Industry chain and key enterprise construction method and analysis system based on tax data

Abstract

The method for constructing industry chain and key enterprise based on tax data comprises the following steps: collecting the invoicing amount of each enterprise in a selected area to determine the core industry; constructing a recommended main commodity list of each industry and a recommended main material list of each industry, and constructing the industry chain and the key enterprise according to the recommended main commodity list and the recommended main material list of the core industry. The application can quickly and accurately establish the industry chain view of the key industry based on the enterprise value-added tax invoice data, and can deeply analyze the evolution process of the industry chain after superimposing the time dimension, thereby helping to analyze the overall development of the industry at the macro and mesoscopic levels, and greatly enabling the analysis personnel to be far away from the complicated data processing, processing and modeling process. Based on the constructed industry chain, the analysis and decision-making personnel can deeply understand the dynamic development of the key industry, thereby assisting in establishing a long-term tracking mechanism for the development of the industry.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis and mining, in particular to an industry chain and key enterprise construction method and analysis system based on tax data. BACKGROUND

[0002] Tax economic analysis plays an important role in serving national macroeconomic decision-making and social management. Deep mining and full use of tax data information can make tax economic analysis play a greater role in national governance. Tax big data, especially value-added tax invoice data, has the advantages of wide coverage and strong timeliness. Through detailed analysis of the problems to be solved in the industry chain, upstream and downstream, and deepening the analysis of the industry chain, upstream and downstream, and economic dynamics, it can serve decision-making from a macro perspective, serve localities from a medium-term perspective, and serve enterprises from a micro perspective.

[0003] The construction of the industry chain based on tax data belongs to the medium-term economic analysis in tax economic analysis, mainly studies the relationship between industries and constructs the upstream and downstream structure view of the industry chain. The industry chain is a concept in industrial economics, which is the chain-like correlation form objectively formed between various industrial departments based on certain technical and economic correlation and according to specific logical relationship and spatial layout relationship. The industry chain is an industrial economics concept that includes value chain, enterprise chain, supply and demand chain, and spatial chain. Industry chain analysis involves complex relationships between multiple industries and multiple levels, and traditional tax query and statistical analysis methods can only display query results for one layer of enterprise upstream and downstream relationship, and cannot support flexible and comprehensive in-depth analysis of the industry chain.

[0004] The invoice data has the characteristics of using the simple data structure of the seller and the buyer to express complex transaction relationships, which is naturally suitable for constructing data community networks and conducting chain data community network analysis. Therefore, the present application provides an industry chain construction method and analysis system based on tax data. Through community network analysis and other technologies, based on invoice data analysis, from a global macro perspective, the present application can comprehensively grasp the current situation, development trend and operation trajectory of the regional industry chain. According to the actual business situation of the enterprise, the industry to which the enterprise belongs is dynamically determined. From the invoice data of an industry for many years, the product issuing rules of the industry are refined. Then, the rules are used to determine whether the enterprise belongs to the industry, so as to correct and real-time correct the industry to which the enterprise belongs. Further, the effectiveness and reliability of the industry chain analysis results are ensured. SUMMARY

[0005] In order to solve the problems in the prior art, the present application provides an industry chain construction method and analysis system based on tax data, which can construct effective industry chains and provide multi-dimensional industry chain analysis display based on invoice data analysis and community network discovery.

[0006] The present invention adopts the following technical solution:

[0007] The method for constructing an industry chain and key enterprises based on tax data includes the following steps:

[0008] Step 1: Collect the invoice amount for each enterprise in the selected region, and select and determine the core industry based on the invoice amount;

[0009] Step 2: Select the recommended main products and recommended main materials for each industry, and construct a list of recommended main products and a list of recommended main materials for each industry;

[0010] Step 3: Construct the industry chain based on the recommended main product list and recommended main material list of the core industries;

[0011] Step 4: Determine the list of companies in each industry chain;

[0012] Step 5: Identify the key companies in the company list;

[0013] Step 6: Showcase the industry chain and key enterprises.

[0014] In step 1, enterprises are classified according to their registered industries, and the total invoicing amount of all enterprises in each industry is calculated and ranked according to the total invoicing amount; the industries ranked before the core industry ranking threshold are selected as core industries.

[0015] In step 2, the total invoice amount for each item is calculated according to the product code on each company's sales invoice, as well as the percentage of the invoice amount for that item in its industry. For each industry, items with an invoice amount percentage greater than the recommended main product screening threshold are selected as the recommended main products for that industry.

[0016] The total amount of the input invoice for each item is calculated based on the commodity code on the input invoice of each enterprise, as well as the proportion of the total amount of the input invoice for that item in the industry. For each industry, items whose proportion of the total amount of the input invoice is greater than the screening threshold for recommended main materials are selected as the recommended main materials for that industry.

[0017] In step 3, identify the downstream and upstream industries of each core industry, and refer to each core industry and its downstream and upstream industries as an industrial chain;

[0018] Upstream industries refer to industries that provide raw materials to this industry, while downstream industries refer to industries that sell products to this industry.

[0019] Let H be a core industry. The method for determining downstream industries is as follows:

[0020] Step 301.1: Set the maximum number of downstream industry levels and initialize the current number of downstream industry levels. =1;

[0021] In the present embodiment, the maximum number of downstream industries is 3; one skilled in the art can set it according to the actual situation;

[0022] Step 301.2: Find all the sales invoices of the enterprises registered in the core industry H, and aggregate the invoice amounts according to the commodity codes and the registered industries of the purchasing enterprises; all the industries involved in the purchasing enterprises are the candidate downstream industries of the core industry H. For each candidate downstream industry, calculate the total invoice amount of the enterprises in the core industry H to the enterprises in the candidate downstream industry, and if the amount is greater than a given threshold , go to step 301.3, otherwise, determine that the candidate downstream industry is not a downstream industry of the core industry H. Select the larger value between 500 million yuan and 3% of the total amount of invoices issued by all enterprises in the core industry H;

[0023] Step 301.3: If the recommended main materials of the candidate downstream industry do not completely coincide with the recommended main commodities of the core industry H, determine that the candidate downstream industry is not a downstream industry of the core industry H, otherwise go to step 301.4;

[0024] Step 301.4: Find the set of recommended main commodities of the core industry H that coincide with the recommended main materials of the candidate downstream industry, and determine whether each recommended main material in the set has a total invoice amount in the candidate downstream industry greater than the downstream industry main material threshold ; if there is no recommended main material in the set greater than the downstream industry main material threshold , consider that the candidate downstream industry is not a downstream industry of the core industry H; otherwise, add the candidate downstream industry to the set of downstream industries of the first level, and go to step 301.5; the downstream industry main material threshold is set to the larger value between 10 million yuan and 5% of the total amount of invoices issued by all enterprises in the core industry H to the enterprises in the candidate downstream industry;

[0025] Step 301.5: Update the current downstream industry level ; if is greater than the maximum level of the downstream industry, merge all levels of the downstream industry and output, and end the extraction of the downstream industry; otherwise, return to step 301.2 for the next iteration, and take each downstream industry in the set of downstream industries of the first level as the core industry H of the next iteration.

[0026] Let H be a core industry, and the method for determining the upstream industry is as follows:

[0027] Step 302.1: Set the maximum number of upstream industries, and initialize the current upstream industry level is 1;

[0028] Step 302.2: Find all the input invoices of the enterprises registered in the industry as the core industry H, and aggregate the invoice amounts according to the commodity codes of the input invoice traders and the registered industries of the selling enterprises; all the industries involved in the selling enterprises are the candidate upstream industries of the core industry H. For each candidate upstream industry, calculate the total invoice amount received by the enterprises in the core industry H from the enterprises in the candidate downstream industry, and if the amount is greater than a given threshold , go to step 302.3, otherwise, determine that the candidate downstream industry is not an upstream industry of the core industry H; is the greater value between 50 million yuan and 3% of the total amount of invoices received by all enterprises in the core industry H;

[0029] Step 302.3: If there is no recommended main commodity of the candidate upstream industry in the recommended main materials of the core industry H, determine that the candidate upstream industry is not an upstream industry of the core industry H, otherwise go to step 302.4;

[0030] Step 302.4: Find the same recommended main commodity set of the recommended main commodities of the candidate upstream industry and the recommended main materials of the core industry H, and respectively determine whether each recommended main commodity in the set has a total input invoice amount greater than the upstream industry main commodity threshold in the candidate upstream industry; if there is no recommended main commodity in the same main commodity set that is greater than the upstream industry main commodity threshold , consider that the candidate upstream industry does not belong to the upstream industry of the core industry H; otherwise, put the recommended main commodities in the same main commodity set that are greater than the upstream industry main commodity threshold into the associated commodity set, and add the candidate upstream industry to the first level upstream industry set, and go to step 302.5; the upstream industry main material threshold is set to the greater value between 10 million yuan and 5% of the total amount of invoices received by all enterprises in the core industry H from the enterprises in the candidate upstream industry;

[0031] Step 302.5: Update the current upstream industry level ; if is greater than the maximum level of the downstream industry, merge all levels of upstream industries and output, and end the extraction of upstream industries; otherwise, return to step 302.2 for the next iteration, and take each upstream industry in the first level upstream industry set as the core industry H of the next iteration.

[0032] Step 4 includes the following:

[0033] Step 401: aggregate the recommended main commodities of each industry in each industrial chain to obtain a recommended main commodity list of the industry;

[0034] Step 402: for each enterprise in the selected region, if the enterprise's sales invoice contains the commodities in the recommended main commodity list of the industry, and the total sales amount of the commodities in the recommended main commodity list of the industry on the enterprise's sales invoice divided by the total sales amount of the enterprise's sales invoice is greater than the industrial chain determination threshold, the enterprise is determined to be an enterprise within the industrial chain; the industrial chain determination threshold is preferably 0.8, and those skilled in the art can select it according to the actual situation;

[0035] Step 403: for the enterprises within the industrial chain, rejudge the industries to which the enterprises belong according to the commodity amount of the enterprise's sales invoice; if the judged industry of the enterprise is in the industrial chain, the enterprise is retained; if the judged industry of the enterprise is not in the industrial chain, the enterprise is removed from the industrial chain;

[0036] Step 404: extract the enterprises in each industrial chain, and constitute an enterprise list of each industrial chain according to the industrial chain junction.

[0037] The specific method for judging whether the industry to which the enterprise belongs is in the industrial chain in step 403 is: if the registered industry of a certain enterprise in the current industrial chain originally belongs to the industry in the industrial chain, it is judged whether the ratio of the total sales amount of the commodities in the recommended main commodity list of the industry existing on the enterprise's sales invoice to the total sales amount of the enterprise is greater than the first industry determination threshold If it is greater, the enterprise is determined to be an enterprise within the industrial chain, and the industry of the enterprise is consistent with the registered industry; otherwise, the enterprise is removed from the industrial chain;

[0038] If the registered industry of a certain enterprise in the current industrial chain originally does not belong to the industry in the industrial chain, it is judged whether the sales invoice of the enterprise contains the commodities in the recommended main commodity list of any industry in the industrial chain, and if so, it is calculated in turn whether the ratio of the total sales amount of the commodities in the recommended main commodity list of the enterprise in these industries to the total sales amount of the enterprise is greater than the second industry determination threshold If it is greater, the enterprise is attributed to the corresponding industry; if there are no commodities in the recommended main commodity list of any industry in the industrial chain or the calculated ratios are all not greater than the second industry determination threshold , the enterprise is removed from the industrial chain.

[0039] Step 5 includes the following steps:

[0040] Step 501: for the enterprise list of each industrial chain obtained in step 4, an enterprise transaction network diagram is established;

[0041] Wherein, the network node is an enterprise, the edge connecting the nodes represents the invoice transaction relationship between enterprises, the transaction amount and transaction commodity are taken as the attributes of the edge, and the direction of the edge is consistent with the transaction direction, that is, from the commodity seller to the commodity purchaser. If there are multiple transactions between two enterprises, the transactions are first aggregated according to the transaction products, and an edge is established between the enterprises for each commodity;

[0042] Step 502: filtering the nodes and edges in the enterprise transaction network graph in step 501 to generate an industry chain main transaction network graph;

[0043] Step 503: merging the edges in the industry chain main transaction network graph obtained in step 502;

[0044] Step 504: calculating the centrality indexes of each node in the industry chain main transaction network graph, including node degree centrality, node total import and export amount, node eigenvector centrality and node betweenness centrality;

[0045] Step 504: calculating each centrality of each enterprise in the network and ranking them in the network respectively. Each enterprise can obtain four ranking data, and the sum of the four ranking data can obtain the total centrality ranking of the enterprise. The top N enterprises in the total centrality ranking are the key enterprises in the network, wherein, .

[0046] In step 504,

[0047] The node degree centrality is defined as the number of edges connected to the node;

[0048] The node total import and export amount is defined as the total amount of the edges connected to the node;

[0049] The node eigenvector centrality is defined as:

[0050] For each industry chain main transaction network graph , V represents the set of all nodes in the industry chain main transaction network graph, E represents the set of all edges in the industry chain main transaction network graph; define its adjacency matrix as When the node v is connected to the node t , otherwise , d represents the total number of all nodes;

[0051] The node v eigenvector centrality satisfies the following relationship:

[0052] Wherein M(v) is the set of nodes connected to the node v, is the eigenvalue of the adjacency matrix A; The node betweenness centrality of node t connected to node v is represented.

[0053] The betweenness centrality of nodes satisfies the following relationship:

[0054] in, This represents the number of shortest paths from node j through node v to node k.

[0055] In step 6, the industrial chain and key enterprises are showcased, including enterprise chain display, supply and demand and product chain display, value chain display and spatial chain display;

[0056] The enterprise chain showcases the composition and distribution of key enterprises in each industry chain;

[0057] The supply and demand and product chain show the correspondence between the main commodities and main materials of the upstream and downstream industries in each industrial chain, which is used to show the transaction links between the upstream and downstream industries in the industrial chain;

[0058] The value chain shows the total transaction amount of each industry in each industry chain, as well as the comparison of the transaction amount of main commodities and main materials that overlap between upstream and downstream industries;

[0059] The spatial chain shows the geographical relationship between the enterprises corresponding to the purchase and sales invoices issued by each key enterprise in each industrial chain and the key enterprise itself.

[0060] This invention also discloses an analysis system for constructing the industrial chain and key enterprises based on tax data, including an indicator model management module, an industry overview module, an industrial chain view module, an industry view module, and a key enterprise view module;

[0061] The indicator model management module is responsible for collecting invoicing data from each enterprise in the selected region and constructing the industry chain based on this data. This module includes a data management submodule, an initialization submodule, an industry chain maintenance submodule, and an industry chain commodity category library submodule.

[0062] The Industry Overview module can analyze and present the selected industry. This module can display the upstream and downstream industry structure view of any specified industry chain, the recommended main product list of the industry, and the information of key enterprises in the industry chain.

[0063] The industry chain view module can display enterprise chain, supply and demand chain, value chain, and spatial chain views; the supply and demand and product chains show the correspondence between the main commodities and main materials of the upstream and downstream industries in each industry chain; the value chain shows the total transaction amount of each industry in each industry chain; the spatial chain display uses a vector map to show the geographical relationship between the enterprises corresponding to the purchase and sales invoices issued by each key enterprise in each industry chain and the key enterprise itself.

[0064] The industry view module includes a sub-module for enterprise distribution, a sub-module for main products and main materials in the industry, and a sub-module for industry scale. The enterprise distribution sub-module can analyze the distribution and increase / decrease of enterprises of different sizes within the industry, the industry changes of enterprises, and the proportion and growth of enterprises within the industry. The main products and main materials sub-module can analyze the composition and changes of main products and main materials in a single industry. The industry scale sub-module can view the overall development of the industry from the perspective of the overall industry scale, and analyze the development of enterprises within the industry and the development of enterprises within the industry from different sales scale ranges.

[0065] The key enterprise view module can provide an overall analysis and display of key enterprises in each industry chain, including basic enterprise information, enterprise purchase and sales information, and information on other enterprises in the upstream and downstream industries of that enterprise.

[0066] In the indicator model management module, the data management submodule collects the invoicing data of each enterprise in the selected region;

[0067] The initialization submodule determines the core industry, the list of recommended main products for the industry, and the list of recommended main materials for the industry based on the collected invoice data amount, and identifies the upstream and downstream industries of each core industry to build the industrial chain;

[0068] The supply chain maintenance submodule is used to add and delete supply chains, query and maintain supply chains containing industries, and the supply chains to which industries belong.

[0069] The product category library submodule of the industrial chain provides the ability to query recommended main products and main materials in the industry, as well as calculate invoice amounts.

[0070] The beneficial effects of this invention are that, compared with existing technologies, the method and analysis system for constructing and analyzing the industrial chain and key enterprises based on tax data can quickly, accurately, and dynamically establish a view of the industrial chain of key industries based on enterprise value-added tax invoice data. By overlaying the time dimension, it can deeply analyze the evolution of the industrial chain, thereby helping to analyze the overall development of the industry at both macro and meso levels. This largely frees analysts from the tedious data processing, handling, and modeling processes. Based on the constructed industrial chain, it is possible to analyze the rationality of the industrial structure, the advantages of industrial development resources and industrial layout, the exploration of industrial upgrading paths, and issues related to regional industrial cooperation. This helps analysts and decision-makers gain a deeper understanding of the dynamic development of key industries, thereby assisting in the establishment of a long-term tracking mechanism for industrial development. Attached Figure Description

[0071] Figure 1 This is a flowchart of the method for constructing the industrial chain and key enterprises based on tax data according to the present invention;

[0072] Figure 2This is a module diagram of the industry chain and key enterprise analysis system of the present invention. Detailed Implementation

[0073] The present application will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, and should not be construed as limiting the scope of protection of the present application.

[0074] This invention proposes a method for constructing an industry chain and key enterprises based on tax data, the process of which is as follows: Figure 1 As shown, it includes the following steps:

[0075] Step 1: Collect the invoice amount for each enterprise in the selected region, and select and determine the core industry based on the invoice amount;

[0076] After classifying enterprises by their registered industry, the total invoicing amount of all enterprises in each industry is calculated, and the industries are ranked according to the total invoicing amount; the industries ranked above the core industry ranking threshold are selected as core industries;

[0077] In this invention, a region includes more than one industry, an industry includes more than one sub-industry, and a sub-industry includes more than one enterprise;

[0078] In this embodiment, the core industry ranking threshold is 11, so the top 10 industries are considered core industries.

[0079] Step 2: Select the recommended main products and recommended main materials for each industry, and construct a list of recommended main products and a list of recommended main materials for each industry;

[0080] The total invoice amount for each item is calculated based on the product code on each company's sales invoice, as well as the percentage of the invoice amount for that item within its industry. For each industry, items with an invoice amount percentage greater than 5% are selected as the recommended main products for that industry.

[0081] Similarly, based on the commodity code on each enterprise's input invoice, the total amount of the input invoice for each item is calculated, as well as the percentage of the total input invoice amount for that item in the industry. For each industry, items with a percentage of the total input invoice amount greater than 5% are selected as the recommended main materials for that industry.

[0082] Step 3: Construct the industry chain based on the recommended main product list and recommended main material list of the core industries;

[0083] Identify the downstream and upstream industries of each core industry, and refer to each core industry and its downstream and upstream industries as an industrial chain.

[0084] Upstream and downstream industries are relative concepts. Upstream industries refer to industries that provide raw materials to the upstream industry, while downstream industries refer to industries that sell products to the downstream industry.

[0085] Let H be a core industry. The method for extracting the downstream industries of core industry H is as follows:

[0086] Step 301.1: Set the maximum number of downstream industry levels and initialize the current number of downstream industry levels. =1;

[0087] In this embodiment, the maximum number of downstream industries is 3; those skilled in the art can set it according to the actual situation.

[0088] Step 301.2: Identify all sales invoices from companies registered under the core industry H, and summarize the invoice amounts based on the commodity codes on the sales invoices and the industries registered by the purchasing companies. All industries involved in the purchasing companies constitute the candidate downstream industries of core industry H. For each candidate downstream industry, calculate the total invoice amount issued by companies in core industry H to companies in that candidate downstream industry. If this amount exceeds a given threshold... If so, proceed to step 301.3; otherwise, it is determined that the candidate downstream industry is not a downstream industry of the core industry H. The larger of 50 million yuan and 3% of the total amount of invoices issued by all enterprises in core industry H is selected;

[0089] Step 301.3: If the recommended main material of the candidate downstream industry does not overlap with the recommended main product of the core industry H, then the candidate downstream industry is determined not to be a downstream industry of the core industry H; otherwise, proceed to step 301.4.

[0090] Step 301.4: Identify the set of recommended main products in core industry H that overlap with the recommended main materials in the candidate downstream industry, and determine whether the total invoiced amount of each recommended main material in the set in the candidate downstream industry exceeds the threshold for main materials in the downstream industry. If the set does not contain any materials exceeding the threshold of the downstream industry's main materials... If the recommended main material is not found, then the candidate downstream industry is considered not to be a downstream industry of core industry H; otherwise, the candidate downstream industry is added to the list of recommended main materials. After identifying downstream industries, proceed to step 301.5; threshold for main materials in downstream industries. Set to the larger of 10 million yuan and 5% of the total amount offered by all companies in core industry H to companies in the candidate downstream industry;

[0091] Step 301.5: Update the current downstream industry level ;if If the number of downstream industries exceeds the maximum level of downstream industries, then all downstream industries at all levels are merged and output, and the extraction of downstream industries ends; otherwise, return to step 301.2 for the next iteration, and the number of downstream industries is... Each downstream industry in the downstream industry set is designated as the core industry H for the next iteration.

[0092] The method for extracting the upstream industries of core industry H is as follows:

[0093] Step 302.1: Set the maximum number of upstream industry levels and initialize the current number of upstream industry levels. =1;

[0094] In this embodiment, the maximum number of upstream industries is 3; those skilled in the art can set it according to the actual situation.

[0095] Step 302.2: Identify all input invoices from companies registered in the core industry H, and summarize the invoice amounts based on the commodity code of the invoice vendor and the industry registered by the seller company; all industries involved in the seller companies are candidate upstream industries of the core industry H. For each candidate upstream industry, calculate the total invoice amount received by companies in the core industry H from companies in that candidate downstream industry. If this amount exceeds a given threshold... If so, proceed to step 302.3; otherwise, it is determined that the candidate downstream industry is not an upstream industry of the core industry H. The larger of 50 million yuan and 3% of the total amount of invoices received by all enterprises in core industry H is selected;

[0096] Step 302.3: If the recommended main materials of the core industry H do not include the recommended main products of the candidate upstream industry, then the candidate upstream industry is deemed not to be the upstream industry of the core industry H; otherwise, proceed to step 302.4.

[0097] Step 302.4: Identify the set of identical recommended main products for the core industry H's recommended main material and the candidate upstream industry's recommended main product, and determine whether the total input invoice amount for each recommended main product in the set is greater than the upstream industry's main product threshold. If there is no product in the same main product set that is greater than the threshold of the upstream industry's main product... If the recommended main product is not a core industry H, then the upstream industry is considered not to belong to the upstream industry of the core industry H; otherwise, the main products in the same set that exceed the upstream industry threshold will be included. The recommended main product is added to the related product set, and the alternative upstream industry is added to the next set. The upstream industry cluster proceeds to step 302.5; upstream industry main material threshold. Set to the greater of 10 million yuan and 5% of the total amount received by all enterprises in core industry H from enterprises in the alternative upstream industry;

[0098] Step 302.5: Update the current upstream industry level. ;if If the number of upstream industries exceeds the maximum level of downstream industries, then all upstream industries at all levels are merged and output, and the extraction of upstream industries ends; otherwise, return to step 302.2 for the next iteration, and output the first... Each upstream industry in the upstream industry set is designated as the core industry H for the next iteration.

[0099] Step 4: Determine the list of companies in each industry chain;

[0100] The specific method is as follows:

[0101] Step 401: Summarize the recommended main products of each industry in each industrial chain to obtain the recommended main product list of this industry;

[0102] Step 402: For each enterprise in the selected region, if its sales invoices contain products from the recommended main product list of the industry, and the total sales amount of the enterprise's sales invoices containing products from the recommended main product list of the industry divided by the enterprise's total sales amount is greater than the industry chain determination threshold, then the enterprise is identified as an enterprise within the industry chain; the enterprise chain determination threshold is preferably 0.8, and those skilled in the art can select it according to the actual situation;

[0103] Step 403: For enterprises within the industry chain, re-determine their industry affiliation based on the sales invoice amount of the goods. If the determined industry affiliation is within the industry chain, retain the enterprise; if the determined industry affiliation is not within the industry chain, remove the enterprise from the industry chain.

[0104] The specific determination method is as follows: If a company's registered industry in the current industry chain is already part of that industry chain, then determine whether the ratio of the total sales amount of the products listed in the recommended main product list of this industry on the company's sales invoices to the company's total sales amount is greater than the first industry determination threshold. If the threshold is greater than 1, the enterprise is considered to be within the industry chain, and its industry is consistent with its registered industry; otherwise, the enterprise is removed from the industry chain. (First industry determination threshold) The preferred value is 0.5;

[0105] If a company's registered industry in the current supply chain is not originally part of that supply chain, then it is determined whether the company's sales invoices contain any of the recommended main products in any industry within the supply chain. If so, the ratio of the company's total sales revenue in the recommended main products lists of these industries to the company's total sales revenue is calculated to see if it exceeds the second industry determination threshold. If the ratio is greater than the threshold, the enterprise is classified into the corresponding industry; otherwise, the enterprise is classified into the second industry. If so, the company will be removed from the supply chain; in order to determine the threshold for the second industry. To ensure that a company belongs to only one industry, preferably, a second industry determination threshold is used. Take 0.7;

[0106] Step 404: Extract the companies in each industry chain and combine them into a list of companies for each industry chain;

[0107] With the list of industry enterprises, a directed graph can be constructed, using enterprises as nodes and transaction relationships between enterprises as edges. Note that transaction relationships between enterprises need to be filtered, retaining only those where the goods are the main products of the seller's industry or the main materials of the buyer's industry. Simultaneously, transaction amounts on edges should be filtered, keeping only edges exceeding a threshold.

[0108] Step 5: Identify the key companies in the company list;

[0109] Step 501: For the enterprise list of each industry chain obtained in Step 4, establish an enterprise transaction network diagram; where network nodes are enterprises, and the edges connecting nodes represent the invoice transaction relationships between enterprises, with transaction amount and traded goods as the attributes of the edges, and the direction of the edges is consistent with the transaction direction, that is, from the seller to the buyer. If there are multiple transactions between two enterprises, they are first summarized according to the traded products, and an edge is established for each type of product between the enterprises.

[0110] Step 502: Filter the nodes and edges in the enterprise transaction network graph from Step 501 to generate the main transaction network graph of the industrial chain; the filtering rules are as follows:

[0111] 1) Only retain borders with transaction amounts greater than 500,000;

[0112] 2) Only retain edges for specific products; specifically, the transaction product corresponding to the edge must satisfy the condition that it is both the recommended main product of the seller's industry and the recommended main material of the buyer's industry;

[0113] 3) Remove nodes that have no edges after edge filtering;

[0114] Step 503: Merge the edges in the main transaction network diagram of the industrial chain obtained in Step 502. That is, if there are multiple edges between two enterprises in the main transaction network, merge them into one edge according to the amount. The amount attribute of the merged edge is the sum of the amounts of the original edges.

[0115] Step 504: Calculate the centrality index of each node in the main transaction network diagram of the industry chain. The centrality index includes: node degree centrality, total purchase and sales amount of the node, node feature vector centrality, and node betweenness centrality.

[0116] Node degree centrality is defined as the number of edges connecting a node.

[0117] The total purchase and sales amount of a node is defined as the total amount of the edges connecting that node.

[0118] The centrality of a node's feature vector is defined as:

[0119] For each industry chain backbone transaction network diagram Let V represent the set of all nodes in the main transaction network graph of the industry chain, and E represent the set of all edges in the main transaction network graph of the industry chain; its adjacency matrix is ​​defined as follows. When node v is connected to node t ,otherwise , Let d represent the total number of nodes. Then the eigenvector centrality of node v is... The following relationship must be satisfied:

[0120] Where M(v) is the set of nodes connected to node v. These are the eigenvalues ​​of the adjacency matrix A; Let represent the node betweenness centrality of node t connected to node v.

[0121] The betweenness centrality of nodes satisfies the following relationship:

[0122] in This represents the number of shortest paths from node j through node v to node k.

[0123] Step 504: For each enterprise in the network, calculate its centrality for each criterion and rank it within the network. Each enterprise will have four ranking data points. Adding these four ranking data points together gives the enterprise's overall centrality ranking. For example, if an enterprise ranks first in node centrality, total sales amount, eigenvector centrality, and betweenness centrality, then adding these four ranking data points together gives 1 + 1 + 1 + 1 = 4.

[0124] The top N companies in terms of overall centrality ranking are considered key enterprises in the network. .

[0125] Step 6: Showcase the industry chain and key enterprises;

[0126] The multi-dimensional display includes enterprise chain display, supply and demand and product chain display, value chain display and spatial chain display. Based on the multi-dimensional perspective and combined with tax and economic related needs, further specific analysis and application can be carried out. For detailed functions, please refer to the industrial chain analysis system.

[0127] Among them, the enterprise chain showcases the composition and distribution of key enterprises in each industrial chain;

[0128] The supply and demand and product chain show the correspondence between the main commodities and main materials of the upstream and downstream industries in each industrial chain, which is used to show the transaction links between the upstream and downstream industries in the industrial chain;

[0129] The value chain shows the total transaction amount of each industry in each industry chain, as well as the comparison of the transaction amount of main commodities and main materials that overlap between upstream and downstream industries;

[0130] The spatial chain shows the geographical relationship between the enterprises corresponding to the purchase and sales invoices issued by each key enterprise in each industrial chain and the key enterprise itself.

[0131] This invention also discloses a supply chain and key enterprise analysis system based on tax data, including an indicator model management module, an industry overview module, a supply chain view module, an industry view module, and a key enterprise view module; specific module diagrams are shown below. Figure 2 As shown.

[0132] The indicator model management module is responsible for collecting invoicing data from each enterprise in the selected region and constructing the industry chain based on this data. This module includes a data management submodule, an initialization submodule, an industry chain maintenance submodule, and an industry chain product category library submodule. The data management submodule collects invoicing data from each enterprise in the selected region. The initialization submodule determines the core industries, recommended main products for each industry, and recommended main materials for each industry based on the collected invoicing data amount, and identifies the upstream and downstream industries of each core industry to construct the industry chain. The industry chain maintenance submodule is used to add and delete industry chains, query and maintain the industries contained within an industry chain, and the industry chains to which each industry belongs. The industry chain product category library submodule provides queries for recommended main products and main materials for each industry, as well as calculations of invoice amounts.

[0133] The Industry Overview module can analyze and present the selected industry. This module can display the upstream and downstream industry structure view of any specified industry chain, the recommended main product list of the industry, and the information of key enterprises in the industry chain.

[0134] The industry chain view module can display enterprise chains, supply and demand chains, value chains, and spatial chains. Through the enterprise chain view, analysts can clearly understand the distribution and composition of enterprises within the industry chain. The supply and demand and product chains show the correspondence between the main commodities and main materials of upstream and downstream industries in each industry chain, helping analysts understand the overall supply and demand situation of the industry, as well as the transaction links between upstream and downstream industries. The value chain shows the total transaction amount of each industry in each industry chain, helping analysts quickly understand the overall value distribution of an industry. The spatial chain view uses a vector map to show the geographical relationship between the enterprises corresponding to the purchase and sales invoices issued by each key enterprise in each industry chain and the key enterprise itself.

[0135] The industry view module includes sub-modules for enterprise distribution, main industry commodities and main materials, and industry scale. The enterprise distribution sub-module analyzes the distribution and changes in enterprises of different sizes within the industry, industry evolution, and the proportion and growth of enterprises within the industry, enabling analysts to understand the development of private enterprises. The main industry commodities and main materials sub-module analyzes the composition and changes of main commodities and main materials in a single industry. The industry scale sub-module examines the overall development of the industry from the perspective of its total size, analyzing the development of enterprises within the industry across different sales scale ranges.

[0136] The Key Enterprise View module provides an overall analysis and presentation of key enterprises in each industry chain, displaying detailed invoice information for these enterprises. The Enterprise View shows information for each enterprise, including basic information, purchase and sales information, and information on other enterprises in the upstream and downstream industries of that enterprise.

[0137] The applicant of this invention has provided a detailed description of the embodiments of the invention in conjunction with the accompanying drawings. However, those skilled in the art should understand that the above embodiments are merely preferred embodiments of the invention. The detailed description is only intended to help readers better understand the spirit of the invention and is not intended to limit the scope of protection of the invention. On the contrary, any improvements or modifications made based on the inventive spirit of the invention should fall within the scope of protection of the invention.

Claims

1. A method for constructing industry chains and key enterprises based on tax data, characterized in that: The method for constructing an industry chain based on tax data includes the following steps: Step 1: Collect the invoice amount for each enterprise in the selected region, and select and determine the core industry based on the invoice amount; Step 2: Select the recommended main products and recommended main materials for each industry, and construct a list of recommended main products and a list of recommended main materials for each industry; Step 3: Construct the industry chain based on the recommended main product list and recommended main material list of the core industries, and refer to each core industry and its downstream and upstream industries as an industry chain; Step 4: Determine the list of companies in each industry chain; Step 5: Identify the key companies in the company list; Step 5 includes the following steps: Step 501: For each industry chain enterprise list obtained in Step 4, establish an enterprise transaction network diagram; In this network, the nodes are enterprises, and the edges connecting the nodes represent the invoice transaction relationship between enterprises. The transaction amount and the traded goods are used as the attributes of the edges, and the direction of the edges is consistent with the transaction direction, that is, from the seller of the goods to the buyer of the goods. If there are multiple transactions between two companies, they are first summarized according to the products traded, and an edge is established between the companies for each type of product. Step 502: Filter the nodes and edges in the enterprise transaction network diagram in Step 501 to generate the main transaction network diagram of the industrial chain; Step 503: Merge the edges in the main transaction network graph of the industrial chain obtained in Step 502; Step 504: Calculate the centrality index of each node in the main transaction network diagram of the industry chain, including node degree centrality, total purchase and sales amount of the node, node feature vector centrality, and node betweenness centrality; Step 504: For each enterprise in the network, calculate its centrality for each type and rank it in the network. Each enterprise receives four ranking data points. Adding these four ranking data points together yields the enterprise's overall centrality ranking. The top N enterprises in terms of overall centrality ranking are considered key enterprises in the network. ; Step 6: Showcase the industry chain and key enterprises.

2. The method for constructing the industrial chain and key enterprises based on tax data according to claim 1, characterized in that: In step 1, enterprises are classified according to their registered industries, the total invoicing amount of all enterprises in each industry is calculated, and the industries are ranked according to the total invoicing amount. Industries whose ranking is above the core industry ranking threshold are selected as core industries.

3. The method for constructing the industrial chain and key enterprises based on tax data according to claim 1, characterized in that: In step 2, the total invoice amount for each item is calculated according to the product code on each enterprise's sales invoice, as well as the percentage of the invoice amount for that item in its industry. For each industry, items with an invoice amount percentage greater than the recommended main product screening threshold are selected as the recommended main products for that industry. The total amount of the input invoice for each item is calculated based on the commodity code on the input invoice of each enterprise, as well as the proportion of the total amount of the input invoice for that item in the industry. For each industry, items whose proportion of the total amount of the input invoice is greater than the screening threshold for recommended main materials are selected as the recommended main materials for that industry.

4. The method for constructing the industrial chain and key enterprises based on tax data according to claim 1 or 3, characterized in that: In step 3, the downstream and upstream industries of each core industry are identified. Upstream industries refer to industries that provide raw materials to this industry, while downstream industries refer to industries that sell products to this industry.

5. The method for constructing the industrial chain and key enterprises based on tax data according to claim 4, characterized in that: Let H be a core industry, and the method for determining the downstream industries is as follows: Step 301.1: Set the maximum number of downstream industry levels to 3, and initialize the current number of downstream industry levels. =1; Step 301.2: Locate all sales invoices for companies registered under the core industry H, and summarize the invoice amounts based on the commodity codes on the sales invoices and the industries registered by the purchasing companies. All industries involved by the purchasing companies are considered as potential downstream industries of the core industry H. For each potential downstream industry, calculate the total invoice amount issued by companies in the core industry H to companies in that potential downstream industry. If this amount exceeds a given threshold... If so, proceed to step 301.3; otherwise, it is determined that the candidate downstream industry is not a downstream industry of the core industry H. The larger of 50 million yuan and 3% of the total amount of invoices issued by all enterprises in core industry H is selected; Step 301.3: If the recommended main material of the candidate downstream industry does not overlap with the recommended main product of the core industry H, then the candidate downstream industry is determined not to be a downstream industry of the core industry H; otherwise, proceed to step 301.

4. Step 301.4: Identify the set of recommended main products in core industry H that overlap with the recommended main materials in the candidate downstream industry, and determine whether the total invoiced amount of each recommended main material in the set in the candidate downstream industry exceeds the threshold for main materials in the downstream industry. If the set does not contain any materials exceeding the threshold of the downstream industry's main materials... If the recommended main material is not a downstream industry of core industry H, then it is considered that the downstream industry is not a downstream industry of core industry H. Otherwise, add the alternative downstream industry to the first... The downstream industry cluster is then proceeded to step 301.5; Downstream industry main material threshold Set to the larger of 10 million yuan and 5% of the total amount offered by all companies in core industry H to companies in the candidate downstream industry; Step 301.5: Update the current downstream industry level. = +1; if If the number of downstream industries exceeds the maximum level of downstream industries, then all downstream industries at all levels are merged and output, and the extraction of downstream industries ends; otherwise, return to step 301.2 for the next iteration, and the number of downstream industries is... Each downstream industry in the downstream industry set is designated as the core industry H for the next iteration.

6. The method for constructing the industrial chain and key enterprises based on tax data according to claim 4, characterized in that: Let H be a core industry. The method for determining the upstream industry is as follows: Step 302.1: Set the maximum number of upstream industry levels and initialize the current number of upstream industry levels. =1; Step 302.2: Find all the input invoices of enterprises whose registered industry is core industry H, and summarize the invoice amounts according to the commodity code of the input invoice supplier and the registered industry of the selling enterprise; The industries involved by the selling company are all considered as alternative upstream industries of the core industry H. For each alternative upstream industry, calculate the total invoice amount received by companies in the core industry H from companies in that alternative upstream industry. If this amount exceeds a given threshold... If so, proceed to step 302.3; otherwise, it is determined that the candidate upstream industry is not an upstream industry of the core industry H. The larger of 50 million yuan and 3% of the total amount of invoices received by all enterprises in core industry H is selected; Step 302.3: If the recommended main materials of the core industry H do not include the recommended main products of the candidate upstream industry, then the candidate upstream industry is deemed not to be the upstream industry of the core industry H; otherwise, proceed to step 302.

4. Step 302.4: Identify the set of identical recommended main products for the core industry H's recommended main material and the candidate upstream industry's recommended main product, and determine whether the total input invoice amount for each recommended main product in the set is greater than the upstream industry's main product threshold. If there is no product in the same main product set that is greater than the threshold of the upstream industry's main product... If the recommended main product is not a core industry H, then the upstream industry is considered not to belong to the upstream industry of the core industry H; otherwise, the main products in the same set that exceed the upstream industry threshold will be included. The recommended main product is added to the related product set, and the alternative upstream industry is added to the next set. The upstream industry cluster proceeds to step 302.5; upstream industry main material threshold. Set to the greater of 10 million yuan and 5% of the total amount received by all enterprises in core industry H from enterprises in the candidate upstream industry; Step 302.5: Update the current upstream industry level. = +1; if If the number of upstream industries exceeds the maximum level of the upstream industry, then merge and output all upstream industries at all levels, and end the extraction of upstream industries; otherwise, return to step 302.2 for the next iteration, and output the first... Each upstream industry in the upstream industry set is designated as the core industry H for the next iteration.

7. The method for constructing an industrial chain and key enterprises based on tax data according to claim 5 or 6, characterized in that: Step 4 includes the following: Step 401: Summarize the recommended main products of each industry in each industrial chain to obtain the recommended main product list of this industry; Step 402: For each enterprise in the selected region, if its sales invoices contain products from the recommended main product list of the industry, and the total sales amount of the products from the recommended main product list of the industry on the enterprise's sales invoices divided by the enterprise's total sales amount is greater than the industry chain determination threshold, then the enterprise is identified as an enterprise within the industry chain. The threshold for determining the supply chain is 0.8; Step 403: For companies within the industrial chain, re-determine their industry based on the sales invoice amount of the goods. If the company's industry is determined to be within this industry chain, then the company is retained. If the industry to which the company belongs is not in the industry chain, then the company is removed from the industry chain. Step 404: Extract the companies in each industry chain and combine them into a list of companies for each industry chain.

8. The method for constructing the industrial chain and key enterprises based on tax data according to claim 7, characterized in that: The specific method for determining whether a company's industry is within the industrial chain in step 403 is as follows: If a company's registered industry is already part of the industrial chain, then determine whether the ratio of the total sales amount of the recommended main product list of the company's products existing in this industry to the company's total sales amount is greater than the first industry determination threshold. If the value is greater than the value, the enterprise is considered to be within the industrial chain, and its industry is consistent with its registered industry; otherwise, the enterprise is removed from the industrial chain. If a company's registered industry in the current supply chain is not originally part of that supply chain, then it is determined whether the company's sales invoices contain any of the recommended main products in any industry within the supply chain. If so, the ratio of the company's total sales revenue in the recommended main products lists of these industries to the company's total sales revenue is calculated to see if it exceeds the second industry determination threshold. If the value is greater than 1, the company is classified into the corresponding industry. If no product is listed in the recommended main product list of any industry in the industrial chain, or if the calculated ratio is not greater than the second industry judgment threshold. If so, the company will be removed from the supply chain.

9. The method for constructing the industrial chain and key enterprises based on tax data according to claim 1, characterized in that: In step 504, The degree centrality of a node is defined as the number of edges connecting that node; The total purchase and sales amount of a node is defined as the total amount of the edges connecting that node. The centrality of a node's feature vector is defined as: For each industry chain backbone transaction network diagram , This represents the set of all nodes in the main transaction network diagram of the industry chain. This represents the set of all edges in the main transaction network graph of the industry chain; its adjacency matrix is ​​defined as follows. When node v is connected to node t ,otherwise , d represents the total number of all nodes; Then the eigenvector centrality of node v The following relationship must be satisfied: Where M(v) is the set of nodes connected to node v. These are the eigenvalues ​​of the adjacency matrix A; The node betweenness centrality of node t connected to node v is represented. The betweenness centrality of nodes satisfies the following relationship: in, This represents the number of shortest paths from node j through node v to node k.

10. The method for constructing the industrial chain and key enterprises based on tax data according to claim 1, characterized in that: In step 6, the industrial chain and key enterprises are displayed, including enterprise chain display, supply and demand and product chain display, value chain display and spatial chain display. The enterprise chain showcases the composition and distribution of key enterprises in each industry chain; The supply and demand and product chain show the correspondence between the main commodities and main materials of the upstream and downstream industries in each industrial chain, which is used to show the transaction links between the upstream and downstream industries in the industrial chain; The value chain shows the total transaction amount of each industry in each industry chain, as well as the comparison of the transaction amount of main commodities and main materials that overlap between upstream and downstream industries; The spatial chain shows the geographical relationship between the enterprises corresponding to the input and output invoices issued by each key enterprise in each industrial chain and the key enterprise itself.

11. An analysis system based on the method for constructing industrial chains and key enterprises based on tax data as described in any one of claims 1-10, characterized in that: The analysis system includes an indicator model management module, an industry overview module, an industry chain view module, an industry view module, and a key enterprise view module. The indicator model management module is responsible for collecting invoicing data from each enterprise in the selected region and constructing the industrial chain based on this data. This module includes a data management submodule, an initialization submodule, an industrial chain maintenance submodule, and an industrial chain commodity category library submodule. The Industry Overview module can analyze and present the selected industry. The content displayed in this module can be arbitrarily specified as the upstream and downstream industry structure view in the industrial chain, the recommended main product list of the industry, and the information of key enterprises in the industrial chain. The industry chain view module can display enterprise chains, supply and demand chains, value chains, and spatial chains. The supply and demand and product chains show the correspondence between the main commodities and main materials of the upstream and downstream industries in each industrial chain; the value chain shows the total transaction amount of each industry in each industrial chain. The spatial chain uses a vector map to show the geographical relationship between the enterprises corresponding to the input and output invoices issued by each key enterprise in each industrial chain and the key enterprise itself. The industry view module includes a sub-module for enterprise distribution, a sub-module for main products and main materials in the industry, and a sub-module for industry scale. The enterprise distribution sub-module can analyze the distribution and increase / decrease of enterprises of different sizes in the industry, the industry changes of enterprises, and the proportion and growth of enterprises in the industry. The main products and main materials sub-module can analyze the composition and changes of main products and main materials in a single industry. The industry scale sub-module can view the overall development of the industry from the perspective of the overall industry scale and analyze the development of enterprises in the industry from different sales scale ranges. The key enterprise view module can provide an overall analysis and display of key enterprises in each industry chain, including basic enterprise information, enterprise purchase and sales information, and information on other enterprises in the upstream and downstream industries of that enterprise.

12. The industry chain and key enterprise analysis system based on tax data according to claim 11, characterized in that: In the indicator model management module, the data management submodule collects invoicing data for each enterprise in the selected region; The initialization submodule determines the core industry, the list of recommended main products for the industry, and the list of recommended main materials for the industry based on the collected invoice data amount, and identifies the upstream and downstream industries of each core industry to construct the industrial chain; The industry chain maintenance submodule is used to add and delete industry chains, query and maintain the industries contained in the industry chain, and the industry chain to which the industry belongs; The industry chain commodity category library submodule provides the ability to query recommended main commodities and main materials in the industry, as well as calculate invoice amounts.

Images