Enterprise process matching method and system based on support vector machine, and storable medium

Automatically match the enterprise quota library and the estimated budget process through the support vector machine model, solving the problem of inaccurate matching in the existing technology, improving matching accuracy and efficiency, and enhancing the competitiveness of engineering project management.

CN118966618APending Publication Date: 2024-11-15INTELLIGENT & TECHNOLOGY BRANCH OF CHINA RAILWAY FIRST GROUP CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410946738.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-15
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

In the prior art, the name matching of the enterprise quota database and the estimated budget process is not comprehensive enough, resulting in inaccurate matching and time-consuming and labor-consuming.

Method used

Using the support vector machine-based method, by obtaining and preprocessing process data and estimated process data, calculating the similarity and comparing it with the preset threshold, re-identifying it using the support vector machine model to realize the automatic matching of the enterprise quota library and the estimated process. Related to avoid manual matching errors.

Benefits of technology

It improves the accuracy and efficiency of matching, reduces cost waste, improves the efficiency and competitiveness of project management, and ensures competitiveness in the engineering market.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118966618A_ABST
    Figure CN118966618A_ABST
Patent Text Reader

Abstract

The invention provides an enterprise process matching method and system based on a support vector machine and a storable medium, and relates to the technical field of data processing, and the method comprises the following steps: obtaining process data included in a plurality of enterprise quota libraries and rough calculation process data of a project; performing feature extraction on the process data and the budget process data to obtain corresponding process feature vector data and budget process feature vector data, and preprocessing the budget process feature vector data and the process feature vector data; calculating similarity data of the preprocessed rough calculation process feature vector data and the process feature vector data, comparing the similarity data with a preset threshold value, and determining whether matching succeeds or not according to a comparison result; according to the method, automatic matching and association of the enterprise quota library and the budget process are achieved, and errors and inconsistency possibly generated in the manual matching process are avoided.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more specifically to an enterprise process matching method, system and storable medium based on support vector machine. Background Art

[0002] At present, in enterprise project management, the enterprise quota library usually refers to the quota standards corresponding to various processes formulated by the enterprise based on its own business conditions and production and operation characteristics. The enterprise quota library can help enterprises better improve operating efficiency, optimize resource allocation, and enhance the competitiveness and profitability of enterprises. The project budget process refers to the process of preliminary estimation, analysis and accounting of the cost of the project based on the existing planning and design schemes and related materials in the initial stage of project construction. By determining the project budget process, it can provide a reference for the subsequent construction schedule.

[0003] However, the existing technology has the problem of insufficient matching between the enterprise quota library and the name of the budget process. Although the enterprise may have a large quota library and detailed budget processes, the lack of effective algorithm support in the matching process often leads to problems such as inaccurate matching, time-consuming and labor-intensive.

[0004] Therefore, how to provide an enterprise process matching method that can solve the above problems is an issue that technical personnel in this field urgently need to solve. Summary of the invention

[0005] In view of this, the present invention provides an enterprise process matching method, system and storable medium based on support vector machine, which realizes automatic matching and association between enterprise quota library and budgetary process, and avoids errors and inconsistencies that may occur in the manual matching process.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A method for enterprise process matching based on support vector machine comprises the following steps:

[0008] Obtain process data included in multiple enterprise quota databases, as well as project budget process data;

[0009] Performing feature extraction on the process data and the estimated process data respectively to obtain corresponding process feature vector data and estimated process feature vector data, and preprocessing the estimated process feature vector data and the process feature vector data;

[0010] The similarity data between the pre-processed estimated process feature vector data and the process feature vector data is calculated, and the similarity data is compared with a preset threshold value, and whether the match is successful is determined according to the comparison result.

[0011] Preferably, the specific process of determining whether the match is successful according to the comparison result includes:

[0012] The similarity data is judged against a preset threshold;

[0013] When the similarity data is less than or equal to a preset threshold, the match is determined to be successful;

[0014] The successfully matched estimated process data and the enterprise quota library corresponding to the process data are bound.

[0015] Preferably, the specific process of determining whether the match is successful according to the comparison result also includes:

[0016] When the similarity data is greater than a preset threshold, it is determined that the match fails;

[0017] Constructing a support vector machine model, using the support vector machine model to re-recognize the similarity data, and determining whether the match is successful based on the recognition result;

[0018] If the identification result determines that the match is successful, the successfully matched estimated process data and the enterprise quota library corresponding to the process data are bound.

[0019] Preferably, the specific processing process of determining whether the match is successful according to the judgment result also includes:

[0020] When there are multiple similarity data less than or equal to the preset threshold, the matching result is determined to be uncertain;

[0021] The plurality of similarity data are input into the support vector machine model, the support vector machine model is used to identify the similarity data again, and whether the matching is successful is determined according to the identification result.

[0022] Preferably, the specific processing process of constructing the support vector machine model includes:

[0023] Acquire successfully matched historical budget process data as a training data set, and set labels for the training data set;

[0024] A support vector machine model is constructed, and the support vector machine model is trained using the training data set.

[0025] The present invention also provides an enterprise process matching system based on a support vector machine, comprising:

[0026] The acquisition module is used to obtain the process data included in the quota database of multiple enterprises and the estimated process data of the project;

[0027] An extraction module, used to perform feature extraction on the process data and the estimated process data respectively, obtain corresponding process feature vector data and estimated process feature vector data, and pre-process the estimated process feature vector data and the process feature vector data;

[0028] The matching module is used to calculate the similarity data between the pre-processed estimated process feature vector data and the process feature vector data, and compare the similarity data with a preset threshold value, and determine whether the match is successful based on the comparison result.

[0029] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the enterprise process matching method as described in any one of the above items is implemented.

[0030] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method, system and storable medium for enterprise process matching based on support vector machines. By using similarity and support vector machine models, it has the ability to handle exceptions for matching failures or uncertain situations, and can more accurately match the processes in the quota library with the estimated processes, avoiding errors caused by manual matching, thereby improving the accuracy of bidding; improving matching accuracy and saving time costs, so that subsequent bidders can more accurately and quickly formulate reasonable cost estimates and construction plans, thereby improving competitiveness in the fiercely competitive engineering market; the present invention can effectively improve the accuracy and efficiency of matching, provide a more accurate and efficient cost estimation method for enterprise project management, and help reduce cost waste and improve project management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0032] Figure 1 The overall flow chart of the enterprise process matching method based on support vector machine provided by the present invention;

[0033] Figure 2 A structural schematic diagram of an enterprise process matching system based on a support vector machine provided by the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] See also Figure 1 As shown, the embodiment of the present invention discloses a method for enterprise process matching based on support vector machine, comprising the following steps:

[0036] Obtain process data included in multiple enterprise quota databases, as well as project budget process data;

[0037] Extracting features from the process data and the estimated process data respectively to obtain corresponding process feature vector data and the estimated process feature vector data, and preprocessing the estimated process feature vector data and the process feature vector data;

[0038] The similarity data between the preprocessed estimated process feature vector data and the process feature vector data is calculated, and the similarity data is compared with a preset threshold value, and whether the match is successful is determined based on the comparison result.

[0039] Specifically, similarity measures the similarity between strings by calculating the minimum number of editing operations required to convert one string into another. Editing operations include inserting, deleting, and replacing characters. The specific process of calculating similarity data may include the following steps:

[0040] Assume that the enterprise quota database includes a process called concrete pouring and a budget process called concrete mixing. Convert the two processes into corresponding strings A and B.

[0041] 1. Initialize a two-dimensional array dp of size (6+1)×(6+1), because the length of string A and string B are both 6.

[0042] 2. Initialize the first row and first column

[0043] Initialize the first row of the array from 0 to 6, representing the number of insertion operations required to convert an empty string to string B. Similarly, initialize the first column of the array from 0 to 6, representing the number of insertion operations required to convert an empty string to string A.

[0044] Copy code

[0045]

[0046]

[0047] 3. Filling a two-dimensional array

[0048] Starting from the second row and second column of the array, compare the characters of string A and B one by one, and calculate the edit distance according to the following rules:

[0049] If A[i - 1] is equal to B[j - 1], then dp[i][j] = dp[i - 1][j - 1], indicating that no editing operation is required;

[0050] Otherwise, dp[i][j] = min(dp[i - 1][j], dp[i][j - 1], dp[i - 1][j - 1]) + 1, indicating that an insertion, deletion, or replacement operation is required.

[0051] Specific calculation steps: Compare "混" and "混":

[0052] Equal, dp[1][1] = dp[0][0] = 0

[0053]

[0054] Compare "混" and "凝":

[0055] Not equal, dp[1][2] = min(dp[0][2], dp[1][1], dp[0][1]) + 1 = 1

[0056]

[0057] Compare "混" and "土":

[0058] Not equal, dp[1][3] = min(dp[0][3], dp[1][2], dp[0][2]) + 1 = 2

[0059]

[0060] Compare "混" and "浇":

[0061] Not equal, dp[1][4] = min(dp[0][4], dp[1][3], dp[0][3]) + 1 = 3

[0062]

[0063] Compare "混" and "筑":

[0064] Not equal, dp[1][5] = min(dp[0][5], dp[1][4], dp[0][4]) + 1 = 4

[0065]

[0066] Compare "mix" and "stir":

[0067] Not equal, dp[1][6] = min(dp[0][6], dp[1][5], dp[0][5]) + 1 = 5

[0068]

[0069]

[0070] Compare "coagulate" and "mix":

[0071] Not equal, dp[2][1] = min(dp[1][1], dp[2][0], dp[1][0]) + 1 = 1

[0072]

[0073] Compare "coagulate" and "coagulate":

[0074] Equal, dp[2][2] = dp[1][1] = 0

[0075]

[0076] Compare "coagulate" and "soil":

[0077] Not equal, dp[2][3] = min(dp[1][3], dp[2][2], dp[1][2]) + 1 = 1

[0078]

[0079]

[0080] Compare "coagulate" and "pour":

[0081] Not equal, dp[2][4] = min(dp[1][4], dp[2][3], dp[1][3]) + 1 = 2

[0082]

[0083] Compare "coagulate" and "build":

[0084] Not equal, dp[2][5] = min(dp[1][5], dp[2][4], dp[1][4]) + 1 = 3

[0085]

[0086] Compare "coagulate" and "stir":

[0087] Not equal, dp[2][6] = min(dp[1][6], dp[2][5], dp[1][5]) + 1 = 4

[0088]

[0089] Repeat the above steps to calculate the edit distance of each position (i, j) in turn.

[0090] The final filled two-dimensional array dp is as follows:

[0091]

[0092] The final edit distance is the value of the array dp[6][6], indicating that the minimum number of edit operations required to convert string A to string B is 2. Possible operations include replacing the character "pour" with "stir", and replacing the character "build" with "mix".

[0093] In a specific embodiment, the specific processing procedure for determining whether the matching is successful according to the comparison result includes:

[0094] Judge the similarity data with the preset threshold;

[0095] When the similarity data is less than or equal to the preset threshold, it is determined that the matching is successful;

[0096] Bind the estimated process data with successful matching and the enterprise quota library corresponding to the process data.

[0097] In a specific embodiment, the specific processing procedure for determining whether the matching is successful according to the comparison result further includes:

[0098] When the similarity data is greater than the preset threshold, it is determined that the matching fails;

[0099] Construct a support vector machine model, use the support vector machine model to re-identify the similarity data, and determine whether the matching is successful according to the recognition result;

[0100] If the recognition result determines that the matching is successful, bind the estimated process data with successful matching and the enterprise quota library corresponding to the process data.

[0101] In a specific embodiment, the specific processing procedure for determining whether the matching is successful according to the judgment result further includes:

[0102] When there are multiple similarity data less than or equal to the preset threshold, it is determined that the matching result is uncertain;

[0103] Input the multiple similarity data into the support vector machine model, use the support vector machine model to re-identify the similarity data, and determine whether the matching is successful according to the recognition result.

[0104] In a specific embodiment, the specific process of constructing the support vector machine model includes:

[0105] Obtain the successfully matched historical budget process data as a training data set, and set labels for the training data set;

[0106] Build a support vector machine model and train it using the training data set.

[0107] Specifically, the specific processing process of constructing the support vector machine model can also optimize the support vector machine model, which may specifically include the following steps:

[0108] 1. Improvement of input features

[0109] Text feature extraction: When processing process names, in addition to conventional character-level and word-level features, more natural language processing (NLP) features can be introduced, such as TF-IDF and word embeddings. These features can better capture the semantic information of process names.

[0110] Edit distance as a feature: Inputting the edit distance calculation result into the SVM model as a feature can help the model better understand the basic similarity between names. The improved feature engineering can more comprehensively capture the similarity between process names, thereby improving the classification accuracy and matching accuracy of the SVM model.

[0111] 2. Kernel function optimization

[0112] Common kernel functions include linear kernel, polynomial kernel and radial basis function (RBF) kernel. According to the specific process name data distribution, the optimal kernel function can be selected through cross-validation, or multiple kernel functions can be combined (for example, using kernel techniques) to enhance the nonlinear expression ability of the model.

[0113] Kernel parameter adjustment: Optimize the parameters of the kernel function (such as the γ parameter of the RBF kernel and the degree of the polynomial kernel). The optimal parameter combination can be found through grid search or random search.

[0114] The optimized kernel function and parameters can improve the nonlinear ability of the decision boundary of the SVM model, enabling it to better handle complex name matching problems.

[0115] 3. Data enhancement and processing

[0116] Data enhancement: Increase the diversity and magnitude of data by generating more training samples (such as replacing synonyms and character perturbations on process names).

[0117] Sample balancing: When dealing with possible class imbalance, undersampling, oversampling, or synthetic minority class techniques (such as SMOTE) can be used to balance the training data.

[0118] Data augmentation and sample balancing techniques can improve the generalization ability and robustness of the model and avoid the overfitting problem of the model on small samples or unbalanced data sets.

[0119] 4. Model Ensemble

[0120] Ensemble learning: You can try to combine other machine learning algorithms (such as random forest, gradient boosting tree, etc.) and use ensemble learning methods (such as soft voting, stacking, etc.) to improve the performance of the overall model.

[0121] Model fusion: Fusion of SVM with other similarity calculation methods (such as word vector similarity, context similarity, etc.) to improve the final matching accuracy through weighted averaging or other fusion strategies.

[0122] Ensemble learning and model fusion can fully utilize the advantages of different algorithms and improve the overall performance and robustness of the model.

[0123] 5. Online learning and adaptive adjustment

[0124] An online learning mechanism is introduced into the model so that it can continuously learn and update from new data to adapt to the changing process names and matching requirements.

[0125] Adaptive adjustment: Dynamically adjust and optimize the model based on the matching results and feedback in actual applications to ensure that the model always maintains high matching accuracy.

[0126] Online learning and adaptive adjustment enable the model to have dynamic adaptability, which can maintain the stability and improvement of matching performance in long-term applications.

[0127] See also Figure 2 As shown, an embodiment of the present invention further provides a matching system of an enterprise process matching method based on a support vector machine using any one of the above embodiments, comprising:

[0128] The acquisition module is used to obtain the process data included in the quota database of multiple enterprises and the estimated process data of the project;

[0129] An extraction module is used to extract features from the process data and the estimated process data respectively, obtain corresponding process feature vector data and the estimated process feature vector data, and pre-process the estimated process feature vector data and the process feature vector data;

[0130] The matching module is used to calculate the similarity data between the preprocessed estimated process feature vector data and the process feature vector data, and compare the similarity data with a preset threshold value, and determine whether the match is successful based on the comparison result.

[0131] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the enterprise process matching method as described in any one of the above embodiments is implemented.

[0132] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0133] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for enterprise process matching based on support vector machine, characterized in that: The following steps are involved: Obtain process data included in multiple enterprise quota databases, as well as project budget process data; Performing feature extraction on the process data and the estimated process data respectively to obtain corresponding process feature vector data and estimated process feature vector data, and preprocessing the estimated process feature vector data and the process feature vector data; The similarity data between the pre-processed estimated process feature vector data and the process feature vector data is calculated, and the similarity data is compared with a preset threshold value, and whether the match is successful is determined according to the comparison result.

2. The enterprise process matching method based on support vector machine according to claim 1 is characterized in that: The specific process of determining whether a match is successful based on the comparison results includes: The similarity data is judged against a preset threshold; When the similarity data is less than or equal to a preset threshold, the match is determined to be successful; The successfully matched estimated process data and the enterprise quota library corresponding to the process data are bound.

3. The enterprise process matching method based on support vector machine according to claim 2 is characterized in that: The specific process of determining whether the match is successful based on the comparison results also includes: When the similarity data is greater than a preset threshold, it is determined that the match fails; Constructing a support vector machine model, using the support vector machine model to re-recognize the similarity data, and determining whether the match is successful based on the recognition result; If the identification result determines that the match is successful, the successfully matched estimated process data and the enterprise quota library corresponding to the process data are bound.

4. The enterprise process matching method based on support vector machine according to claim 3 is characterized in that: The specific process of determining whether the match is successful based on the judgment result also includes: When there are multiple similarity data less than or equal to the preset threshold, the matching result is determined to be uncertain; The plurality of similarity data are input into the support vector machine model, the support vector machine model is used to identify the similarity data again, and whether the matching is successful is determined according to the identification result.

5. The enterprise process matching method based on support vector machine according to claim 3 is characterized in that: The specific process of building a support vector machine model includes: Acquire successfully matched historical budget process data as a training data set, and set labels for the training data set; A support vector machine model is constructed, and the support vector machine model is trained using the training data set.

6. A matching system using the enterprise process matching method based on support vector machine according to any one of claims 1 to 5, characterized in that: include: The acquisition module is used to obtain the process data included in the quota database of multiple enterprises and the estimated process data of the project; An extraction module, used to perform feature extraction on the process data and the estimated process data respectively, obtain corresponding process feature vector data and estimated process feature vector data, and pre-process the estimated process feature vector data and the process feature vector data; The matching module is used to calculate the similarity data between the pre-processed estimated process feature vector data and the process feature vector data, and compare the similarity data with a preset threshold value, and determine whether the match is successful based on the comparison result.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the enterprise process matching method according to any one of claims 1 to 5 is implemented.