An intelligent regional summary accounting method for ocean industry data

Abstract

The present invention relates to the technical field of marine data management, and specifically discloses an intelligent regional summary accounting method for marine industry data, including setting a characteristic time period based on changes in target data, generating a recommendation coefficient based on changes in marine gross domestic product of the target data and multiple matching results within the characteristic time period; selecting several subcategories from the multiple matching results as recommended categories of the target data according to the recommendation coefficient; performing verification analysis on the selected several recommended categories, and using a subcategory that passes the verification analysis as the verification category of the target data. The present invention processes marine industry data that is not in a comparison table of marine and related industries through in-depth data utilization and analysis, thereby solving the problem that the national economic industry categories to which some marine-related units belong are not in the comparison table of marine and related industries and are difficult to classify directly.

Description

Technical Field

[0001] The present invention relates to the technical field of marine data management, and in particular to an intelligent regional summary accounting method for marine industry data. Background Art

[0002] Monitoring of marine-related enterprises is one of the important basic sources for calculating marine gross domestic product in marine data management, and provides support for determining the stripping coefficient, i.e. the proportion of marine-related operating income, and the extrapolated growth rate of marine industry added value calculation in non-economic years.

[0003] The main problems in the existing accounting practices include that some marine-related data belong to national economic industry categories that are not included in the marine and related industries comparison table. Such marine-related unit output values ​​are part of the marine gross domestic product accounting that is not included, that is, under the current marine and related industries classification standard system, they are not yet within the marine gross domestic product accounting scope. For example, the national economic industry "communication system equipment manufacturing" (3921) is not included in the marine and related industries classification standard. According to the marine production activities it engages in, it should belong to the marine information service industry and marine-related equipment manufacturing industry.

[0004] This type of marine-related data is difficult to classify directly and needs to be manually checked. Manual checking often requires multiple verifications or even field inspections to avoid inaccurate classification of marine industry attributes of marine-related units.

[0005] In view of this, the present invention proposes an intelligent regional summary accounting method for marine industry data. Through in-depth data utilization and analysis, the marine industry data that is not in the marine and related industries comparison table is processed, thereby solving the problem that the national economic industry categories of some marine-related units are not in the marine and related industries comparison table and are difficult to classify directly. Summary of the invention

[0006] The purpose of the present invention is to provide an intelligent regional summary accounting method for marine industry data to solve the following technical problems:

[0007] How to solve the problem that the national economic industry categories of some marine-related units are not included in the marine and related industries comparison table, making it difficult to classify them directly.

[0008] The purpose of the present invention can be achieved by the following technical solutions:

[0009] An intelligent regional summary accounting method for marine industry data, including:

[0010] During the monitoring period, the marine industry data is collected and sorted according to the monitoring period, and the acquired marine industry data is directly classified according to the preset table, which includes two classification modes: major category and minor category;

[0011] Obtain the marine industry data that cannot be directly classified through a preset table as target data, and perform semantic matching based on the name of the target data and the subclass names in the preset table, and then obtain multiple matching results;

[0012] Then, set a characteristic period based on the change of the target data, and generate a recommendation coefficient based on the change of the gross marine product value of the target data and the multiple matching results within the characteristic period;

[0013] Select several subclasses from the multiple matching results as the recommended classification of the target data, and the recommended classification is used to provide the classification for the verification personnel to give priority to manual verification;

[0014] Perform verification analysis on the selected several recommended classifications, and use one subclass that passes the verification analysis as the verification classification of the target data, and the verification classification is used to skip the manual verification of the current target data for the verification personnel.

[0015] Through the above technical solution: Analyze and process the marine industry data that cannot be directly classified according to the preset table through two processes of recommendation analysis and verification analysis, which can directly determine the corresponding subclass of the marine industry data in some cases, and give several recommended classifications when the subclass cannot be directly given, thus greatly reducing the workload of subsequent verification personnel and improving the verification efficiency.

[0016] As a further technical solution of the present invention: The process of performing semantic matching and then obtaining multiple matching results includes:

[0017] Perform word segmentation on the target text, and set corresponding word vectors for each word segment. The target text includes the title text of the target data and the classification title text of the subclass;

[0018] Obtain the semantic vector of the target text based on the multiple word vectors;

[0019] Obtain the vector distance between the semantic vector of the target data and the semantic vector of the classification title of the subclass, and match and obtain multiple matching results from the classification titles of the subclasses based on a preset judgment distance.

[0020] As a further technical solution of the present invention: The process of setting a characteristic period based on the change of the target data includes:

[0021] Obtain the gross marine product value of the target data within several monitoring periods;

[0022] Set a change threshold, obtain the difference between the gross marine product value in one monitoring period and the gross marine product value in the previous monitoring period, and compare the numerical part of the difference with the change threshold;

[0023] Set the monitoring periods in which the numerical part of the difference exceeds the change threshold and the difference signs are the same in multiple adjacent monitoring periods as characteristic periods.

[0024] As a further technical solution of the present invention: The process of generating a recommendation coefficient based on the changes in the gross ocean product within the characteristic periods of the target data and multiple matching results includes:

[0025] Calculate and obtain the average change amount of the target data and multiple matching results in each monitoring period within the characteristic periods.

[0026] Set standard change amounts for the target data and multiple matching results respectively according to the contemporaneous data in the historical data.

[0027] Obtain a reference value based on the ratio of the average change amount within the characteristic periods corresponding to the target data and multiple matching results to the corresponding standard change amounts.

[0028] Obtain the number of differences in which the sign of the difference between the gross ocean product in one monitoring period and the gross ocean product in the previous monitoring period within the characteristic periods of multiple matching results is different from the sign of the corresponding difference of the target data, and compare this number of differences with the total number of differences generated by the characteristic periods based on the target data to obtain a coefficient value.

[0029] Calculate and obtain the recommendation coefficient according to the reference value and the coefficient value.

[0030] Through the above technical solution: Recommendation verification is carried out by obtaining the recommendation coefficient. The recommendation coefficient is obtained based on the target data with unknown classification. First, multiple relevant sub - classes are obtained through semantic matching, and then several relevant sub - classes with the closest change status are selected as recommended classifications by comparing the change status of multiple relevant sub - classes with the target data. In this way, a suitable classification can be quickly found among a large number of sub - classes, thus realizing intelligence and making in - depth use of the data.

[0031] As a further technical solution of the present invention: The process of selecting several sub - classes from multiple matching results as the recommended classification of the target data includes:

[0032] Set a recommendation threshold and compare the recommendation coefficient with the recommendation threshold.

[0033] If the recommendation coefficient is not less than the recommendation threshold, then use the matching result corresponding to this recommendation coefficient as one of the recommended classifications.

[0034] If the recommendation coefficient is less than the recommendation threshold, then do not recommend this matching result.

[0035] As a further technical solution of the present invention: The process of verifying and analyzing the selected several recommended classifications includes:

[0036] If the number of recommended classifications is zero, no verification analysis is performed, and the output verification classification is none;

[0037] If the number of recommended classifications is not less than 1, obtain the list of target data source units, their business data, and the list of business units registered under the corresponding major categories of the recommended classifications;

[0038] Then, based on the income of the target data source units and the missing rate of the source units by the units registered under the corresponding major categories, obtain the verification coefficient;

[0039] Based on the verification coefficient, obtain the verification analysis result, which includes passing the verification and failing the verification.

[0040] As a further technical solution of the present invention: The process of obtaining the verification coefficient includes:

[0041] Obtain the total offshore income of the target data source units and the income of the industries corresponding to the business target data, obtain the proportion of the income of the industries corresponding to the business target data in the total offshore income, and calculate the average value of this proportion for all target data source units;

[0042] Select several business scopes with the highest repetition rate of the offshore business scope of the target data source units as the target scopes;

[0043] Select several business units that include the target scope from the business units corresponding to the recommended classifications, and set a standard proportion based on the historical business data of the target units. The standard proportion is the average value of the ratio of the income of the industries corresponding to the recommended classifications of the target units to the total offshore income of the target units;

[0044] Set a first weight coefficient for the ratio of the average value of this proportion to the standard proportion, set a second weight value for the missing rate of the source units by the enterprises registered under the corresponding major categories of the target data, and perform weighted summation to obtain the verification coefficient.

[0045] Through the above technical solution: Further verify several recommended classifications through the verification coefficient. The recommended classifications are obtained based on the data change status. Obviously, there will be two sub-categories with different classifications but similar changes in the gross ocean product value. The verification coefficient of the present invention is obtained based on the missing rate of the enterprises corresponding to the target data among all the enterprises corresponding to the recommended classifications and the proportion of the income of the industries corresponding to the target data in the total income of the enterprises. By adding the enterprise's offshore income and offshore registration status as verification dimensions, further verify the recommended classifications. Under the condition of setting a reasonable status in the verification interval, it can accurately find the sub-categories that match the target data in the recommended classifications, thus achieving further verification analysis.

[0046] As a further technical solution of the present invention: The process of obtaining the verification analysis result based on the verification coefficient includes:

[0047] Based on the number of recommended categories, the validation coefficient is compared with the validation interval;

[0048] If the number of recommended categories is 1, the verification will be passed if the verification coefficient is not greater than the right endpoint of the verification interval, otherwise it will fail;

[0049] If the number of recommended categories is greater than 1, the verification passes if the verification coefficient is less than the left endpoint of the verification interval, otherwise it fails.

[0050] As a further technical solution of the present invention: it also includes data recording of recommended classifications and verified classifications, and at the same time, classifying and storing the marine industry data collected and obtained during the monitoring period according to the monitoring time period.

[0051] Beneficial effects of the present invention:

[0052] (1) The present invention obtains marine industry data that cannot be directly classified according to preset tables for analysis and processing through two processes: recommendation analysis and verification analysis. In some cases, the present invention can directly determine the subcategories corresponding to the marine industry data, and give several recommended classifications when the subcategories cannot be given directly, thereby greatly reducing the workload of subsequent verification personnel and improving verification efficiency.

[0053] (2) The present invention verifies the recommendation by obtaining the recommendation coefficient. The recommendation coefficient is obtained based on the target data with unknown classification. It first obtains multiple related subcategories through semantic matching, and then compares the change status of the multiple related subcategories with the target data to screen out several related subcategories with the closest change status as recommended categories. In this way, it can quickly find the appropriate classification among a large number of subcategories, thereby realizing intelligence and making in-depth use of data.

[0054] (3) The verification coefficient of the present invention is obtained based on the missing rate of the target data corresponding to all enterprises in the recommended classification and the proportion of the target data corresponding to the industry income in the total enterprise income. The enterprise's marine-related income and marine-related registration status are added as verification dimensions to further verify the recommended classification. When the verification interval is set reasonably, it can accurately find the subcategories in the recommended classification that are consistent with the target data, thereby realizing further verification analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The present invention will be further described below in conjunction with the accompanying drawings.

[0056] Figure 1 It is a flow chart of the overall steps of the accounting method of the present invention;

[0057] Figure 2 It is a flow chart of the steps of obtaining the recommended coefficients of the present invention. DETAILED DESCRIPTION

[0058] 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.

[0059] See also Figure 1 - Figure 2 As shown, in one embodiment, a smart regional summary accounting method for marine industry data is provided, including:

[0060] S100. Collect and sort the marine industry data according to the monitoring period during the monitoring cycle, and directly classify the acquired marine industry data according to the preset table. The preset table includes two classification modes: major category and minor category. The monitoring period can be set to day or week without specific restrictions. The monitoring period is the interval between two statistical accounting behaviors. The collection channels include: sharing by marine-related departments, mainly the national economic accounting reporting system of the statistical department, and the data sharing of the professional statistical reporting system of marine-related departments such as tourism, agriculture, and transportation; relevant data collected by large marine-related enterprise groups such as the Salt Industry Group, CNOOC, the Energy Bureau, and the Shipbuilding Industry Corporation; data submitted by the marine economic statistical survey and the marine gross domestic product accounting reporting system; and marine-related enterprise data obtained from marine economic operation monitoring. The preset table is a comparison table of marine and related industries, which usually includes two classification modes: major category and minor category. For example, the marine aquaculture minor category under the marine fishery major category is one of the bases for dividing the attributes of the marine industry. Direct classification requires that the category name corresponding to the marine industry data is the same as that recorded in the preset table;

[0061] S200, obtaining marine industry data that cannot be directly classified through the preset table as target data, and performing semantic matching based on the name of the target data and the sub-category name in the preset table to obtain multiple matching results. Marine-related data, such as the national economic industry "communication system equipment manufacturing" (3921), is not included in the current marine and related industry classification standards. According to the marine production activities engaged in, it should belong to the marine information service industry and the marine equipment manufacturing industry;

[0062] S300, setting a characteristic period based on the change of target data, and generating a recommendation coefficient based on the change of the gross marine product of the target data and multiple matching results within the characteristic period;

[0063] S400, selecting several subcategories from multiple matching results as recommended categories of target data according to the recommendation coefficients, the recommended categories being used to provide priority categories for manual verification to the verification personnel;

[0064] S500. Verify and analyze several selected recommended classifications, and use a subclass that passes the verification analysis as the verification classification for the target data. The verification classification is used to skip the manual verification of the current target data by the verification personnel.

[0065] In this embodiment, through two processes of recommendation analysis and verification analysis, the marine industry data that cannot be directly classified according to the preset table is obtained for analysis and processing. In some cases, the subclass corresponding to the marine industry data can be directly determined, and several recommended classifications are given when the subclass cannot be directly given, thus greatly reducing the workload of subsequent verification personnel and improving the verification efficiency.

[0066] Among them, the process of performing semantic matching and then obtaining multiple matching results includes:

[0067] S210. Perform word segmentation on the target text and set corresponding word vectors for each word segment. The target text includes the title text of the target data and the classification title text of the subclass.

[0068] S220. Obtain the semantic vector of the target text based on multiple word vectors.

[0069] Obtain the vector distance between the semantic vector of the target data and the semantic vector of the classification title of the subclass, and match and obtain multiple matching results from the classification titles of the subclasses based on the preset judgment distance. Taking "Communication System Equipment Manufacturing" and "Marine Information Service Industry and Marine Equipment Manufacturing Industry" as examples, the communication system includes at least four word segments: "communication", "system", "equipment", and "manufacturing". Set multi-dimensional corresponding word vectors for each word segment, and then combine multiple corresponding word vectors to obtain a semantic vector. Similarly, in the same dimension, "Marine Information Service Industry and Marine Equipment Manufacturing Industry" can also be obtained by combining multiple corresponding word vectors to obtain a semantic vector. When each corresponding word vector is unique, the closer the semantics of the two are, the smaller the vector distance between the two semantic vectors.

[0070] The process of setting the characteristic time period based on the change of the target data in S230 includes:

[0071] S240. Obtain the gross marine product value of the target data within several monitoring time periods.

[0072] S250. Set a change threshold, obtain the difference between the gross marine product value within one monitoring time period and the gross marine product value within the previous monitoring time period, and compare the numerical part of the difference with the change threshold.

[0073] S260. Set the monitoring time periods in which the numerical parts of the differences in multiple adjacent monitoring time periods exceed the change threshold and have the same difference sign as the characteristic time periods.

[0074] Reference Figure 2, the process of generating a recommendation coefficient based on the changes in the gross ocean product during the characteristic period for the target data and multiple matching results includes:

[0075] S310. Calculate and obtain the average change amount of the target data and multiple matching results in each monitoring period during the characteristic period;

[0076] S320. Set the standard change amounts for the target data and multiple matching results respectively according to the contemporaneous data in the historical data;

[0077] S330. Obtain a reference value based on the ratio of the average change amount to the corresponding standard change amount during the characteristic period for the target data and multiple matching results;

[0078] S340. Obtain the number of differences where the sign of the difference between the gross ocean product in one monitoring period and the gross ocean product in the previous monitoring period during the characteristic period in multiple matching results is different from the sign of the corresponding difference in the target data, and compare this number of differences with the total number of differences generated for the target data during the characteristic period to obtain a coefficient value;

[0079] S350. Calculate and obtain the recommendation coefficient according to the reference value and the coefficient value using an exponential model.

[0080] As an example, the recommendation coefficient is obtained through the formula:

[0081]

[0082] where the recommendation coefficient is , is the average change amount of the target data in the th characteristic period, is the standard change amount preset for the target data in the th characteristic period, is the average change amount in a corresponding th characteristic period in a matching result, is the standard change amount preset for the current matching result in the th characteristic period, is the number of characteristic periods, is a non - zero natural number less than or equal to , is the total number of differences generated for the target data during the characteristic period, is the number of differences where the sign of the difference between the gross ocean product in one monitoring period and the gross ocean product in the previous monitoring period during the characteristic period in a matching result is different from the sign of the corresponding difference in the target data, is a preset correction coefficient, which is a constant; it should be noted that the reference value is , and at the same time the coefficient value is ;

[0083] In this embodiment, recommendation verification is performed by obtaining a recommendation coefficient, which is obtained based on target data with unknown classification. First, multiple relevant subcategories are obtained through semantic matching, and then several relevant subcategories with the closest change status are selected as recommended classifications by comparing the change status of the multiple relevant subcategories with the target data. In this way, a suitable classification can be quickly found among a large number of subcategories, thereby realizing intelligence and making in-depth use of the data.

[0084] Among them, the process of selecting several subcategories as the recommended classification of the target data from multiple matching results includes:

[0085] Set a recommendation threshold , and compare the recommendation coefficient with the recommendation threshold ;

[0086] If , then take the matching result corresponding to the recommendation coefficient as one of the recommended classifications;

[0087] If , then do not recommend the matching result.

[0088] The process of verifying and analyzing the selected several recommended classifications includes:

[0089] If the number of recommended classifications is zero, no verification and analysis are performed, and the output verification classification is none;

[0090] If the number of recommended classifications is not less than 1, obtain the list of source units of the target data, the business data, and the list of business units registered for the corresponding major categories of the recommended classifications. The list of marine-related enterprises provides a basis for determining the proportion of the marine-related industry, lays a foundation for the scientific implementation of marine economic operation monitoring, and provides support for determining the stripping coefficient, that is, the proportion of the marine-related operating income of marine industry enterprises and the extrapolated growth rate of the non-economic census year's marine industry added value;

[0091] Then, obtain a verification coefficient based on the income of the source unit of the target data and the missing rate of the source unit by the registration unit of the corresponding major category;

[0092] Obtain a verification analysis result based on the verification coefficient, and the verification analysis result includes passing the verification and not passing the verification.

[0093] The process of obtaining the verification coefficient includes:

[0094] Obtain the total marine-related income of the source unit of the target data and the income of the industry corresponding to the business target data, obtain the proportion of the income of the industry corresponding to the business target data in the total marine-related income, and calculate the average value of this proportion for all source units of the target data;

[0095] Select several business scopes with the highest duplication rate of the sea-related business scope of the target data source unit as the target scope;

[0096] Select several business units containing the target scope from the business units corresponding to the recommended classification as the target units, and set a standard ratio based on the historical business data of the target units. The standard ratio is the average value of the ratio of the income of the target units operating in the industry corresponding to the recommended classification to the total sea-related income of the target units' operations;

[0097] Set a first weight coefficient for the ratio of the average value of the ratio to the standard ratio, set a second weight value for the missing rate of the enterprises corresponding to the target data in the large category to the source unit, and perform weighted summation to obtain a verification coefficient.

[0098] As an example, through the formula:

[0099]

[0100] Among them, the verification coefficient is , is the first weight coefficient set for the first reference value, which is set based on empirical data, is the second weight coefficient set for the second reference value, which is also set based on empirical data, is the income of the target data source unit operating in the industry corresponding to the target data, is the total sea-related income of the target data source unit, is the number of operations of the target data source unit, is the average value of the ratio of the income of the target unit operating in the industry corresponding to the recommended classification to the total sea-related income of the target unit's operations, is the number of units not covered by the list of business units registered in the large category corresponding to the recommended classification in the target data source unit, is the total number of business units registered in the large category corresponding to the recommended classification.

[0101] In this embodiment, several recommended classifications are further verified through the verification coefficient. The recommended classifications are obtained based on the data change status. Obviously, there will be two small categories with different classifications but similar sea-related gross product change status. The verification coefficient of the present invention is obtained based on the missing rate of the enterprises corresponding to the target data in all enterprises corresponding to the recommended classification and the proportion of the income of the industry corresponding to the target data in the total income of the enterprises. By adding the enterprise's sea-related income and sea-related registration status as verification dimensions, the recommended classification is further verified. Under the condition of reasonable setting of the verification interval, it can accurately find the small category in the recommended classification that conforms to the target data, so as to achieve further verification and analysis.

[0102] The process of obtaining verification analysis results based on verification coefficients includes:

[0103] Based on the number of recommended categories, the validation coefficient and validation interval are Make comparisons;

[0104] If the number of recommended categories is 1, the verification coefficient Not greater than the right endpoint of the verification interval If yes, the verification is passed, otherwise it fails;

[0105] If the number of recommended categories is greater than 1, the verification coefficient Less than the left endpoint of the validation interval If yes, the verification is passed; otherwise, it fails.

[0106] When classifying marine data, different types of marine data have different levels of difficulty in classification. Although some marine data are not included in the marine and related industry comparison table, their appropriate classification in the marine and related industry comparison table can be determined relatively easily based on their names and corresponding business scopes of enterprises. However, some marine data, such as marine scientific research projects, are usually associated with data in multiple marine and related industry comparison tables. In this case, recommendation analysis and verification analysis can be performed to improve statistical speed.

[0107] In this embodiment, data can also be recorded for the recommended classification and the verified classification, and at the same time, the marine industry data collected during the monitoring period can be classified and stored according to the monitoring time period to facilitate manual verification.

[0108] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An intelligent regional summary accounting method for marine industry data, characterized in that: The steps include: During the monitoring period, the marine industry data is collected and sorted according to the monitoring period, and the acquired marine industry data is directly classified according to the preset table, which includes two classification modes: major category and minor category; Obtaining marine industry data that cannot be directly classified through a preset table as target data, and performing semantic matching based on the name of the target data and the sub-category name in the preset table to obtain multiple matching results; Then, a characteristic period is set based on the change of the target data, and a recommendation coefficient is generated based on the change of the gross marine production value of the target data and multiple matching results within the characteristic period; The process of setting characteristic time periods based on changes in target data includes: Obtain the total marine production value of target data within a certain monitoring period; Set a change threshold, obtain the difference between the total marine production value in one monitoring period and the total marine production value in the previous monitoring period, and compare the numerical value of the difference with the change threshold; The monitoring periods in which the numerical values ​​of the differences in multiple adjacent monitoring periods exceed the change threshold and the difference signs are the same are set as characteristic periods; The process of generating the recommended coefficient based on the target data and the changes in the total marine production value within the characteristic period includes: Calculate and obtain the average change of target data and multiple matching results in each monitoring period within the characteristic period; According to the data of the same period in the historical data, standard changes are set for the target data and multiple matching results respectively; Obtaining a benchmark value based on the ratio of the average change amount in the characteristic time period corresponding to the target data and the multiple matching results to the corresponding standard change amount; Obtain the number of differences between the total marine production value in a monitoring period within a characteristic period and the total marine production value in a previous monitoring period in the multiple matching results, the signs of which are different from the signs of the differences corresponding to the target data, and compare the number of differences with the total number of differences generated by setting the characteristic period for the target data to obtain the coefficient value; Calculate and obtain the recommended coefficient based on the benchmark value and the coefficient value; Selecting several subcategories from multiple matching results as recommended categories of target data according to the recommendation coefficients, wherein the recommended categories are used to provide priority categories for manual checking to the checking personnel; Verification analysis is performed on the selected several recommended categories, and a subcategory that passes the verification analysis is used as the verification category of the target data. The verification category is used for the checker to skip the manual check of the current target data.

2. According to claim 1, a smart regional summary accounting method for marine industry data is characterized in that: The process of performing semantic matching and then obtaining multiple matching results includes: Perform word segmentation on the target text and set a corresponding word vector for each word segmentation. The target text includes the title text of the target data and the classification title text of the sub-category. Obtain the semantic vector of the target text based on multiple word vectors; The vector distance between the semantic vector of the target data and the semantic vector of the classification title of the subcategory is obtained, and multiple matching results are obtained from the classification title of the subcategory based on a preset judgment distance.

3. According to claim 1, the intelligent regional summary accounting method for marine industry data is characterized in that: The process of selecting several subcategories from multiple matching results as recommended categories of target data according to the recommendation coefficient includes: Setting a recommendation threshold and comparing the recommendation coefficient with the recommendation threshold; If the recommendation coefficient is not less than the recommendation threshold, the matching result corresponding to the recommendation coefficient is used as one of the recommended categories; If the recommendation coefficient is less than the recommendation threshold, the matching result is not recommended.

4. According to claim 1, the intelligent regional summary accounting method for marine industry data is characterized in that: The process of verifying and analyzing the selected recommended categories includes: If the number of recommended categories is zero, no verification analysis is performed and the output verification category is None; If the number of recommended categories is not less than 1, obtain the list of target data source units and the list of business units registered in the major category corresponding to the business data and the recommended categories; Then, the verification coefficient is obtained based on the income of the target data source unit and the missing rate of the corresponding major registration unit to the source unit; The verification analysis results are obtained based on the verification coefficient, and the verification analysis results include passing the verification and failing the verification.

5. According to claim 4, a smart regional summary accounting method for marine industry data is characterized in that: The process of obtaining the validation coefficient includes: Obtain the total marine-related income of the target data source unit and the income of the industry corresponding to the operating target data, obtain the proportion of the income of the industry corresponding to the operating target data in the total marine-related income, and calculate the average of the proportions of all target data source units; Select several business scopes whose business scopes contain the highest duplication rate of the marine business scopes of the target data source units as the target scopes; Select several business units that include the target range from the business units corresponding to the recommended classification as target units, and set a standard ratio based on the historical operating data of the target units. The standard ratio is the average value of the ratio of the income of the target unit's business corresponding to the industry of the recommended classification to the total marine-related income of the target unit; A first weight coefficient is set for the ratio of the average value of the proportion to the standard proportion, a second weight value is set for the missing rate of the source units of the major registered enterprises corresponding to the target data, and the weighted sum is performed to obtain the verification coefficient.

6. According to claim 5, a smart regional summary accounting method for marine industry data is characterized in that: The process of obtaining verification analysis results based on verification coefficients includes: Based on the number of recommended categories, the validation coefficient is compared with the validation interval; If the number of recommended categories is 1, the verification will be passed if the verification coefficient is not greater than the right endpoint of the verification interval, otherwise it will fail; If the number of recommended categories is greater than 1, the verification passes if the verification coefficient is less than the left endpoint of the verification interval, otherwise it fails.

7. According to claim 1, a smart regional summary accounting method for marine industry data is characterized in that: Also includes: Data records are kept for recommended classifications and verified classifications, and marine industry data collected during the monitoring period are classified and stored according to the monitoring period.

Images