary, and the preset implicit semantic dictionary represents an indirect semantic analysis dictionary. The classification set is subjected to further dependency syntactic analysis using the implicit semantic dictionary to obtain the second semantic analysis result. The aim is to reveal the implicit semantic information and event structure in the file, ensuring the comprehensiveness and accuracy of semantic analysis. Combining the first semantic analysis result and the second semantic analysis result can comprehensively and deeply understand the semantic content in the tender documents, and extract the tender response content from the tender documents for subsequent review and decision-making processes.

[0092] Beneficial effects of the above technical solution: Classifying and organizing the tender documents facilitates improving the file recognition efficiency. The domain of each classification set is determined, and based on the dictionary of the corresponding domain, semantic analysis is carried out directly and indirectly to obtain the first semantic analysis result and the second semantic analysis result, thereby facilitating the accurate obtaining of the tender response content.

[0093] According to some embodiments of the present invention, before compiling the tender documents and extracting the tender response content from the tender documents, it further includes:

[0094] Collect the risk information of the tender enterprise and related enterprises in the tender documents, and determine the enterprise risk value. When it is determined that the enterprise risk value is greater than the preset risk threshold, an unqualified enterprise prompt message is generated.

[0095] Advantages of the above technical solution: Collect risk information of the bidding enterprise and related enterprises in the bidding documents, and determine the enterprise risk value. When it is determined that the enterprise risk value is greater than the preset risk threshold, generate a prompt message for unqualified enterprises. This facilitates the early assessment of the overall risk of enterprises, screens out enterprises that do not meet the risk requirements, reduces the number of enterprises to be reviewed, and improves the review speed.

[0096] According to some embodiments of the present invention, extracting the bid response content from the bidding documents according to the first semantic analysis result and the second semantic analysis result includes:

[0097] According to the first semantic analysis result and the second semantic analysis result, obtain a number of semantic information, and obtain the corresponding analysis function in the preset quantity - analysis function relationship according to the quantity of the semantic information;

[0098] Calculate the matching degree between each semantic information and the target semantic information corresponding to the bid response content according to the analysis function;

[0099]

[0100] Among them, is the semantic feature value of the P-th semantic information; is the number of keywords in the P-th semantic information; is the character feature value of the i-th keyword in the P-th semantic information; is the fusion coefficient of each keyword in the P-th semantic information, and the value range is (0, 2);

[0101]

[0102] Among them, is the matching degree between the P-th semantic information and the target semantic information corresponding to the bid response content; is the semantic feature value of the target semantic information; is the first coefficient of the analysis function Act, and the value range is (0, 1); is the second coefficient of the analysis function Act, which is proportional to the size of the semantic information, and the value range is (0, 1), and + = 1; is the average value of the semantic feature values of all semantic information; the analysis function means expanding by 10𝛾2 times;

[0103] Determine the semantic information with the highest matching degree, and use the content corresponding to the semantic information with the highest matching degree as the bid response content.

[0104] Working principle and beneficial effects of the above technical solution: According to the first semantic analysis result and the second semantic analysis result, a number of semantic information are obtained. According to the quantity of the semantic information, the corresponding analysis function is obtained in the preset quantity - analysis function relationship; the matching degree between each semantic information and the target semantic information corresponding to the tender response content is calculated according to the analysis function; during the calculation process, the larger the semantic information, the larger it is. There is a preset control data table for the variable coefficients of the preset semantic information and the analysis function Act. The analysis function means expanding X by 10 times. For example, when X = 3, is equal to 0.2, then = 6; by way of example, = 1.2, m = 2, and the character feature values of the i-th keyword in the P-th semantic information are 2 and 1.5 respectively, then the semantic feature value of the P-th semantic information is 4.2. When the quantity of the semantic information is 6, the semantic feature value of the target semantic information is 1.5, the first coefficient of the analysis function Act is 0.6, the second coefficient is 0.4, the average value of the semantic feature values of all semantic information is 3, and the matching degree between the P-th semantic information and the target semantic information corresponding to the tender response content is 13.3%. Determine the semantic information with the highest matching degree, and use the content corresponding to the semantic information with the highest matching degree as the tender response content. This is convenient for accurately determining the tender response content.

[0105] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and its equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. An intelligent assisted bidding evaluation method configured by rules, characterized in that: include: Determine the evaluation criteria for the bid evaluation task based on rule configuration; Prepare bidding documents, extract bidding response content from bidding documents, and convert bidding response content into structured data through OCR recognition; According to the evaluation criteria, the bid response contents converted into structured data are compared item by item to determine whether they are responsive and give the evaluation results; Send the evaluation results to the bid evaluation expert end, receive the result confirmation information returned by the bid evaluation expert end, and generate bid evaluation information; Extract bid response content from bid documents, including: Analyze the bidding document to determine feature information; the feature information includes the extension name, file header feature and file tail feature of the bidding document; calculate a corresponding feature value according to the feature information of the bidding document, compare the feature value with a preset feature value range corresponding to a preset set, determine a corresponding target preset set according to the comparison result, cluster the bidding document to obtain a plurality of classification sets; Perform part-of-speech tagging on any file in the classification set, perform dependency syntactic analysis based on the result of the part-of-speech tagging, obtain a dependency syntactic tree corresponding to the file, and use the root of the dependency syntactic tree as the keyword of the file; determine the corresponding preset domain dictionary based on the keyword; perform dependency syntactic analysis within the domain of the corresponding classification set based on the preset domain dictionary to obtain a first semantic analysis result; Determine a preset latent semantic dictionary associated with the preset domain dictionary, and perform dependency syntactic analysis on the corresponding classification set in the domain according to the preset latent semantic dictionary to obtain a second semantic analysis result; the preset latent semantic dictionary includes a part-of-speech aggregation dictionary and an event recognition dictionary; extracting the bid response content from the bid document according to the first semantic analysis result and the second semantic analysis result; Extracting bid response content from the bid document according to the first semantic analysis result and the second semantic analysis result includes: According to the first semantic analysis result and the second semantic analysis result, a plurality of semantic information is obtained, and according to the amount of the semantic information, a corresponding analysis function is obtained in a preset quantity-analysis function relationship; Calculate the matching degree of each semantic information with the target semantic information corresponding to the bid response content according to the analysis function; in, is the semantic feature value of the Pth semantic information; is the number of keywords in the Pth semantic information; is the word feature value of the i-th keyword in the P-th semantic information; is the fusion coefficient of each keyword in the Pth semantic information, and its value range is (0,2); in, is the matching degree between the Pth semantic information and the target semantic information corresponding to the bid response content; is the semantic feature value of the target semantic information; is the first coefficient of the analysis function Act, and its value range is (0,1); is the second coefficient of the analysis function Act, which is proportional to the size of the semantic information and has a value range of (0,1), and + =1; is the average value of the semantic feature values ​​of all semantic information; analysis function Indicates that Expand by 10𝛾2 times; The semantic information with the highest matching degree is determined, and the content corresponding to the semantic information with the highest matching degree is used as the bid response content.

2. The intelligent assisted bidding evaluation method according to claim 1, characterized in that: The evaluation criteria for the bid evaluation task are determined based on the rule configuration, including: Define review rule configuration information and build a review rule library; Prepare bidding documents, identify them, and extract the bid evaluation method from them; Determine the evaluation criteria for the bid evaluation task based on the evaluation rule base and the bid evaluation method.

3. The intelligent assisted bidding evaluation method according to claim 2, characterized in that: Prepare bidding documents, including: setting requirements, evaluation procedures and scoring mechanism; among which, The required standards include qualification requirements, credibility requirements and qualifications of key personnel; Qualification requirements include production safety license, business license and professional qualification requirements.

4. The intelligent assisted bidding evaluation method according to claim 1, characterized in that: Preparation of bidding documents, including: Prepare the response file directory, export the file template under the corresponding chapter, fill in the file template, import it after preparation, sign and seal the imported file, and generate the response file as the bidding document.

5. The intelligent assisted bidding evaluation method according to claim 1, characterized in that: According to the evaluation criteria, the bid response content converted into structured data is compared item by item to determine whether it is responsive and give the evaluation results, including: The bid response content converted into structured data will be subject to hardware information review, price revision, preliminary review, business review, technical review, first price scoring, second price scoring and comprehensive scoring in sequence; the review items of the hardware information review include MAC address and IP address.

6. The intelligent assisted bidding evaluation method according to claim 1, characterized in that: The review results include: The results of the compliance analysis include inspection of fees, taxes, non-competitive fees, and consistency of lists, which is a consistency check between the bid list and the tender list of each bidding unit; The computational analysis results are used to check the calculation errors of the bid list of each bidding unit; The rationality analysis result is the imbalance analysis of the bidding list of each bidding unit; The result of the similarity analysis is the similarity analysis of the bid lists and quotations of each bidding unit.

7. The intelligent assisted bidding evaluation method according to claim 1, characterized in that: Before generating the bid evaluation information, it also includes: Summarize the scores of each bidding document in the evaluation results, rank them, and get the first place; Determine the ranking of each bidding document according to the result confirmation information returned by the bid evaluation expert, and obtain the second ranking; The first ranking is compared with the second ranking, and when it is determined that the first ranking is inconsistent with the second ranking, a prompt message of the modification basis is generated.

8. The intelligent assisted bidding evaluation method according to claim 1, characterized in that: Before preparing the bidding documents and extracting the bidding response content from the bidding documents, it also includes: Collect risk information of bidding enterprises and affiliated enterprises in bidding documents, and determine enterprise risk value. When it is determined that the enterprise risk value is greater than the preset risk threshold, generate unqualified enterprise prompt information.

Images