Intelligent data classification method for chemical engineering cost database

Abstract

The invention relates to the field of intelligent classification of chemical cost databases, in particular to an intelligent data classification method for a chemical cost database, which comprises the following steps: acquiring a chemical material dimension classification result based on a three-dimensional logic judgment system by utilizing the chemical cost database; performing code conversion preprocessing according to the chemical material dimension classification result to obtain a chemical material code conversion standardization processing result; the intelligent data classification result of the chemical cost database is obtained by using the chemical material code conversion standardization processing result, a breakthrough algorithm architecture is established, the classification accuracy is improved to 99% or above, meanwhile, the efficient processing capacity is achieved, concurrent processing of large-batch data is supported, the processing efficiency is improved, and the method is suitable for popularization and application. And meanwhile, dynamic mark low-confidence classification is carried out and abnormal logs are generated, so that the rechecking efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent classification of chemical engineering cost database, and particularly relates to a data intelligent classification method for chemical engineering cost database. BACKGROUND

[0002] In the field of chemical engineering design, cost management is the core link of project cost control, involving the fine classification of tens of thousands of materials, equipment and costs. The traditional method relies on manual consultation of standard documents such as "Petroleum and Chemical Engineering Design Budget Compilation Methods" and "Refinery Process Design Specification", and manually matches material attributes (such as material quality, pressure grade), process parameters and standard numbers, and enters the cost subject code one by one. With the development of chemical industry towards large-scale and fine, new materials (such as nano catalysts, composite materials) and process standards (such as SH / T 3541-2021) are emerging, the traditional manual classification faces the double challenges of efficiency and accuracy: on the one hand, the scale of single project material list can reach tens of thousands (such as 100 million tons of ethylene plant containing 3,872 materials), and the manual classification takes 6-8 hours per thousand, and the error rate is as high as 3-5%; on the other hand, the existing Excel auxiliary tool only supports simple IF / AND / OR logical judgment, and cannot handle complex conditions, resulting in a significant increase in the error rate of non-standard equipment classification. In addition, the classification standards of the financial system and the engineering database are not compatible, forming a data island, and the cross-system integration needs manual secondary mapping, further reducing the efficiency. Therefore, an intelligent classification method based on dynamic rule engine is needed to solve the problems of manual efficiency bottleneck, rule update lag and complex condition processing.

[0003] The current mainstream technical scheme has the following significant defects:

[0004] 1. High dependence on manual work: The traditional method relies on professional personnel to classify one by one, which is inefficient when faced with massive data, and prone to errors (error rate 3-5%) due to fatigue. For example, a coal chemical project mistakenly classified "Incoloy 825 + carbon fiber composite material" as ordinary alloy material, causing cost budget deviation.

[0005] 2. Rule update is rigid: The existing system uses a fixed rule base (such as IF / vlookup nested formula based on Excel), when new materials or process changes occur, the code logic needs to be modified manually, and the update cycle is as long as several weeks. For example, after the release of SH / T3541-2021 standard, it took a design institute 3 weeks to complete the rule adaptation, causing delay in the progress of multiple projects at the same period.

[0006] 3. Insufficient multi-condition coupling capability: existing technologies only support simple logical combinations (AND / OR) and cannot handle hierarchical nesting or multi-criteria compound judgments. Tests show that when faced with a classification rule containing 7 layers of logical conditions ("Equipment location number starts with P and material ≠ 304L or pressure rating ≥ Class 600 and standard number contains GB / T"), the matching accuracy of traditional Excel formulas drops sharply.

[0007] 4. Data island problem: the financial system and the engineering database use different classification standards (e.g., financial subject code is "FG-2314-07" and engineering code is "RMB-2314-07"), existing tools lack automatic mapping mechanism, and cross-system integration requires manual conversion through comparison table, increasing time consumption.

[0008] 5. Lack of abnormality handling: for missing parameters (e.g., pressure rating labeled as "Class-XX") or format error data (e.g., standard number "GB / T50316-2021" miswritten as "GB / T503162021"), existing technologies cannot automatically identify and warn, relying on manual inspection, resulting in serious efficiency loss.

[0009] The above defects seriously restrict the intelligent level of chemical cost management, and it is urgent to realize an automatic, high-precision and strongly adaptive classification solution through technical innovation. SUMMARY

[0010] To overcome the shortcomings of the prior art, the present application provides a data intelligent classification method for a chemical cost database, which realizes automatic classification of material lists through establishment of a three-dimensional logical judgment system, the three-dimensional logical judgment system including three basic dimensions of main item, specialty, and classification, and two extended dimensions of engineering quantity and average cost, improving and enhancing classification accuracy, and supporting concurrent processing of large quantities of data, improving processing efficiency.

[0011] To achieve the above purpose, the present application provides a data intelligent classification method for a chemical cost database, comprising:

[0012] S1, obtaining a chemical material dimension classification result based on a three-dimensional logical judgment system using a chemical cost database;

[0013] S2, performing code conversion preprocessing to obtain a chemical material code conversion standardized processing result according to the chemical material dimension classification result;

[0014] S3, obtaining a data intelligent classification result of the chemical cost database using the chemical material code conversion standardized processing result.

[0015] Preferably, the obtaining of the chemical material dimension classification result based on the three-dimensional logical judgment system using the chemical cost database comprises:

[0016] Obtaining chemical material list data from the chemical cost database;

[0017] Establishing main items, specialties and classifications as chemical material basic dimensions according to the chemical material list data;

[0018] Establishing quantities and prices as chemical material extended dimensions according to the chemical material list data;

[0019] Obtaining chemical material dimension classification results by regular expression processing based on hierarchical nested structure using the chemical material basic dimensions and the chemical material extended dimensions.

[0020] Further, obtaining chemical material dimension classification results by regular expression processing based on hierarchical nested structure using the chemical material basic dimensions and the chemical material extended dimensions includes:

[0021] Obtaining chemical material classification matching processing results by classification matching processing using the chemical material basic dimensions and the chemical material extended dimensions;

[0022] Obtaining chemical material dimension classification results by regular expression processing based on hierarchical nested structure using the chemical material classification matching processing results according to logical operators;

[0023] The classification matching includes logical classification and conditional classification, the logical classification includes mandatory content, optional content, excluded content and classification name, and the logical operators include exclusive or operator and non-operator.

[0024] Further, obtaining chemical material coding conversion standardization processing results by coding conversion preprocessing according to the chemical material dimension classification results includes:

[0025] Obtaining corresponding item identification codes using the chemical material dimension classification results;

[0026] Obtaining corresponding main classification codes using the chemical material dimension classification results;

[0027] Obtaining corresponding main identification codes using the chemical material dimension classification results;

[0028] Obtaining corresponding sequence codes using the chemical material dimension classification results;

[0029] Using the chemical material dimension classification results, item identification codes, main classification codes, main identification codes and sequence codes as chemical material coding conversion results;

[0030] Obtaining chemical material coding conversion standardization processing results by standardization processing according to the chemical material coding conversion results.

[0031] Further, the chemical material code conversion result is standardized to obtain a chemical material code conversion standardized processing result, which includes:

[0032] According to the chemical material code conversion result, a non-standard symbol replacement processing result is obtained by performing non-standard symbol replacement processing.

[0033] According to the non-standard symbol replacement processing result, a unit normalization processing result is obtained by performing unit normalization processing.

[0034] According to the unit normalization processing result, a chemical material code conversion standardized processing result is obtained by performing fuzzy field extraction processing.

[0035] Further, the chemical material code conversion standardized processing result is used to obtain a data intelligent classification result of the chemical cost database, which includes:

[0036] The chemical material code conversion standardized processing result is obtained.

[0037] The chemical material code conversion standardized processing result and the historical chemical material code conversion standardized processing result are used to obtain a primary verification processing result by performing primary verification processing.

[0038] According to the primary verification processing result, a data intelligent classification result of the chemical cost database is obtained.

[0039] Further, the chemical material code conversion standardized processing result and the historical chemical material code conversion standardized processing result are used to obtain a primary verification processing result, which includes:

[0040] If the confidence of the classification data corresponding to the chemical material code conversion standardized processing result and the classification data corresponding to the historical chemical material code conversion standardized processing result is less than a warning threshold, a warning is given, and the chemical material code conversion standardized processing result is output. Otherwise, the current chemical material code conversion standardized processing result is retained.

[0041] The classification data is the classification corresponding data of the chemical material basic dimension, the confidence is the similarity of the classification data corresponding to the chemical material code conversion standardized processing result and the classification data corresponding to the historical chemical material code conversion standardized processing result, and the warning threshold range is 85% to 90%.

[0042] Further, according to the primary verification processing result, a data intelligent classification result of the chemical cost database is obtained, which includes:

[0043] A new and old rule conflict index at the current time is obtained.

[0044] determining whether the new and old rule conflict index of the current time is greater than an automatic migration threshold value, if yes, not processing, otherwise, performing automatic migration processing to obtain an automatic migration processing result;

[0045] performing multi-dimensional test processing using the automatic migration processing result to obtain a multi-dimensional test processing result as a data intelligent classification result of the chemical cost database;

[0046] The calculation formula of the new and old rule conflict index is as follows:

[0047]

[0048] The CI is a new and old rule conflict index, x is the number of rules that exist transformation, and X is the total number of rules;

[0049] The automatic migration threshold value ranges from 6% to 10%.

[0050] Further, performing multi-dimensional test processing using the automatic migration processing result to obtain a multi-dimensional test processing result as a data intelligent classification result of the chemical cost database includes:

[0051] Performing stress test processing using the automatic migration processing result to obtain a stress test processing result;

[0052] Performing precision test processing using the automatic migration processing result to obtain a precision test processing result;

[0053] Performing environmental compatibility test processing using the automatic migration processing result to obtain an environmental compatibility test processing result;

[0054] Determining whether the stress test processing result, the precision test processing result and the environmental compatibility test processing result all meet the test standard, if yes, obtaining the data intelligent classification result of the chemical cost database using the automatic migration processing result, otherwise, returning to S1;

[0055] The test standard of the stress test processing result is that the response time of 100,000 data concurrent processing is not greater than 30 minutes, the test standard of the precision test processing result is that the consistency test degree with the classification result of the chemical cost database is not less than 0.9, and the test standard of the environmental compatibility test processing result is that the Excel full version operation is supported.

[0056] Compared with the closest prior art, the present application has the beneficial effects:

[0057] A breakthrough algorithm architecture is established to support AND / OR / XOR / NOT multi-level logic nesting, the classification accuracy is improved to more than 99%, and meanwhile, high processing capacity is achieved, supporting 10000-20000 concurrent processing per minute, being capable of carrying out real-time classification of 100,000+ items per single project, realizing 30-90% efficiency improvement, and having intelligent error correction and adaptability, double verification process, automatic marking of low confidence classification and generation of exception logs, 30-50% efficiency improvement of manual review, cross-project rule migration, rule reuse rate of 70-90%, and 60% reduction in new project start-up period. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 is a flowchart of a data intelligent classification method for a chemical cost database provided by the present application. DETAILED DESCRIPTION

[0059] The specific embodiments of the present application will be further described in detail below with reference to the accompanying drawings.

[0060] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0061] Embodiment 1:

[0062] The present application provides a data intelligent classification method for a chemical cost database, as shown in Figure 1 , comprising:

[0063] S1, obtaining a chemical material dimension classification result based on a three-dimensional logical judgment system of a chemical cost database;

[0064] S2, obtaining a chemical material coding conversion standardization processing result by coding conversion preprocessing according to the chemical material dimension classification result;

[0065] S3, obtaining a data intelligent classification result of the chemical cost database by using the chemical material coding conversion standardization processing result.

[0066] S1 specifically comprises:

[0067] S1-1, obtaining a chemical material list data by using the chemical cost database;

[0068] S1-2, sequentially establishing a main item, a specialty and a classification as a basic dimension of a chemical material according to the chemical material list data.

[0069] S1-3, sequentially establishing the quantities and prices as the chemical material extended dimensions according to the chemical material list data;

[0070] S1-4, obtaining the chemical material dimension classification result by using the chemical material basic dimension and the chemical material extended dimension based on the hierarchical nested structure for regular expression processing.

[0071] S1-4 specifically includes:

[0072] S1-4-1, obtaining the chemical material classification matching processing result by using the chemical material basic dimension and the chemical material extended dimension for classification matching processing;

[0073] S1-4-2, obtaining the chemical material dimension classification result by using the chemical material classification matching processing result based on the hierarchical nested structure for regular expression processing according to the logical operator;

[0074] Among them, the classification matching includes logical classification and conditional classification, the logical classification includes mandatory content, optional content, exclusion content and classification name, and the logical operator includes exclusive or operator and non-operator.

[0075] S2 specifically includes:

[0076] S2-1, obtaining the corresponding project identification code by using the chemical material dimension classification result;

[0077] S2-2, obtaining the corresponding main classification code by using the chemical material dimension classification result;

[0078] S2-3, obtaining the corresponding main identification code by using the chemical material dimension classification result;

[0079] S2-4, obtaining the corresponding sequence code by using the chemical material dimension classification result;

[0080] S2-5, using the chemical material dimension classification result, the project identification code, the main classification code, the main identification code and the sequence code as the chemical material coding conversion result;

[0081] S2-6, obtaining the chemical material coding conversion standardization processing result by using the chemical material coding conversion result for standardization processing.

[0082] S2-6 specifically includes:

[0083] S2-6-1, obtaining the corresponding non-standard symbol replacement processing result by using the chemical material coding conversion result for non-standard symbol replacement processing;

[0084] S2-6-2, performing unit normalization processing according to the non-symbol replacement processing result to obtain a corresponding unit normalization processing result;

[0085] S2-6-3, performing fuzzy field extraction processing according to the unit normalization processing result to obtain a chemical material code conversion standardization processing result.

[0086] S3 specifically includes:

[0087] S3-1, obtaining a historical chemical material code conversion standardization processing result corresponding to the chemical material code conversion standardization processing result;

[0088] S3-2, performing primary verification processing on the chemical material code conversion standardization processing result and the historical chemical material code conversion standardization processing result to obtain a primary verification processing result

[0089] S3-3, obtaining a data intelligent classification result of the chemical cost database according to the primary verification processing result.

[0090] S3-2 specifically includes:

[0091] S3-2-1, determining whether the confidence of the classification data corresponding to the chemical material code conversion standardization processing result and the classification data corresponding to the historical chemical material code conversion standardization processing result is less than a warning threshold, if yes, performing a warning and outputting the chemical material code conversion standardization processing result, otherwise, retaining the current chemical material code conversion standardization processing result;

[0092] The classification data is classification corresponding data of the chemical material basic dimension, the confidence is the similarity of the classification data corresponding to the chemical material code conversion standardization processing result and the classification data corresponding to the historical chemical material code conversion standardization processing result, and the warning threshold range is 85% to 90%.

[0093] In this embodiment, a data intelligent classification method for a chemical cost database is provided, and the warning threshold range can be determined according to different projects in actual application to meet the requirements.

[0094] S3-3 specifically includes:

[0095] S3-3-1, obtaining a new and old rule conflict index at a current time;

[0096] S3-3-2, determining whether the new and old rule conflict index at the current time is greater than an automatic migration threshold, if yes, not performing processing, otherwise, performing automatic migration processing to obtain an automatic migration processing result;

[0097] S3-3-3, obtaining a multi-dimensional test processing result by using the automatic migration processing result as a data intelligent classification result of the chemical cost database;

[0098] The calculation formula of the new and old rule conflict index is as follows:

[0099]

[0100] The CI is a new and old rule conflict index, x is the number of rules with transformation, and X is the total number of rules;

[0101] The automatic migration threshold range is 6% to 10%.

[0102] In this embodiment, the automatic migration threshold can be appropriately set according to the actual situation in actual application.

[0103] S3-3-3 specifically includes:

[0104] S3-3-3-1, obtaining a stress test processing result by using the automatic migration processing result for stress test processing;

[0105] S3-3-3-2, obtaining a precision test processing result by using the automatic migration processing result for precision test processing;

[0106] S3-3-3-3, obtaining an environment compatibility test processing result by using the automatic migration processing result for environment compatibility test processing;

[0107] S3-3-3-4, determining whether the stress test processing result, the precision test processing result, and the environment compatibility test processing result all meet the test standard, if yes, obtaining a data intelligent classification result of the chemical cost database by using the automatic migration processing result, otherwise, returning to S1;

[0108] The test standard of the stress test processing result is that the response time of 100,000 data concurrent processing is not more than 30 minutes, the test standard of the precision test processing result is that the consistency test degree with the classification result of the chemical cost database is not less than 0.9, and the test standard of the environment compatibility test processing result is that the Excel full version operation is supported.

[0109] In this embodiment, a data intelligent classification method for a chemical cost database includes the following steps:

[0110] Obtaining a feature vector based on a database, a chemical cost name, a feature description, and a process parameter;

[0111] Classify the data based on the predefined labels using the feature vector as a training set according to a decision tree;

[0112] Obtain new project data as data to be classified;

[0113] Use the data to be classified to remove irrelevant features to obtain standard data to be classified;

[0114] Determine whether the standard data to be classified has duplicate type data, if yes, obtain the duplicate type data to establish an updated classification label, otherwise, retain the current classification label;

[0115] Adjust the current historical data weight to 95% of the adjacent last quarter.

[0116] In this embodiment, a data intelligent classification method for a chemical cost database is provided, and the specific system implementation of the method is as follows:

[0117] The dynamic rule matrix engine, the intelligent mapping mechanism, and the abnormality early warning module realize automatic classification of the bill of materials through a preset three-dimensional logical judgment system, and the three-dimensional logical judgment system includes three basic dimensions of main items (WBS), specialties, and classifications (CBS) and two extended dimensions of quantities and unit prices.

[0118] The dynamic rule matrix engine adopts a hierarchical nested structure, the basic layer includes AND / OR logical judgment units, the senior layer introduces XOR / NOT operators, and supports regular expression matching, and the specific rule format is:

[0119] Intelligent matching program name (logical classification area, condition to be classified area)

[0120] The logical area includes {mandatory content, optional content, excluded content, and classification name}.

[0121] The intelligent mapping mechanism includes a standard code conversion table, which automatically converts the classification result into a corresponding cost subject code, and the conversion table includes a ten-level coding system composed of a 6-digit project identification code (T12345), a 12-digit main classification code (P1234-56789-ab), a 4-digit main item identification code (0001), and a 3-digit sequence code.

[0122] The preprocessing module executes a data standardization and cleaning process: non-standard symbol replacement, unit normalization, and fuzzy field extraction.

[0123] The abnormality processing mechanism includes a double-checking module: primary checking: yellow warning is triggered when the classification confidence is less than 90%; and final checking: a manual review interface supports batch correction and rule feedback, and the corrected data automatically updates the knowledge base.

[0124] Support cross-project rule migration, including rule version control system, can automatically calculate the conflict index (CI = Δ rule number / total rule number × 100%) of new and old rules, and trigger the automatic migration protocol when CI ≤ 8%.

[0125] The user interface includes a three-dimensional configuration panel, including: condition area: supporting drag-and-drop logic block combination (AND / OR / XOR); mapping table editor: providing bidirectional conversion of standard encoding and custom encoding; real-time monitoring board: displaying classification progress, abnormal data distribution and rule hit rate.

[0126] The multi-dimensional test system includes: stress test: 100,000 concurrent data processing (response time ≤ 30 minutes); precision test: Kappa consistency test (κ ≥ 0.90) with CBS classification results in the database historical data module; compatibility test: supporting Excel (2003 and above versions) / WPS full version operation.

[0127] The cross-platform data interface: input interface: compatible with CSV / TXT / XML format, automatically parsing device bit number coding rules; output interface: generating XML review files in accordance with "Petroleum and Chemical Construction Project Cost File Preparation Specification"; cloud synchronization module: encrypted transmission of classification results to cloud knowledge base, using AES-256 encryption.

[0128] The dynamic learning model: based on historical classification data to train decision tree model (accuracy ≥ 99.5%); implement incremental learning mechanism: automatically update rule weight coefficient every month;

[0129] a. Extract names, feature descriptions, process parameters, etc. from the database into feature vectors to lay the foundation for the learning model;

[0130] b. Use the decision tree model to classify according to the pre-defined labels, i.e. automatically classify according to features, for example: containing any one of gate valve, stop valve, butterfly valve, check valve, and pressure rating greater than Class600 and not safety valve → classified as "high pressure valve";

[0131] c. Incremental learning mechanism and weight adjustment mechanism:

[0132] ① Every month, new project data automatically enters the classification queue, and data unrelated to existing features are excluded (using the exclusion word dictionary).

[0133] ② If a certain type of data appears continuously in multiple samples, a new classification label is automatically generated.

[0134] ③ The weight of historical data is reduced by 5% every quarter to ensure that the model focuses on the classification of recent data.

[0135] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0136] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0137] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0138] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0139] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A data intelligent classification method for chemical cost database, characterized in that: include: S1. Obtain the chemical material dimensional classification results based on the three-dimensional logic judgment system using the chemical cost database; S2. Perform code conversion preprocessing according to the chemical material dimensional classification results to obtain a chemical material code conversion standardization processing result; S3. Obtain data intelligent classification results of the chemical cost database using the chemical material coding conversion standardization processing results.

2. The data intelligent classification method for chemical cost database according to claim 1, characterized in that: The method of obtaining the chemical material dimension classification results based on the three-dimensional logic judgment system using the chemical cost database includes: Obtaining chemical material list data using the chemical cost database; According to the chemical material list data, main items, specialties and categories are sequentially established as basic dimensions of chemical materials; According to the chemical material list data, engineering quantity and total price are sequentially established as chemical material extension dimensions; The chemical material basic dimension and the chemical material extended dimension are used to perform regular expression processing based on a hierarchical nested structure to obtain a chemical material dimension classification result.

3. The data intelligent classification method for chemical cost database according to claim 2, characterized in that: The chemical material dimension classification results obtained by performing regular expression processing based on the hierarchical nested structure using the chemical material basic dimension and the chemical material extended dimension include: Using the chemical material basic dimension and the chemical material extended dimension to perform classification matching processing to obtain a chemical material classification matching processing result; Utilizing the chemical material classification matching processing result, regular expression processing is performed based on a hierarchical nested structure according to a logical operator to obtain a chemical material dimension classification result; The classification matching includes logical classification and conditional classification, the logical classification includes required content, optional content, excluded content and classification name, and the logical operator includes an exclusive-OR operator and a negation operator.

4. The data intelligent classification method for chemical cost database according to claim 3, characterized in that: The coding conversion preprocessing is performed according to the chemical material dimensional classification results to obtain the chemical material coding conversion standardization processing results including: Obtaining a corresponding project identification code using the chemical material dimension classification result; Obtaining a corresponding primary classification code using the chemical material dimension classification result; Obtaining a corresponding primary identification code using the chemical material dimension classification result; Obtaining corresponding sequence codes using the chemical material dimension classification results; Utilize the chemical material dimension classification result to correspond to the project identification code, main classification code, main identification code and sequence code as the chemical material coding conversion result; A standardization process is performed on the chemical material code conversion result to obtain a chemical material code conversion standardization process result.

5. The data intelligent classification method for chemical cost database according to claim 4, characterized in that: The chemical material code conversion standardization processing result obtained by performing standardization processing according to the chemical material code conversion result includes: Perform non-standard symbol replacement processing according to the chemical material coding conversion result to obtain a corresponding non-standard symbol replacement processing result; Performing unit normalization processing on the non-standard symbol replacement processing result to obtain a corresponding unit normalization processing result; Fuzzy field extraction processing is performed based on the unit normalization processing result to obtain the chemical material code conversion standardization processing result.

6. The data intelligent classification method for chemical cost database according to claim 3, characterized in that: The data intelligent classification results of the chemical cost database obtained by using the chemical material coding conversion standardization processing results include: Obtain the chemical material code conversion standardization processing results corresponding to the historical chemical material code conversion standardization processing results; Primary verification processing is performed using the chemical material code conversion standardization processing result and the historical chemical material code conversion standardization processing result to obtain a primary verification processing result The data intelligent classification result of the chemical cost database is obtained according to the primary verification processing result.

7. A data intelligent classification method for a chemical cost database according to claim 6, characterized in that: The primary verification processing is performed using the chemical material code conversion standardization processing result and the historical chemical material code conversion standardization processing result to obtain the primary verification processing result, which includes: Determine whether the confidence level of the classification data corresponding to the chemical material code conversion standardization processing result and the classification data corresponding to the historical chemical material code conversion standardization processing result is less than the warning threshold. If so, issue a warning and output the chemical material code conversion standardization processing result. Otherwise, retain the current chemical material code conversion standardization processing result. Among them, the classification data is the classification corresponding data of the basic dimension of chemical materials, the confidence level is the degree of similarity between the classification data corresponding to the chemical material code conversion standardization processing results and the classification data corresponding to the historical chemical material code conversion standardization processing results, and the warning threshold range is 85% to 90%.

8. The data intelligent classification method for chemical cost database according to claim 7, characterized in that: The data intelligent classification results of the chemical cost database obtained according to the primary verification processing results include: Get the conflict index of new and old rules at the current moment; Determine whether the conflict index between the new and old rules at the current moment is greater than the automatic migration threshold; if so, do not perform any processing; otherwise, perform automatic migration processing to obtain an automatic migration processing result; Using the automatic migration processing results to perform multi-dimensional test processing to obtain multi-dimensional test processing results as data intelligent classification results of the chemical cost database; The calculation formula of the new and old rule conflict index is as follows: The CI is the conflict index between the new and old rules, x is the number of rules with transformation, and X is the total number of rules; The auto-migration threshold ranges from 6% to 10%.

9. The method for intelligent data classification for a chemical cost database according to claim 8, characterized in that: The multi-dimensional test processing is performed using the automatic migration processing result to obtain the multi-dimensional test processing result as the data intelligent classification result of the chemical cost database, including: Performing a stress test using the automatic migration processing result to obtain a stress test processing result; Performing an accuracy test using the automatic migration processing result to obtain an accuracy test processing result; Performing an environmental compatibility test process using the automatic migration process result to obtain an environmental compatibility test process result; Determine whether the pressure test processing results, the accuracy test processing results, and the environmental compatibility test processing results all meet the test standards. If so, use the automatic migration processing results to obtain the data intelligent classification results of the chemical cost database. Otherwise, return to S1; Among them, the test standard for the stress test processing results is that the response time for concurrent processing of 100,000 data is no more than 30 minutes, the test standard for the accuracy test processing results is that the consistency test degree with the classification results of the chemical cost database is no less than 0.9, and the test standard for the environmental compatibility test processing results is to support the operation of all versions of Excel.

Images