Intelligent bidding document identifying and screening platform and method

Abstract

The invention is suitable for the technical field of file processing, and provides a bidding document intelligent identification and screening platform and method, and the method comprises the following steps: identifying the text information of each bidding document, and extracting credit content in the text information according to basic credit requirements; performing primary screening on the bidding files according to the basic credit qualification requirements and credit qualification contents, and determining the number of reserved bidding files; determining advanced bidding requirements according to the number of the bidding files, and extracting advanced contents in the text information; performing secondary screening on the bidding files according to advanced bidding requirements and advanced contents; and classifying the bidding files reserved by secondary screening according to the overall project scheme. According to the method, on the basis of screening the bidding files based on the basic credit requirements, comprehensive and accurate evaluation based on advanced bidding requirements can be realized, the advanced bidding requirements are dynamically determined according to the number of the bidding files, and advanced contents in text information are extracted and evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of document processing, and in particular to a bidding document intelligent identification and screening platform and method. Background Art

[0002] In the current bidding process, the processing and screening of bid documents is a tedious and time-consuming task. Traditional screening methods mainly rely on manual reading and analysis, which is not only inefficient, but also prone to inconsistency and subjectivity in screening results due to human factors. With the increase in bidding projects, the number and complexity of bid documents are also increasing. Traditional screening methods can no longer meet the needs of efficient and accurate processing. Therefore, it is necessary to provide a bidding document intelligent identification and screening platform and method to solve the above problems. Summary of the invention

[0003] In view of the deficiencies in the prior art, the purpose of the present invention is to provide a bidding document intelligent identification and screening platform and method to solve the problems existing in the above-mentioned background technology.

[0004] The present invention is implemented as follows: a method for intelligently identifying and screening bidding documents, the method comprising the following steps:

[0005] Identify the text information of each bidding document and extract the credit content in the text information according to the basic credit requirements;

[0006] Conduct a preliminary screening of the bidding documents based on the basic credit requirements and credit content, and determine the number of bidding documents to be retained;

[0007] Determine advanced bidding requirements based on the number of bidding documents and extract advanced content from text information;

[0008] Conduct secondary screening of bidding documents according to advanced bidding requirements and advanced contents;

[0009] Classify the bid documents retained after the second screening according to the overall project plan to obtain several bid categories;

[0010] Compare several overall project proposals in each bidding category.

[0011] As a further solution of the present invention: the step of determining the advanced bidding requirements according to the number of bidding documents and extracting the advanced content in the text information specifically includes:

[0012] Input the number of bidding documents into the advanced requirements library and output the corresponding advanced bidding requirements. Each document quantity range corresponds to different advanced bidding requirements.

[0013] Determine the parameter items and project requirements in the advanced bidding requirements, extract the advanced content in the text information based on the parameter items, and determine the project content corresponding to each of the parameter items.

[0014] As a further solution of the present invention: the step of secondary screening of the bidding documents according to the advanced bidding requirements and advanced contents specifically includes:

[0015] Calculate the degree of compliance of parameter items based on the project content and corresponding project requirements;

[0016] Calculate the matching degree R of each bidding document relative to the advanced bidding requirements based on the degree of compliance. Where m represents the number of parameter items in the advanced bidding requirements, Vz represents the weight of the zth parameter item, and Fz is the degree of compliance of the bidding document on the zth parameter item;

[0017] The bidding documents are arranged in descending order according to the degree of compliance, and the bidding documents ranked after the nth place are screened out, where n is a fixed value.

[0018] As a further solution of the present invention: the step of classifying the bid documents retained after the secondary screening according to the overall project plan to obtain several bid categories specifically includes:

[0019] Extract the overall project plan from the text information, preprocess the overall project plan and perform text vectorization;

[0020] A similarity matrix is ​​constructed based on cosine similarity. For n bidding documents, an n×n similarity matrix S is constructed, where Sij represents the similarity between the i-th and j-th bidding documents;

[0021] Determine the number of clusters K and perform K-means clustering, specifically: randomly select K data points as the initial cluster centers, and the initial cluster centers are the starting points of iterations; calculate the distance from each data point to the K cluster centers, and assign each data point to the cluster with the closest cluster center, and recalculate the cluster center of each cluster; until the termination condition is met;

[0022] Output and verify the clustering results to obtain several bidding categories.

[0023] As a further solution of the present invention: the steps of preprocessing the overall project plan and text vectorization specifically include:

[0024] Use word segmentation tools to split the overall project plan into individual words and phrases, and remove stop words;

[0025] The processed text is converted into vector form. The method for text vectorization is the bag-of-words model or term frequency-inverse document frequency.

[0026] As a further solution of the present invention: the step of comparing the overall plans of several projects in each bidding category specifically includes:

[0027] Retrieve the comparison dimension and create a comparison table based on the comparison dimension;

[0028] According to the comparison dimensions, extract the corresponding dimension content in the overall plan of each project and complete the comparison table.

[0029] Another object of the present invention is to provide a bidding document intelligent identification and screening platform, the platform comprising:

[0030] The credit content extraction module is used to identify the text information of each bidding document and extract the credit content in the text information according to the basic credit requirements;

[0031] The document initial screening module is used to conduct initial screening of bidding documents based on basic credit requirements and credit content, and determine the number of bidding documents to be retained;

[0032] An advanced content extraction module is used to determine advanced bidding requirements based on the number of bidding documents and extract advanced content from text information;

[0033] The document secondary screening module is used to conduct secondary screening of bidding documents according to advanced bidding requirements and advanced content;

[0034] The bidding category determination module is used to classify the bidding documents retained after the secondary screening according to the overall project plan to obtain several bidding categories;

[0035] The project plan comparison module is used to compare several overall project plans in each bidding category.

[0036] As a further solution of the present invention: the advanced content extraction module includes:

[0037] The advanced bidding requirement unit is used to input the number of bidding documents into the advanced requirement library and output the corresponding advanced bidding requirements. Each document quantity range corresponds to a different advanced bidding requirement.

[0038] The project content determination unit is used to determine the parameter items and project requirements in the advanced bidding requirements, extract the advanced content in the text information based on the parameter items, and determine the project content corresponding to each of the parameter items.

[0039] As a further solution of the present invention: the file secondary screening module includes:

[0040] The conformity determination unit is used to calculate the conformity of the parameter project based on the project content and the corresponding project requirements; the matching degree calculation unit is used to calculate the matching degree R of each bidding document relative to the advanced bidding requirements according to the conformity degree, Where m represents the number of parameter items in the advanced bidding requirements, Vz represents the weight of the zth parameter item, and Fz is the degree of compliance of the bidding document on the zth parameter item;

[0041] The bidding document screening unit is used to arrange the bidding documents in descending order according to the degree of compliance, and the bidding documents arranged after the nth position are screened out, where n is a fixed value.

[0042] As a further solution of the present invention: the bidding category determination module includes:

[0043] A scheme text processing unit is used to extract the overall project scheme from the text information, and to pre-process and vectorize the text of the overall project scheme;

[0044] A similarity matrix construction unit is used to construct a similarity matrix based on cosine similarity. For n bidding documents, an n×n similarity matrix S is constructed, where Sij represents the similarity between the i-th and j-th bidding documents.

[0045] The clustering execution unit is used to determine the number of clusters K and perform K-means clustering, specifically: randomly select K data points as the initial cluster centers, and the initial cluster centers are the starting points of iterations; calculate the distance from each data point to the K cluster centers, and assign each data point to the cluster with the closest cluster center, and recalculate the cluster center of each cluster; until the termination condition is met;

[0046] The clustering result output unit is used to output and verify the clustering result to obtain several bidding categories.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention can not only screen the bidding documents based on the basic credit requirements, but also realize a comprehensive and accurate evaluation based on the advanced bidding requirements, dynamically determine the advanced bidding requirements according to the number of bidding documents, and extract and evaluate the advanced content in the text information. In addition, the bidding documents retained by the secondary screening can be classified according to the overall project plan, and several overall project plans in each bidding category can be compared, which is convenient for the auditor to review and compare, and improves the auditor's subsequent audit work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 The present invention is a flowchart of a bidding document intelligent identification and screening method.

[0050] Figure 2 The present invention is a flowchart for extracting advanced content in a bidding document intelligent recognition and screening method.

[0051] Figure 3 The present invention is a flow chart of secondary screening of bidding documents in a bidding document intelligent identification and screening method.

[0052] Figure 4 The present invention is a flowchart for classifying bidding documents in a bidding document intelligent identification and screening method.

[0053] Figure 5 This is a flowchart for preprocessing and text vectorization of the project overall plan in an intelligent recognition and screening method for bidding documents.

[0054] Figure 6 The present invention is a flowchart for comparing overall project plans in a bidding document intelligent identification and screening method.

[0055] Figure 7 The figure is a schematic diagram of the structure of an intelligent identification and screening platform for bidding documents. DETAILED DESCRIPTION

[0056] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0057] The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.

[0058] like Figure 1 As shown, an embodiment of the present invention provides a method for intelligently identifying and screening bidding documents, the method comprising the following steps:

[0059] S100, identifying the text information of each bidding document, and extracting the credit content in the text information according to the basic credit requirements;

[0060] S200, conduct a preliminary screening of the bidding documents based on the basic credit requirements and credit content, and determine the number of bidding documents to be retained;

[0061] S300, determining advanced bidding requirements according to the number of bidding documents, and extracting advanced content from the text information;

[0062] S400, secondary screening of bidding documents based on advanced bidding requirements and advanced content;

[0063] S500, classifying the bid documents retained after the secondary screening according to the overall project plan to obtain a number of bid categories;

[0064] S600, compare several overall project plans in each bidding category.

[0065] It should be noted that in the process of processing the bidding documents, it is first necessary to identify and extract the key credit information in the documents, such as company qualifications, personnel qualification certificates, credibility, etc., which are important bases for judging whether the bidder meets the basic bidding requirements. However, due to the large differences in the format and content of the bidding documents, manual extraction of this information is not only time-consuming and laborious, but also prone to errors. In addition, after the initial screening, the bidding documents need to be further screened according to the advanced bidding requirements. Advanced bidding requirements usually include delivery deadlines, payment methods, after-sales guarantees, financial status, staffing, etc. These requirements are often more complex and specific, and require more refined processing and analysis. Traditional screening methods are often difficult to achieve comprehensive and accurate evaluation when facing these advanced requirements. Finally, after screening out the bidding documents that meet the requirements, it is also necessary to classify and compare the construction design schemes in order to further screen out the best bidding scheme. However, due to the diversity and complexity of the construction design schemes, manual classification and comparison are not only inefficient, but also difficult to ensure the accuracy of classification and the comprehensiveness of comparison. The embodiments of the present invention are intended to solve the above problems.

[0066] In an embodiment of the present invention, the text information of each bidding document is firstly identified based on OCR (optical character recognition) or NLP (natural language processing), and the credit content in the text information is extracted according to the basic credit requirements, which are reflected in the company qualifications, company size, personnel qualification certificates and corporate credit. Then, the bidding documents are initially screened according to the basic credit requirements and the credit content. After the screening, the number of retained bidding documents is determined; then, the advanced bidding requirements are determined according to the number of bidding documents, and the advanced content in the text information is extracted. The advanced bidding requirements are mainly reflected in the project cycle, payment method, after-sales guarantee, financial status, staffing, etc. Then, the bidding documents are secondary screened according to the advanced bidding requirements and the advanced content. It is easy to understand that after the initial screening, the more bidding documents are retained, the more and stricter the corresponding advanced bidding requirements are, so that the secondary screening can be well controlled. Then, the embodiment of the present invention will classify the bidding documents retained by the secondary screening according to the overall project plan to obtain several bidding categories. The overall project plans in several bidding documents in each bidding category have certain similarities. In this way, the classification of bidding documents is automatically realized. In addition, the embodiment of the present invention will also compare the overall project plans of several bidding categories in each bidding category. This greatly facilitates the subsequent review work of the reviewer.

[0067] like Figure 2As shown, as a preferred embodiment of the present invention, the step of determining the advanced bidding requirements according to the number of bidding documents and extracting the advanced content in the text information specifically includes:

[0068] S301, inputting the number of bidding documents into the advanced requirement library, and outputting corresponding advanced bidding requirements, where each document quantity range corresponds to different advanced bidding requirements;

[0069] S302, determining parameter items and project requirements in the advanced bidding requirements, extracting advanced content in the text information based on the parameter items, and determining the project content corresponding to each of the parameter items.

[0070] In an embodiment of the present invention, an advanced requirement library is established in advance, and the advanced requirement library contains several levels of advanced bidding requirements, and each advanced bidding requirement corresponds to a different range of document quantity. When the number of bidding documents is input into the advanced requirement library, the corresponding advanced bidding requirements will be automatically output. The more bidding documents there are, the higher the level of the advanced bidding requirements; then the parameter items and project requirements in the advanced bidding requirements will be determined. The advanced bidding requirements are composed of several parameter items and project requirements. For example, the content of a certain advanced bidding requirement is: 1. Project cycle: 4-6 months, 2. Payment method: installment payment. Then, the advanced content in the text information will be extracted based on the parameter items, and the project content corresponding to each of the parameter items will be determined, so as to facilitate the subsequent determination of the degree of compliance.

[0071] like Figure 3 As shown, as a preferred embodiment of the present invention, the step of secondary screening of the bidding documents according to the advanced bidding requirements and advanced contents specifically includes:

[0072] S401, calculating the conformity of parameter items based on the project content and the corresponding project requirements;

[0073] S402, calculating the matching degree R of each bidding document relative to the advanced bidding requirements according to the degree of compliance,

[0074] S403, arranging the bidding documents in descending order according to the degree of compliance, and filtering out the bidding documents arranged after the nth place, where n is a fixed value.

[0075] In the embodiment of the present invention, the compliance degree of each parameter item is calculated based on the project content and the corresponding project requirements. For example, for the project cycle, when the project cycle in the bidding document is within the range of 4-6 months, the compliance degree of the bidding document on the parameter item is 1, and when it is not within the range of 4-6 months, the compliance degree is 0. Then, the matching degree R of each bidding document relative to the advanced bidding requirements is calculated based on the compliance degree. Among them, m represents the number of parameter items in the advanced bidding requirements, Vz represents the weight of the zth parameter item, and the weight of each parameter item is set in advance. Fz is the degree of compliance of the bid document on the zth parameter item. Finally, the bid documents are sorted in descending order according to the degree of compliance, and the bid documents ranked after the nth are screened out. n is a fixed value, for example, if n is 10, only 10 bid documents will be retained after the second screening.

[0076] like Figure 4 As shown, as a preferred embodiment of the present invention, the step of classifying the bid documents retained after the secondary screening according to the overall project plan to obtain several bid categories specifically includes:

[0077] S501, extracting the overall project plan from the text information, preprocessing the overall project plan and performing text vectorization;

[0078] S502, constructing a similarity matrix based on cosine similarity, and constructing an n×n similarity matrix S for n bidding documents;

[0079] S503, determining the number of clusters K, and performing K-means clustering, specifically: randomly selecting K data points as initial cluster centers, with the initial cluster centers being the starting point of iteration; calculating the distance from each data point to the K cluster centers, and assigning each data point to the cluster with the closest cluster center, and recalculating the cluster center of each cluster; until the termination condition is met;

[0080] S504, output and verify the clustering results to obtain several bidding categories.

[0081] In the embodiments of the present invention, for the convenience of subsequent review work by reviewers, the bid documents after secondary screening will be automatically classified so that the overall project plans in each category are similar. Specifically, the overall project plan in the text information will be extracted, preprocessed and vectorized. Then, a similarity matrix will be constructed based on the cosine similarity. For n bid documents, an n×n similarity matrix S will be constructed, where Sij represents the similarity between the i-th and j-th bid documents. Then, clustering analysis will be carried out. First, the number of clusters K needs to be determined. The optimal number of clusters K can be determined by the Elbow Method or the Silhouette Coefficient method. Then, K-means clustering will be performed. Specifically: In the first step, randomly select K data points as the initial cluster centers, which are the starting points of the iteration. In the second step, calculate the distance from each data point to the K cluster centers (here the Euclidean distance is used), and assign each data point to the cluster where the nearest cluster center is located, and recalculate the cluster center of each cluster, that is, calculate the mean of all data points in the cluster as the new cluster center. In the third step, repeat the second step until the termination condition is met (such as the cluster centers no longer change or the maximum number of iterations is reached). Then, the clustering result will be output and the clustering result will be verified. Specifically, calculate the similarity within the category: For each bid category obtained by clustering, calculate the average similarity (such as cosine similarity) between all bid documents within it to verify the rationality of the clustering result.

[0082] As Figure 5 shown, as a preferred embodiment of the present invention, the steps of preprocessing and text vectorization of the overall project plan specifically include:

[0083] S5011, use a word segmentation tool to split the overall project plan into individual words and phrases, and remove stop words;

[0084] S5012, convert the processed text into a vector form, and the methods for text vectorization are the bag-of-words model or term frequency-inverse document frequency.

[0085] In the embodiments of the present invention, for the convenience of carrying out clustering analysis, it is necessary to preprocess and vectorize the text of the overall project plan in advance. Use a word segmentation tool to split the overall project plan into individual words and phrases, and remove stop words, that is, remove common but meaningless words in the text, such as "de", "shi", etc.; for English text, word stemming is also required to restore the word to its basic form, such as restoring "running" to "run". Then, convert the preprocessed text into a vector form, and common methods include the bag-of-words model (Bag of Words) and TF-IDF (term frequency-inverse document frequency).

[0086] like Figure 6 As shown, as a preferred embodiment of the present invention, the step of comparing the overall plans of several projects in each bidding category specifically includes:

[0087] S601, retrieve the comparison dimension, and create a comparison table according to the comparison dimension;

[0088] S602, extracting the corresponding dimension content in each project overall plan according to the comparison dimension, and completing the comparison table.

[0089] In an embodiment of the present invention, the overall project plans in the same category will also be compared. First, the comparison dimensions will be retrieved. The comparison dimensions are set in advance. For example, the comparison dimensions include: technical route, cost control, quality control and risk management. Then, a comparison table is created based on the comparison dimensions, and the corresponding dimension content in each project overall plan is extracted based on the comparison dimensions. The comparison table is filled in. In this way, the reviewer can intuitively see the differences between the various bidding documents through a table, which is efficient and convenient.

[0090] like Figure 7 As shown, the embodiment of the present invention further provides a bidding document intelligent identification and screening platform, the platform comprising:

[0091] The credit content extraction module 100 is used to identify the text information of each bidding document and extract the credit content in the text information according to the basic credit requirements;

[0092] The document initial screening module 200 is used to perform initial screening of the bidding documents according to the basic credit requirements and credit content, and determine the number of bidding documents to be retained;

[0093] An advanced content extraction module 300 is used to determine advanced bidding requirements according to the number of bidding documents and extract advanced content from the text information;

[0094] A document secondary screening module 400 is used to perform secondary screening of the bidding documents according to the advanced bidding requirements and advanced contents;

[0095] The bidding category determination module 500 is used to classify the bidding documents retained after the secondary screening according to the overall project plan to obtain several bidding categories;

[0096] The project plan comparison module 600 is used to compare several project overall plans in each bidding category.

[0097] As a preferred embodiment of the present invention, the advanced content extraction module 300 includes:

[0098] The advanced bidding requirement unit is used to input the number of bidding documents into the advanced requirement library and output the corresponding advanced bidding requirements. Each document quantity range corresponds to a different advanced bidding requirement.

[0099] The project content determination unit is used to determine the parameter items and project requirements in the advanced bidding requirements, extract the advanced content in the text information based on the parameter items, and determine the project content corresponding to each of the parameter items.

[0100] As a preferred embodiment of the present invention, the file secondary screening module 400 includes:

[0101] A conformity determination unit, for calculating the conformity of a parameter item based on the item content and the corresponding item requirements;

[0102] The matching degree calculation unit is used to calculate the matching degree R of each bidding document relative to the advanced bidding requirements according to the degree of compliance. Where m represents the number of parameter items in the advanced bidding requirements, Vz represents the weight of the zth parameter item, and Fz is the degree of compliance of the bidding document on the zth parameter item;

[0103] The bidding document screening unit is used to arrange the bidding documents in descending order according to the degree of compliance, and the bidding documents arranged after the nth position are screened out, where n is a fixed value.

[0104] As a preferred embodiment of the present invention, the bidding category determination module 500 includes:

[0105] A scheme text processing unit is used to extract the overall project scheme from the text information, and to pre-process and vectorize the text of the overall project scheme;

[0106] A similarity matrix construction unit is used to construct a similarity matrix based on cosine similarity. For n bidding documents, an n×n similarity matrix S is constructed, where Sij represents the similarity between the i-th and j-th bidding documents.

[0107] The clustering execution unit is used to determine the number of clusters K and perform K-means clustering, specifically: randomly select K data points as the initial cluster centers, and the initial cluster centers are the starting points of iterations; calculate the distance from each data point to the K cluster centers, and assign each data point to the cluster with the closest cluster center, and recalculate the cluster center of each cluster; until the termination condition is met;

[0108] The clustering result output unit is used to output and verify the clustering results to obtain several bidding categories.

[0109] The above only describes in detail the preferred embodiments of the present invention, which is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0110] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0111] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0112] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the disclosure in the specification and examples. This application is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.

Claims

1. A method for intelligent identification and screening of bidding documents, characterized in that: The method comprises the following steps: Identify the text information of each bidding document and extract the credit content in the text information according to the basic credit requirements; Conduct a preliminary screening of the bidding documents based on the basic credit requirements and credit content, and determine the number of bidding documents to be retained; Determine advanced bidding requirements based on the number of bidding documents and extract advanced content from text information; Conduct secondary screening of bidding documents according to advanced bidding requirements and advanced contents; Classify the bid documents retained after the second screening according to the overall project plan to obtain several bid categories; Compare several overall project proposals in each bidding category.

2. The method for intelligent identification and screening of bidding documents according to claim 1, characterized in that: The step of determining the advanced bidding requirements according to the number of bidding documents and extracting the advanced content in the text information specifically includes: Input the number of bidding documents into the advanced requirements library and output the corresponding advanced bidding requirements. Each document quantity range corresponds to different advanced bidding requirements. Determine the parameter items and project requirements in the advanced bidding requirements, extract the advanced content in the text information based on the parameter items, and determine the project content corresponding to each of the parameter items.

3. The method for intelligent identification and screening of bidding documents according to claim 2 is characterized in that: The step of secondary screening of the bidding documents according to the advanced bidding requirements and advanced contents specifically includes: Calculate the degree of compliance of parameter items based on the project content and corresponding project requirements; Calculate the matching degree R of each bidding document relative to the advanced bidding requirements based on the degree of compliance. Where m represents the number of parameter items in the advanced bidding requirements, Vz represents the weight of the zth parameter item, and Fz is the degree of compliance of the bidding document on the zth parameter item; The bidding documents are arranged in descending order according to the degree of compliance, and the bidding documents ranked after the nth place are screened out, where n is a fixed value.

4. The method for intelligent identification and screening of bidding documents according to claim 1, characterized in that: The step of classifying the bid documents retained after the secondary screening according to the overall project plan to obtain several bid categories specifically includes: Extract the overall project plan from the text information, preprocess the overall project plan and perform text vectorization; A similarity matrix is ​​constructed based on cosine similarity. For n bidding documents, an n×n similarity matrix S is constructed, where Sij represents the similarity between the i-th and j-th bidding documents; Determine the number of clusters K and perform K-means clustering, specifically: randomly select K data points as the initial cluster centers, and the initial cluster centers are the starting points of iterations; calculate the distance from each data point to the K cluster centers, and assign each data point to the cluster with the closest cluster center, and recalculate the cluster center of each cluster; until the termination condition is met; Output and verify the clustering results to obtain several bidding categories.

5. The method for intelligent identification and screening of bidding documents according to claim 4 is characterized in that: The steps of preprocessing and text vectorizing the overall project plan specifically include: Use word segmentation tools to split the overall project plan into individual words and phrases, and remove stop words; The processed text is converted into vector form. The method for text vectorization is the bag-of-words model or term frequency-inverse document frequency.

6. The method for intelligent identification and screening of bidding documents according to claim 1, characterized in that: The step of comparing the overall plans of several projects in each bidding category specifically includes: Retrieve the comparison dimension and create a comparison table based on the comparison dimension; According to the comparison dimensions, extract the corresponding dimension content in the overall plan of each project and complete the comparison table.

7. A bidding document intelligent identification and screening platform, characterized in that: The platform includes: The credit content extraction module is used to identify the text information of each bidding document and extract the credit content in the text information according to the basic credit requirements; The document initial screening module is used to conduct initial screening of bidding documents based on basic credit requirements and credit content, and determine the number of bidding documents to be retained; An advanced content extraction module is used to determine advanced bidding requirements based on the number of bidding documents and extract advanced content from text information; The document secondary screening module is used to conduct secondary screening of bidding documents according to advanced bidding requirements and advanced content; The bidding category determination module is used to classify the bidding documents retained after the secondary screening according to the overall project plan to obtain several bidding categories; The project plan comparison module is used to compare several overall project plans in each bidding category.

8. The bidding document intelligent identification and screening platform according to claim 7 is characterized in that: The advanced content extraction module includes: The advanced bidding requirement unit is used to input the number of bidding documents into the advanced requirement library and output the corresponding advanced bidding requirements. Each document quantity range corresponds to a different advanced bidding requirement. The project content determination unit is used to determine the parameter items and project requirements in the advanced bidding requirements, extract the advanced content in the text information based on the parameter items, and determine the project content corresponding to each of the parameter items.

9. The bidding document intelligent identification and screening platform according to claim 8, characterized in that: The file secondary screening module includes: A conformity determination unit, for calculating the conformity of a parameter item based on the item content and the corresponding item requirements; The matching degree calculation unit is used to calculate the matching degree R of each bidding document relative to the advanced bidding requirements according to the degree of compliance. Where m represents the number of parameter items in the advanced bidding requirements, Vz represents the weight of the zth parameter item, and Fz is the degree of compliance of the bidding document on the zth parameter item; The bidding document screening unit is used to arrange the bidding documents in descending order according to the degree of compliance, and the bidding documents arranged after the nth position are screened out, where n is a fixed value.

10. The bidding document intelligent identification and screening platform according to claim 7, characterized in that: The bidding category determination module includes: A scheme text processing unit is used to extract the overall project scheme from the text information, and to pre-process and vectorize the text of the overall project scheme; A similarity matrix construction unit is used to construct a similarity matrix based on cosine similarity. For n bidding documents, an n×n similarity matrix S is constructed, where Sij represents the similarity between the i-th and j-th bidding documents. The clustering execution unit is used to determine the number of clusters K and perform K-means clustering, specifically: randomly select K data points as the initial cluster centers, and the initial cluster centers are the starting points of iterations; calculate the distance from each data point to the K cluster centers, and assign each data point to the cluster with the closest cluster center, and recalculate the cluster center of each cluster; until the termination condition is met; The clustering result output unit is used to output and verify the clustering results to obtain several bidding categories.

Images