Enterprise industry classification method and device
By extracting deduplication keywords and calculating matching scores in enterprise industry standard classification documents, the classification problem of small and medium-sized enterprises in existing technology is solved, the universality and accuracy of enterprise industry classification is improved, and exogenous databases and artificial intelligence are avoided Limitations of algorithms.
Patent Information
- Application Number
- CN202411075855.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-07
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-08-07
AI Technical Summary
Existing enterprise industry classification technologies cannot be effectively matched in the case of defects in data resources or incomplete exogenous databases, and small enterprises and individual business owners are difficult to cover, and artificial intelligence methods are not applicable to small samples and the algorithm is complex, difficult to predict and over-predict. Fit.
The independent keywords and classification parameters after deduplication are extracted through the enterprise industry standard classification file, the matching score is calculated based on the keywords contained in the enterprise name, the enterprise industry classification is determined, and the universality and accuracy of classification are improved.
It realizes effective industry classification for small enterprises and individual business owners, improves the universality and accuracy of classification, and avoids the complexity of exogenous database dependence and artificial intelligence algorithms.
Smart Images

Figure CN119004177A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to a method and device for enterprise industry classification. Background Art
[0002] There is a widespread demand for enterprise industry classification. Existing enterprise industry classification technologies are mainly implemented by matching enterprise names (or enterprise unified identification codes) with data collected in external databases. This type of existing technology can fully match when the data resources are complete, but there are the following problems when the data resources are defective: (1) The matching is completely limited by the selection and collection of data in the external database. If the external database is defective, the data contains errors or omissions, or the identification codes are not unified, there is no solution; (2) Traditional research focuses on large enterprises, so external databases do not often include small but large micro-companies or individual industrial and commercial households. Existing enterprise industry classification technologies cannot widely cover the latter.
[0003] In addition to the more traditional existing technologies mentioned above, artificial intelligence methods (such as deep learning, clustering, etc.) have also been used in recent years to classify corporate industries. The general problem with this type of method is that it often requires a larger training set and is not applicable to small samples. In addition, the algorithm itself is opaque, difficult to predict, overfitting, and non-standardized. Summary of the invention
[0004] In view of the problems in the prior art, an embodiment of the present invention provides a method and device for enterprise industry classification, which can at least partially solve the problems in the prior art.
[0005] On the one hand, the present invention proposes a method for classifying enterprise industries, comprising:
[0006] Extract the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry based on the enterprise industry standard classification file;
[0007] The enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries;
[0008] Determine the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword;
[0009] The matching scores corresponding to the enterprise to be classified and each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified.
[0010] The extracted independent keywords corresponding to each industry after deduplication according to the enterprise industry standard classification file include:
[0011] Extracting the column where the category name is located from the enterprise industry standard classification file, and performing preliminary extraction on the text content in the column where the category name is located to obtain initial keywords;
[0012] The initial keywords are modified to obtain independent keywords corresponding to each industry after duplication.
[0013] The initial keywords are modified to obtain independent keywords corresponding to each industry after duplication, including:
[0014] Remove words and phrases that are not related to the industry, and among the retained words, correct the mis-split words according to the category name corresponding to the column where the category name is located;
[0015] Correct the wrongly split words with conjunctions, split the corrected words to obtain keywords with no more than a preset number of words, remove duplicates from the keywords with no more than a preset number of words, and obtain independent keywords corresponding to each industry after removal of duplicates.
[0016] The step of determining the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword includes:
[0017] Taking the industry corresponding to at least one independent keyword contained in the enterprise name as the target industry;
[0018] In each target industry, when an independent keyword included in the enterprise name appears, the number of target independent keywords of the corresponding target industry is increased by 1 to obtain the number of target independent keywords of each target industry that matches the enterprise to be classified.
[0019] The enterprise industry classification method also includes:
[0020] The industries corresponding to the independent keywords that are not included in the enterprise name are directly determined as enterprise industry classifications that are unrelated to the enterprise to be classified.
[0021] The matching scores of the enterprise to be classified corresponding to each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of occurrences, including:
[0022] The matching score corresponding to each target industry is calculated according to the following expression:
[0023]
[0024] Wherein, w is the number of target independent keywords, n i is the number of occurrences, N is the total number of keywords, M is the number of keywords, k i is the number of times, i represents the i-th target independent keyword, and I represents the target independent keyword set composed of all target independent keywords.
[0025] Wherein, after the step of determining the target industry with the highest matching score as the enterprise industry classification of the enterprise to be classified, the enterprise industry classification method further includes:
[0026] In response to the audit operation on the enterprise industry classification, if it is determined that the audit operation result is a first audit operation result with incorrect classification, a prompt message is generated to optimize the enterprise industry classification parameters and / or independent keywords.
[0027] In one aspect, the present invention provides an enterprise industry classification device, comprising:
[0028] An acquisition unit is used to extract, according to the enterprise industry standard classification file, independent keywords and enterprise industry classification parameters corresponding to each industry after duplication;
[0029] The enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries;
[0030] A determination unit, configured to determine the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword;
[0031] A classification unit is used to calculate the matching scores corresponding to each target industry of the enterprise to be classified according to the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and determine the target industry with the highest matching score as the enterprise industry classification of the enterprise to be classified.
[0032] In another aspect, an embodiment of the present invention provides an electronic device, including: a processor, a memory, and a bus, wherein:
[0033] The processor and the memory communicate with each other via the bus;
[0034] The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the following method:
[0035] Extract the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry based on the enterprise industry standard classification file;
[0036] The enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries;
[0037] Determine the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword;
[0038] The matching scores corresponding to the enterprise to be classified and each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified.
[0039] An embodiment of the present invention provides a non-transitory computer-readable storage medium, including:
[0040] The non-transitory computer-readable storage medium stores computer instructions, which cause the computer to execute the following method:
[0041] Extract the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry based on the enterprise industry standard classification file;
[0042] The enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries;
[0043] Determine the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword;
[0044] The matching scores corresponding to the enterprise to be classified and each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified.
[0045] The enterprise industry classification method and device provided by the embodiment of the present invention extracts the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry according to the enterprise industry standard classification file; wherein the enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of occurrences of each independent keyword in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of occurrences of each independent keyword in different industries; the number of target independent keywords for each target industry that matches the enterprise to be classified is determined according to whether the enterprise name of the enterprise to be classified contains the independent keyword; the matching score corresponding to each target industry of the enterprise to be classified is calculated according to the target independent keyword number, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified, which can improve the universality of enterprise industry classification. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:
[0047] Figure 1 It is a flowchart of a method for classifying enterprise industries provided in one embodiment of the present invention.
[0048] Figure 2 It is a schematic diagram illustrating part of the contents extracted from an enterprise industry standard classification file provided by an embodiment of the present invention.
[0049] Figure 3 It is a flowchart of an enterprise industry classification method provided by another embodiment of the present invention.
[0050] Figure 4It is a flowchart of an enterprise industry classification method provided by another embodiment of the present invention.
[0051] Figure 5 It is a structural diagram of an enterprise industry classification device provided by an embodiment of the present invention.
[0052] Figure 6 A schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0053] In order to make the purpose, technical scheme and advantages of the embodiments of the present invention more clear, the embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention. It should be noted that, in the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other arbitrarily.
[0054] Explanation of related terms:
[0055] Sliding Window: Sliding window is a data structure and algorithm-related problem-solving technique, applicable to array or list problems. It was originally used for flow control. The sliding window protocol is applied to TCP / IP and is one of the core strategies of TCP to implement flow control, etc. It is used for flow control during network data transmission to avoid congestion. Later, it developed into a sliding window algorithm, which mainly solves different problems by maintaining a window and moving the two boundaries of the window.
[0056] Sliding window resource pool: Borrowing the idea of thread pool, multiple sliding windows are maintained in the sliding window resource pool to allow concurrent execution of tasks.
[0057] Figure 1 FIG. 1 is a flow chart of a method for classifying an enterprise industry according to an embodiment of the present invention. Figure 1 As shown, the enterprise industry classification method provided by the embodiment of the present invention includes:
[0058] Step S1: extracting the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry according to the enterprise industry standard classification file;
[0059] Among them, the enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries.
[0060] Step S2: determining the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword.
[0061] Step S3: Calculate the matching scores of the enterprise to be classified corresponding to each target industry based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and determine the target industry with the highest matching score as the enterprise industry classification of the enterprise to be classified.
[0062] In the above step S1, the device extracts the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry according to the enterprise industry standard classification file;
[0063] Among them, the enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries. The device can be a computer device that executes the method, for example, it can include a server. It should be noted that the acquisition and analysis of data involved in the embodiments of the present invention are authorized by the user. The enterprise industry standard classification file can specifically be a file corresponding to GB / T 4754-2017, such as Figure 2 As shown, a certain keyword may appear multiple times in the row corresponding to the same industry in the "Description" column. The keyword can be deduplicated, and the obtained independent keyword is the keyword retained after deduplication.
[0064] The extracted independent keywords corresponding to each industry after deduplication according to the enterprise industry standard classification file include:
[0065] Extracting the column where the category name is located from the enterprise industry standard classification file, and preliminarily extracting the text content in the column where the category name is located to obtain initial keywords;
[0066] The initial keywords are modified to obtain independent keywords corresponding to each industry after duplication.
[0067] The category name column is the source of industry keywords, such as Figure 2 As shown, the codes, category names and descriptions of various levels of industry classification are listed, such as Figure 2 The industry classification code for "rice planting" is "A0111", where "A" is the category code, the next two digits "01" are the major category code, and the last two digits "1" are the sequential codes for the medium category and minor category respectively.
[0068] Performing a correction process on the initial keywords to obtain de-duplicated independent keywords corresponding to each industry respectively, including:
[0069] Removing words and terms irrelevant to the industry, and correcting mis-split words among the remaining words according to the category names corresponding to the columns where the category names are located;
[0070] Correcting mis-split words with conjunctions, splitting the corrected words to obtain keywords not exceeding a preset number of characters, and de-duplicating the keywords not exceeding the preset number of characters to obtain de-duplicated independent keywords corresponding to each industry respectively.
[0071] Such as Figure 3 As shown, the extraction of independent keywords mainly includes a splitting process and three correction processes, which are described as follows:
[0072] First, the open-source jieba library in Python can be used to perform a rough first-step splitting of the industry category names, and then common conjunctions (such as "and", "and others", "other", "of", etc.) and punctuation marks that appear (such as commas and Chinese and English brackets) are removed to obtain the initial keywords.
[0073] Since the category names are mostly short sentences, there are also many omissions and errors in the splitting by the jieba library. Correct the above initial keywords:
[0074] Screen out relatively obvious useless words and adjectives from all the single characters split out, remove words that are obviously irrelevant to the industry (such as "type", "contain", "in", "west", "similar"), and then combine the category names to correct adjectives and mis-split words into common words (such as combining "light" Figure 2 to correct to "light type", "year" combined with Figure 2 to correct to "annuity", etc.).
[0075] Correct mis-split words with conjunctions. For example, conjunctions such as "and" and "and others" occasionally fail to be separated from other nouns (such as "salt and"), and in this step, correct the mis-split words to normal words (such as "salt").
[0076] Continue to split keywords with four characters and five characters (the longest word after jieba splitting is five characters). Among the keywords with four characters, except for special cases (such as "carbon dioxide"), they are all split into two words according to 2 + 2, and among the keywords with five characters, except for special cases, they are split into two words of 2 + 3 or 3 + 2 according to common language rules.
[0077] Table 1
[0078]
[0079] As shown in Table 1, 10 examples are listed to illustrate the results of keyword splitting of industry category names. Roughly one slightly longer category and keyword splitting result is randomly selected for every 200 data items. Keywords can be extracted from categories, major categories, medium categories, or minor categories. In general, the most common keywords are double-word keywords, covering more representative words in various industries.
[0080] Before industry matching, the keyword list needs to be analyzed and counted.
[0081] first Figure 2 The last column in extracts all the keywords involved in each industry (category), and N is used to represent the total number of all keywords (including repeated occurrences) extracted from each industry (that is, the total number of keywords corresponding to each industry and including repeated occurrences in the enterprise industry standard classification file), and n is used to represent the total number of keywords (including repeated occurrences) extracted from each industry. i To represent the number of times each independent keyword appears (i.e., the number of times each independent keyword appears in the enterprise industry standard classification file), M is used to represent the total number of non-repeated keywords in each industry (i.e., the number of non-repeated keywords in each industry).
[0082] After counting N and n i , M and then analyze the keywords between industries, using k i To represent the number of times a keyword appears in different industries (i.e., the number of times each independent keyword appears in different industries), for example, if a keyword appears in two industries at the same time, k i =2, if a keyword appears in three industries at the same time, then k i =3, and so on. After all statistics are completed, a table containing industry (category code), keywords (independent keywords after deduplication) and enterprise industry classification parameters (N, n i , M, k i ) is used for subsequent processing.
[0083] In the above step S2, the device determines the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword. Figure 4 As shown, the enterprise name of the enterprise to be classified can be pre-processed before determining whether the enterprise name contains independent keywords. Figure 4The left side represents the processing rules for company names, and the right side represents the processing rules for industry matching tables. Here, we take a company "XX Aviation Industry Group Co., Ltd." as an example to explain. First of all, company names generally contain many identical but industry-irrelevant words, such as "limited liability", "group", etc. The method of the present invention includes preprocessing of company names to remove these irrelevant words. The main purpose is to improve the speed of the later matching algorithm and reduce the matching interference that may be caused by irrelevant words.
[0084] For example, after removing irrelevant words from "XXX Aviation Industry Group Co., Ltd.", the remaining part is "Aviation Industry".
[0085] The step of determining the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword includes:
[0086] Taking the industry corresponding to at least one independent keyword contained in the enterprise name as the target industry;
[0087] In each target industry, when an independent keyword included in the enterprise name appears, the number of target independent keywords of the corresponding target industry is increased by 1 to obtain the number of target independent keywords of each target industry that matches the enterprise to be classified.
[0088] The enterprise industry classification method also includes:
[0089] The industries corresponding to the independent keywords that are not included in the enterprise name are directly determined as enterprise industry classifications that are irrelevant to the enterprise to be classified. For enterprise industry classifications that are irrelevant to the enterprise to be classified, there is no need to perform the corresponding industry matching score calculation process steps.
[0090] In the above step S3, the device calculates the matching scores of the enterprise to be classified corresponding to each target industry according to the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and determines the target industry with the highest matching score as the enterprise industry classification of the enterprise to be classified.
[0091] The matching scores of the enterprise to be classified corresponding to each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of occurrences, including:
[0092] The matching score corresponding to each target industry is calculated according to the following expression:
[0093]
[0094] Wherein, w is the number of target independent keywords, n i is the number of occurrences, N is the total number of keywords, M is the number of keywords, k i is the number, i represents the ith target independent keyword, and I represents the target independent keyword set composed of all target independent keywords. Referring to the above description, for the enterprise industry classification irrelevant to the enterprise to be classified, w=0, so the corresponding score=0.
[0095] The internal logic of this formula is to consider the importance of each keyword in the industry, as well as the number of keywords contained in a company name. For example, if a keyword appears frequently in an industry, then the frequency of this keyword It will be relatively high, and the company containing this keyword is more likely to belong to this industry, so it should account for a larger proportion in the score calculation formula.
[0096] For example, if a keyword appears in multiple industries, then the industry representativeness of this keyword is relatively poor, and a relatively small weight should be given in the score statistics.
[0097] Since the number of keywords varies in different industries, generally speaking, the more non-repeated keywords there are, the larger the industry is and the more companies it includes, so M is used as an adjustment item for industry size in the score.
[0098] The more keywords belonging to a certain industry a company name contains, the more likely it is that the company belongs to this industry. The formula uses w to adjust the final matching score.
[0099] After the matching scores of each industry are calculated, an industry matching score table covering all industries can be generated. If the matching scores of all industries are 0, it means that the company name does not contain any industry keywords. In this case, we believe that the industry classification has failed.
[0100] If the industry matching scores are not all 0, then we believe that the industry with the highest score is most likely to be the correct industry classification, and the algorithm will return the category code of this industry, and the industry matching is successful. For example, after the industry matching algorithm calculates the matching score for "XX Aviation Industry Group Co., Ltd.", the matching score table shown in Table 2 will be generated:
[0101] Table 2
[0102] industry Matching score industry Matching score A 0 K 0 B 0 L 0 C 2.806 M 0.3636 D 0 N 0 E 0 O 0 F 0 P 0 G 1.700 Q 0 H 0 R 0 I 0 S 0 J 0 T 0
[0103] Among them, category C (manufacturing) has the highest score, and the algorithm of the present invention will classify the enterprise into industry C.
[0104] The present invention can quickly classify the industries of enterprises in the sample. From the results of industry classification, most of the matching failures are because the enterprise name itself does not contain industry information, such as "XXX (Group) Co., Ltd.", "XXXXXX Group Holdings Co., Ltd.", etc.
[0105] In addition, we manually checked the examples of successful matches in the samples, and the matching accuracy was around 70%. The matching errors were mainly affected by two factors:
[0106] (1) High-frequency words in some industry keywords are not common in company names. For example, the word "retail" is often used in the industry name of the retail industry, but actual retail companies often use words such as "supermarket" and "department store";
[0107] (2) Industry keywords are mainly double-word keywords, and there is overlap between keywords in related industries. Therefore, when a single-word combination appears in the company name, matching errors are likely to occur. For example, in "XX Power Grid", both "electricity" and "network" contain industry information, but in the keyword table, "electricity" basically appears as a double-word combination such as "power generation". In the above cases, the matching accuracy of the algorithm can be improved by further modifying the company's industry classification parameters and / or keyword table.
[0108] The enterprise industry classification method provided by the embodiment of the present invention can be used for the rapid classification of samples of any size, and is particularly suitable for large sample classification research that is difficult to process manually and for situations where the sample enterprises lack effective external database matching information. The more industry-related information contained in the enterprise name, the greater the accuracy of the classification. For industries with obvious differences in nature, such as manufacturing and finance, the accuracy of the algorithm classification is very high. For industries with common properties, such as manufacturing and transportation, the accuracy of the algorithm classification is relatively low. Therefore, the method of the present invention is more suitable for distinguishing enterprises with obvious differences in industry nature.
[0109] The enterprise industry classification method provided by the embodiment of the present invention extracts the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry according to the enterprise industry standard classification file; wherein the enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries; the number of target independent keywords for each target industry that matches the enterprise to be classified is determined according to whether the enterprise name of the enterprise to be classified contains the independent keyword; the matching score corresponding to each target industry of the enterprise to be classified is calculated according to the target independent keyword number, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified, which can improve the universality of enterprise industry classification.
[0110] Furthermore, the extracted independent keywords corresponding to each industry after deduplication according to the enterprise industry standard classification file include:
[0111] The column where the category name is located is extracted from the enterprise industry standard classification file, and the text content in the column where the category name is located is preliminarily extracted to obtain initial keywords; this can be described with reference to the above embodiment and will not be repeated here.
[0112] The initial keywords are modified to obtain independent keywords corresponding to each industry after duplication.
[0113] Furthermore, the initial keywords are modified to obtain independent keywords corresponding to each industry after duplication, including:
[0114] Remove characters and words that are not related to the industry, and correct the wrongly split words among the retained words according to the category name corresponding to the column where the category name is located; refer to the above embodiment for description and no further details will be given.
[0115] Correct the wrongly split words with connecting words, split the corrected words, obtain keywords with no more than a preset number of words, remove duplicates from the keywords with no more than a preset number of words, and obtain independent keywords corresponding to each industry after removing duplicates. The above-mentioned embodiment can be referred to for explanation, and no further description is given.
[0116] Furthermore, the step of determining the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword includes:
[0117] The industry corresponding to at least one independent keyword contained in the enterprise name is taken as the target industry; the description can be made with reference to the above embodiment and will not be repeated here.
[0118] In each target industry, when an independent keyword included in the enterprise name appears, the number of target independent keywords of the corresponding target industry is increased by 1, so as to obtain the number of target independent keywords of each target industry matching the enterprise to be classified.
[0119] Furthermore, the enterprise industry classification method also includes:
[0120] The industries corresponding to the independent keywords that are not included in the enterprise name are directly determined as enterprise industry classifications that are irrelevant to the enterprise to be classified. The above-mentioned embodiments can be referred to for explanation and will not be described in detail.
[0121] Furthermore, the matching scores of the enterprise to be classified corresponding to each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of occurrences, including:
[0122] The matching score corresponding to each target industry is calculated according to the following expression:
[0123]
[0124] Wherein, w is the number of target independent keywords, n i is the number of occurrences, N is the total number of keywords, M is the number of keywords, k i is the number of times, i represents the i-th target independent keyword, and I represents the target independent keyword set composed of all target independent keywords. The above embodiments can be referred to for explanation, and no further details are given.
[0125] Furthermore, after the step of determining the target industry with the highest matching score as the enterprise industry classification of the enterprise to be classified, the enterprise industry classification method further includes:
[0126] In response to the audit operation on the enterprise industry classification, if it is determined that the audit operation result is a first audit operation result with incorrect classification, a prompt message for optimizing the enterprise industry classification parameters and / or independent keywords is generated.
[0127] It should be noted that the enterprise industry classification method provided in the embodiment of the present invention can be used in the financial field, and can also be used in any technical field other than the financial field. The embodiment of the present invention does not limit the application field of the enterprise industry classification method.
[0128] Figure 5 is a schematic diagram of the structure of an enterprise industry classification device provided by an embodiment of the present invention, such as Figure 5 As shown, the enterprise industry classification device provided by the embodiment of the present invention includes an acquisition unit 501, a determination unit 502 and a classification unit 503, wherein:
[0129] The acquisition unit 501 is used to extract the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry according to the enterprise industry standard classification file; wherein the enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries; the determination unit 502 is used to determine the target number of independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name of the enterprise to be classified contains the independent keywords; the classification unit 503 is used to calculate the matching score of the enterprise to be classified corresponding to each target industry according to the target number of independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and determine the target industry with the highest matching score as the enterprise industry classification of the enterprise to be classified.
[0130] Specifically, the acquisition unit 501 in the device is used to extract the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry according to the enterprise industry standard classification file; wherein the enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries; the determination unit 502 is used to determine the target number of independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name of the enterprise to be classified contains the independent keywords; the classification unit 503 is used to calculate the matching score of the enterprise to be classified corresponding to each target industry according to the target number of independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and determine the target industry with the highest matching score as the enterprise industry classification of the enterprise to be classified.
[0131] The enterprise industry classification device provided by the embodiment of the present invention extracts the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry according to the enterprise industry standard classification file; wherein the enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries; the number of target independent keywords for each target industry that matches the enterprise to be classified is determined according to whether the enterprise name of the enterprise to be classified contains the independent keyword; the matching score corresponding to each target industry of the enterprise to be classified is calculated according to the target independent keyword number, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified, which can improve the universality of enterprise industry classification.
[0132] The embodiment of the present invention provides an embodiment of an enterprise industry classification device which can be specifically used to execute the processing flow of the above-mentioned method embodiments. Its functions are not described in detail here, and reference can be made to the detailed description of the above-mentioned method embodiments.
[0133] Figure 6 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as Figure 6 As shown, the electronic device includes: a processor (processor) 601, a memory (memory) 602 and a bus 603;
[0134] The processor 601 and the memory 602 communicate with each other via a bus 603;
[0135] The processor 601 is used to call the program instructions in the memory 602 to execute the methods provided by the above method embodiments, for example, including:
[0136] Extract the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry based on the enterprise industry standard classification file;
[0137] The enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries;
[0138] Determine the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword;
[0139] The matching scores corresponding to the enterprise to be classified and each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified.
[0140] This embodiment discloses a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can perform the methods provided by the above method embodiments, for example, including:
[0141] Extract the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry based on the enterprise industry standard classification file;
[0142] The enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries;
[0143] Determine the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword;
[0144] The matching scores corresponding to the enterprise to be classified and each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified.
[0145] This embodiment provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein the computer program enables the computer to execute the methods provided by the above method embodiments, for example, including:
[0146] Extract the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry based on the enterprise industry standard classification file;
[0147] The enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries;
[0148] Determine the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword;
[0149] The matching scores corresponding to the enterprise to be classified and each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified.
[0150] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, 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 disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0151] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of 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.
[0152] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate 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 A function specified in one or more boxes.
[0153] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0154] In the description of this specification, the description with reference to the terms "one embodiment", "a specific embodiment", "some embodiments", "for example", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.
[0155] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A method for classifying enterprise industries, characterized in that: include: Extract the deduplicated independent keywords and enterprise industry classification parameters corresponding to each industry based on the enterprise industry standard classification file; The enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries; Determine the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword; The matching scores corresponding to the enterprise to be classified and each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and the target industry with the highest matching score is determined as the enterprise industry classification of the enterprise to be classified.
2. The enterprise industry classification method according to claim 1 is characterized in that: The extracted independent keywords corresponding to each industry after deduplication according to the enterprise industry standard classification file include: Extracting the column where the category name is located from the enterprise industry standard classification file, and preliminarily extracting the text content in the column where the category name is located to obtain initial keywords; The initial keywords are modified to obtain independent keywords corresponding to each industry after duplication.
3. The enterprise industry classification method according to claim 2 is characterized in that: The initial keywords are modified to obtain independent keywords corresponding to each industry after duplication, including: Remove words and phrases that are not related to the industry, and among the retained words, correct the mis-split words according to the category name corresponding to the column where the category name is located; Correct the wrongly split words with conjunctions, split the corrected words to obtain keywords with no more than a preset number of words, remove duplicates from the keywords with no more than a preset number of words, and obtain independent keywords corresponding to each industry after removal of duplicates.
4. The enterprise industry classification method according to claim 1 is characterized in that: The step of determining the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword includes: Taking the industry corresponding to at least one independent keyword contained in the enterprise name as the target industry; In each target industry, when an independent keyword included in the enterprise name appears, the number of target independent keywords of the corresponding target industry is increased by 1 to obtain the number of target independent keywords of each target industry that matches the enterprise to be classified.
5. The enterprise industry classification method according to claim 4 is characterized in that: The enterprise industry classification method also includes: The industries corresponding to the independent keywords that are not included in the enterprise name are directly determined as enterprise industry classifications that are unrelated to the enterprise to be classified.
6. The enterprise industry classification method according to claim 1 is characterized in that: The matching scores of the enterprise to be classified corresponding to each target industry are calculated based on the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of occurrences, including: The matching score corresponding to each target industry is calculated according to the following expression: Wherein, w is the number of target independent keywords, n i is the number of occurrences, N is the total number of keywords, M is the number of keywords, k i is the number of times, i represents the i-th target independent keyword, and I represents the target independent keyword set composed of all target independent keywords.
7. The enterprise industry classification method according to any one of claims 1 to 6, characterized in that: After the step of determining the target industry with the highest matching score as the enterprise industry classification of the enterprise to be classified, the enterprise industry classification method further includes: In response to the audit operation on the enterprise industry classification, if it is determined that the audit operation result is a first audit operation result with incorrect classification, a prompt message is generated to optimize the enterprise industry classification parameters and / or independent keywords.
8. An enterprise industry classification device, characterized in that: include: An acquisition unit is used to extract, according to the enterprise industry standard classification file, independent keywords and enterprise industry classification parameters corresponding to each industry after duplication; The enterprise industry classification parameters include the total number of keywords corresponding to each industry that appear repeatedly in the enterprise industry standard classification file, the number of times each independent keyword appears in the enterprise industry standard classification file, the number of non-repeated keywords in each industry, and the number of times each independent keyword appears in different industries; A determination unit, configured to determine the number of target independent keywords for each target industry that matches the enterprise to be classified according to whether the enterprise name to be classified contains the independent keyword; A classification unit is used to calculate the matching scores corresponding to each target industry of the enterprise to be classified according to the number of target independent keywords, the number of occurrences, the total number of keywords, the number of keywords and the number of times, and determine the target industry with the highest matching score as the enterprise industry classification of the enterprise to be classified.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.
Citation Information
Patent Citations
Chinese enterprise name entity accurate identification secondary matching method
CN111597304A
Industrial category determination method and device, storage medium and electronic equipment
CN114297347A
Enterprise industry classification method and device, storage medium and electronic equipment
CN114637842A
Enterprise industry classification method, system and equipment based on big data and medium
CN116933130A
Enterprise industry classification method and device and storage medium
CN118035437A
