Associative Analysis and Mining Method for Tax-related Data

Abstract

The present invention relates to a method for associated analysis and mining of tax-related data. According to the unified social credit code of the initial tax-related enterprise, the tax-related initial enterprise relationship person, the result set of the tax-related initial enterprise invoicing MAC code, the result set of the tax-related initial enterprise invoicing IP address, and the result set of the main board serial number of the tax-related enterprise corresponding to the initial tax-related enterprise can be found respectively. Furthermore, the tax-related associated enterprise information set of the relationship person, the tax-related associated enterprise information set of the invoicing MAC code, the tax-related associated enterprise information set of the invoicing IP address, and the tax-related associated enterprise information set of the main board serial number can be obtained respectively. The tax-related associated enterprises simultaneously included in the tax-related associated enterprise information set of the relationship person, the tax-related associated enterprise information set of the invoicing MAC code, the tax-related associated enterprise information set of the invoicing IP address, and the tax-related associated enterprise information set of the main board serial number are identified as tax-related suspected enterprises, which can effectively realize the associated analysis and mining of tax-related data, and is safe and reliable.

Description

Technical Field

[0001] The present invention relates to an association analysis and mining method, in particular to an association analysis and mining method for tax-related data. Background Art

[0002] At present, in order to obtain huge profits, lawbreakers exist in the situation of wantonly issuing false value-added tax special invoices to defraud export tax rebates and offset taxes, resulting in a large loss of national taxes.

[0003] How to find out the criminal gangs suspected of issuing false invoices and tax fraud is a difficult problem that urgently needs to be solved at present. Summary of the Invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide an association analysis and mining method for tax-related data, which can effectively realize the association analysis and mining of tax-related data, and is safe and reliable.

[0005] According to the technical solution provided by the present invention, for the association analysis and mining method of the tax-related data, the association analysis and mining method includes the following steps:

[0006] Step 1: Obtain the initial tax-related information, and determine the initial tax-related enterprise and the corresponding basic tax-related information of the initial tax-related enterprise according to the obtained initial tax-related information. The basic tax-related information includes the unified social credit code of the tax-related enterprise;

[0007] Step 2: Extract the tax registration information and the electronic bottom ledger invoice information from the tax big data cloud platform through an ETL tool, so as to process the extracted tax registration information and the electronic bottom ledger information to form an associated analysis tax data source;

[0008] Step 3: According to the unified social credit code of the initial tax-related enterprise, respectively find the corresponding initial tax-related enterprise relationship persons, the initial tax-related enterprise invoicing MAC code result set, the initial tax-related enterprise invoicing IP address result set and the main board serial number result set of the tax-related enterprise in the associated analysis tax data source;

[0009] Step 4: According to the relationship persons of the initial tax-related enterprise, find the related tax-related associated enterprises of any relationship person in the associated analysis tax data source, and use all the related tax-related associated enterprises to form a related tax-related associated enterprise information set;

[0010] According to the result set of the invoicing MAC codes of the initial tax-related enterprises, search in the associated analysis tax data source for the invoicing MAC code tax-related associated enterprises with the same invoicing MAC code as any of the invoicing MAC codes in the result set of the invoicing MAC codes of the initial tax-related enterprises. Using all the invoicing MAC code tax-related associated enterprises, an invoicing MAC code tax-related associated enterprise information set can be formed;

[0011] According to the result set of the invoicing IP addresses of the initial tax-related enterprises, search in the associated analysis tax data source for the invoicing IP address tax-related associated enterprises with the same invoicing IP address as any of the invoicing IP addresses in the result set of the invoicing IP addresses of the initial tax-related enterprises. Using all the invoicing IP address tax-related associated enterprises, an invoicing IP address tax-related associated enterprise information set can be formed;

[0012] According to the result set of the initial main board serial numbers related to taxes, search in the associated analysis tax data source for the main board serial number tax-related associated enterprises with the same main board serial number as any of the initial main board serial numbers in the result set of the initial main board serial numbers related to taxes. Using all the main board serial number tax-related associated enterprises, a main board serial number tax-related associated enterprise information set can be formed;

[0013] Step 5: Identify the tax-related suspected enterprises as the tax-related associated enterprises that are simultaneously included in the information set of the tax-related associated enterprises of the related parties, the invoicing MAC code tax-related associated enterprise information set, the invoicing IP address tax-related associated enterprise information set, and the main board serial number tax-related associated enterprise information set.

[0014] In step 2, the associated analysis tax data source includes tax registration personnel information, enterprise invoicing MAC code information, enterprise invoicing IP address information, and enterprise invoicing main board number information;

[0015] Among them, the tax registration personnel information includes the unified social credit code, personnel name, personnel document number, and personnel identity, and the personnel identity includes legal person, invoice recipient, and / or investor; the enterprise invoicing MAC code information includes the unified social credit code and the invoicing MAC code; the enterprise invoicing IP address information includes the unified social credit code and the invoicing IP address; the enterprise invoicing main board number information includes the unified social credit code and the invoicing main board serial number.

[0016] In step 4, when obtaining the information set of the tax-related associated enterprises of the related parties, search for the identity document information of the tax-related enterprise related parties in the associated analysis tax data source, so as to obtain an identity document information set based on all the identity document information;

[0017] Search for the tax-related associated enterprises of the related parties in the associated analysis tax data source using any identity document information in the identity document information set. The tax-related associated enterprises of the related parties obtained by searching using all the identity document information in the identity document information set can form an information set of the tax-related associated enterprises of the related parties.

[0018] Based on the obtained initial tax-related enterprises, the determined initial tax-related enterprises can be one or more.

[0019] Advantages of the present invention: By using an ETL tool to extract tax registration information and electronic ledger invoice information from the tax big data cloud platform, it is possible to process the extracted tax registration information and electronic ledger information to form an associated analysis tax data source; according to the unified social credit code of the initial tax-related enterprise, it is possible to separately find the initial tax-related enterprise relationship persons, the result set of the initial tax-related enterprise invoicing MAC codes, the result set of the initial tax-related enterprise invoicing IP addresses, and the result set of the main board serial numbers of the tax-related enterprises corresponding to the initial tax-related enterprise, and then it is possible to separately obtain the information set of related person tax-related associated enterprises, the information set of invoicing MAC code tax-related associated enterprises, the information set of invoicing IP address tax-related associated enterprises, and the information set of main board serial number tax-related associated enterprises, and use the tax-related associated enterprises simultaneously included in the information set of related person tax-related associated enterprises, the information set of invoicing MAC code tax-related associated enterprises, the information set of invoicing IP address tax-related associated enterprises, and the information set of main board serial number tax-related associated enterprises to confirm tax-related suspected enterprises, which can effectively realize the associated analysis and mining of tax-related data, and is safe and reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a schematic diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] The present invention will be further described below in conjunction with specific drawings and embodiments.

[0022] As Figure 1 shown: In order to effectively realize the associated analysis and mining of tax-related data, the associated analysis and mining method of the present invention includes the following steps:

[0023] Step 1: Obtain initial tax-related information, and determine the initial tax-related enterprise and the corresponding basic tax-related information of the initial tax-related enterprise according to the obtained initial tax-related information. The basic tax-related information includes the unified social credit code of the tax-related enterprise.

[0024] Specifically, the initial tax-related information is generally a single clue reported by the masses, assigned by the superior department, or transferred by other departments. The specific acquisition channels of the initial tax-related information can be determined according to the actual situation, and are well-known to those skilled in the art in this technical field, and will not be elaborated here. In specific implementation, based on the obtained initial tax-related enterprises, the determined initial tax-related enterprises can be one or more. The number of initial tax-related enterprises is related to the initial tax-related information, and is well-known to those skilled in the art in this technical field, and will not be elaborated here. The situation where the initial tax-related enterprise is one will be described below. For the case of multiple enterprises, the corresponding description can be referred to, and it is well-known to those skilled in this technical field, and will not be elaborated here.

[0025] Step 2: Extract tax registration information and e-invoice ledger information from the tax big data cloud platform through an ETL tool, so as to process the extracted tax registration information and e-invoice ledger information to form an associated analysis tax data source;

[0026] In specific implementation, the present invention is based on the tax registration information and e-invoice ledger information of tax big data, and extracts data from the tax big data cloud platform through an ETL tool. Since the tax registration information and e-invoice ledger information of the big data cloud platform are changing in real time, the ETL tool needs to perform incremental extraction at regular intervals every day. The tax registration information is in the tens of millions of data volume, and the e-invoice ledger information is in the hundreds of millions of data volume. To ensure the efficiency of associated analysis and mining, it is necessary to preprocess the extracted data to improve the query efficiency. The preprocessing function needs to be calculated regularly every day to ensure the timeliness of data update. The specific situation of the ETL tool is well known to those skilled in the art and will not be elaborated here.

[0027] In the embodiment of the present invention, a scheduled task is created using the task management plan built into the windows or linux system, and the extraction tool uses the ETL tool kettle for incremental extraction. The extraction data range is from the cut-off time of the last extraction to the current time. After the extraction is completed, the cut-off time of this extraction is recorded for use when the next scheduled task extracts data. The extraction frequency can be daily or weekly according to the requirements of the local tax bureau.

[0028] In specific implementation, taking the invoice table as an example, the solution for scheduled incremental extraction is as follows. Among them, A is the source library invoice table (with a data volume of hundreds of millions, and the invoicing time is used as the index), B is the invoice table in the target library, C is the timestamp table (used to record the reference time for synchronization), TO is the last synchronization time, and Now is the current synchronization time.

[0029] 1) The original record of table C is (T0, T0), and once the synchronization starts, it will become (T0, Now);

[0030] 2) Retrieve all records in table A where the invoicing time >= T0 and the invoicing time < Now, and update them to table B;

[0031] 3) After the synchronization is completed, table C becomes (Now, Now).

[0032] 4) Repeat the above steps, and the time of Now keeps moving forward.

[0033] Scheduled tasks use the task management plan built into the Windows or Linux system to create scheduled tasks. The timer service technology is used to preprocess tax registration information and e-invoice data in the electronic ledger at regular intervals. The preprocessing is performed each time after the incremental extraction by the ETL tool Kettle. The processing results are data such as tax registration personnel information, enterprise invoicing MAC code information, enterprise invoicing IP address information, and enterprise invoicing motherboard number information. Among them, the tax registration personnel information includes the unified social credit code, personnel name, personnel ID number, and personnel identity (legal person, invoice recipient, investor); the enterprise invoicing MAC code information includes the unified social credit code and invoicing MAC code; the enterprise invoicing IP address information includes the unified social credit code and invoicing IP address; the enterprise invoicing motherboard number information includes the unified social credit code and invoicing motherboard serial number.

[0034] Taking the enterprise invoicing MAC information as an example, the preprocessing solution is as follows. Among them, A is the source database invoice table (indexed by invoicing time); B is the invoice table in the target database (indexed by invoicing time); C(TO,Now) is the incremental timestamp table (used to record the benchmark time for synchronization), TO is the previous synchronization time, and Now is the current synchronization time; D is the enterprise invoicing MAC information table in the target database (unified social credit code, invoicing MAC code).

[0035] 1) The original record in table C is (T0,T0), and once the synchronization starts, it will become (T0,Now);

[0036] 2) Retrieve all records in table A where the invoicing time >= T0 and the invoicing time < Now, and update them to table B;

[0037] 3) Retrieve the unified social credit code and invoicing MAC address of all records in table B where the invoicing time >= T0 and the invoicing time < Now, and compare them with the data in D. If there is no identical unified social credit code and invoicing MAC code, update them to table D.

[0038] 4) Note: Tables B and D are in the same database, and the Oracle database syntax merg into can be used to complete the update. The specific statement is as follows:

[0039] Merge into D using (select unified social credit code, invoicing MAC code from B where invoicing time >= T0 and invoicing time < Now) tmp on (d.unified social credit code = tmp.unified social credit code and d.invoicing MAC code = tmp.invoicing MAC code) when not matched then insert into D (unified social credit code, invoicing MAC code) values (tmp.unified social credit code, tmp.invoicing MAC code);

[0040] After preprocessing, Table C becomes (Now, Now).

[0041] Repeat the above steps, and the time of Now moves forward continuously.

[0042] In summary, in the embodiment of the present invention, the associated analysis tax data source includes tax registration personnel information, enterprise invoicing MAC code information, enterprise invoicing IP address information, and enterprise invoicing main board number information;

[0043] Among them, the tax registration personnel information includes unified social credit code, personnel name, personnel ID number, and personnel identity, and the personnel identity includes legal person, invoice receiver, and / or investor; the enterprise invoicing MAC code information includes unified social credit code and invoicing MAC code; the enterprise invoicing IP address information includes unified social credit code and invoicing IP address; the enterprise invoicing main board number information includes unified social credit code and invoicing main board serial number.

[0044] Step 3: According to the unified social credit code of the initial tax-related enterprise, respectively find in the associated analysis tax data source the tax-related initial enterprise relationship person, the tax-related initial enterprise invoicing MAC code result set, the tax-related initial enterprise invoicing IP address result set, and the tax-related enterprise main board serial number result set corresponding to the initial tax-related enterprise;

[0045] Specifically, those skilled in the art know that for an enterprise, the unified social credit code is a unique code, and the tax-related initial enterprise relationship person of the enterprise can be a legal person, an invoice receiver, or an investor. For the same initial tax-related enterprise, the unified social credit code may correspond to one or more tax-related initial enterprise invoicing MAC codes, tax-related initial enterprise invoicing IP addresses, and tax-related enterprise main board serial numbers, and all corresponding tax-related initial enterprise invoicing MAC codes, tax-related initial enterprise invoicing IP addresses, and tax-related enterprise main board serial numbers can be found in the associated analysis tax data source according to the unified social credit code, so as to respectively form a tax-related initial enterprise invoicing MAC code result set, a tax-related initial enterprise invoicing IP address result set, and a tax-related enterprise main board serial number result set.

[0046] Step 4: According to the related persons of the initial tax-related enterprise, search in the associated analysis tax data source for tax-related associated enterprises of the related persons that are associated with any one of the related persons within the related persons of the initial tax-related enterprise. Using all the tax-related associated enterprises of the related persons, a tax-related associated enterprise information set of the related persons can be formed;

[0047] According to the invoice MAC code result set of the initial tax-related enterprise, search in the associated analysis tax data source for invoice MAC code tax-related associated enterprises with the same invoice MAC code as any one of the invoice MAC codes in the invoice MAC code result set of the initial tax-related enterprise. Using all the invoice MAC code tax-related associated enterprises, an invoice MAC code tax-related associated enterprise information set can be formed;

[0048] According to the invoice IP address result set of the initial tax-related enterprise, search in the associated analysis tax data source for invoice IP address tax-related associated enterprises with the same invoice IP address as any one of the invoice IP addresses in the invoice IP address result set of the initial tax-related enterprise. Using all the invoice IP address tax-related associated enterprises, an invoice IP address tax-related associated enterprise information set can be formed;

[0049] According to the main board serial number result set of the initial tax-related enterprise, search in the associated analysis tax data source for main board serial number tax-related associated enterprises with the same main board serial number as any one of the main board serial numbers in the main board serial number result set of the initial tax-related enterprise. Using all the main board serial number tax-related associated enterprises, a main board serial number tax-related associated enterprise information set can be formed;

[0050] Specifically, when obtaining the tax-related associated enterprise information set of the related persons, search for the identity document information of the tax-related enterprise related persons in the associated analysis tax data source, so as to obtain an identity document information set based on all the identity document information;

[0051] Use any one of the identity document information in the identity document information set to search for tax-related associated enterprises of the related persons in the associated analysis tax data source. The tax-related associated enterprises of the related persons found by using all the identity document information in the identity document information set can form a tax-related associated enterprise information set of the related persons.

[0052] In the embodiment of the present invention, the identity document information generally includes the person's name and the corresponding identity document number. Using one of the identity document information can find one or more tax-related associated enterprises of the related persons. The relationship of the identity document information in the tax-related associated enterprises of the related persons may be inconsistent with the relationship in the initial tax-related enterprise. For example, in the initial tax-related enterprise, it can be the legal person, and in the tax-related associated enterprises of the related persons, it can be the person receiving the invoice, etc. The process of specifically finding the tax-related associated enterprises of the related persons according to the identity document information is well known to those skilled in the art and will not be elaborated here.

[0053] For the specific query process and method of the invoicing IP address, invoicing MAC code, and main board serial number of the initial tax-related enterprise, reference can be made to the above description, which is well-known to those skilled in the art and will not be elaborated here.

[0054] Step 5: Identify the tax-related associated enterprises that are simultaneously included in the tax-related associated enterprise information set of the related person, the tax-related associated enterprise information set of the invoicing MAC code, the tax-related associated enterprise information set of the invoicing IP address, and the tax-related associated enterprise information set of the main board serial number as tax-related suspected enterprises.

[0055] In the embodiment of the present invention, based on the distribution of the tax-related associated enterprises in the tax-related associated enterprise information set of the related person, the tax-related associated enterprise information set of the invoicing MAC code, the tax-related associated enterprise information set of the invoicing IP address, and the tax-related associated enterprise information set of the main board serial number, tax-related suspected enterprises can be analyzed and mined. The tax-related suspected enterprises are those enterprises and related personnel that are associated with the initial tax-related enterprise and may be involved in tax evasion or false invoicing. When the initial tax-related enterprise is the initial clue, the above steps 3 to 5 are used to expand the clue.

[0056] In summary, by using the ETL tool to extract the tax registration information and the electronic ledger invoice information from the tax big data cloud platform, the tax-related data source for correlation analysis can be formed after processing the extracted tax registration information and the electronic ledger information; according to the unified social credit code of the initial tax-related enterprise, the tax-related initial enterprise related person, the result set of the invoicing MAC code of the initial tax-related enterprise, the result set of the invoicing IP address of the initial tax-related enterprise, and the result set of the main board serial number of the tax-related enterprise can be respectively found, and then the tax-related associated enterprise information set of the related person, the tax-related associated enterprise information set of the invoicing MAC code, the tax-related associated enterprise information set of the invoicing IP address, and the tax-related associated enterprise information set of the main board serial number can be respectively obtained. By identifying the tax-related associated enterprises that are simultaneously included in the tax-related associated enterprise information set of the related person, the tax-related associated enterprise information set of the invoicing MAC code, the tax-related associated enterprise information set of the invoicing IP address, and the tax-related associated enterprise information set of the main board serial number as tax-related suspected enterprises, the correlation analysis and mining of tax-related data can be effectively realized, which is safe and reliable.

Claims

1. An associated analysis and mining method for tax-related data, characterized in that The related analysis and mining method includes the following steps: Step 1: Obtain the initial tax-related information, determine the initial tax-related enterprise based on the obtained initial tax-related information, and the basic tax-related information corresponding to the initial tax-related enterprise, where the basic tax-related information includes the unified social credit code of the initial tax-related enterprise; Step 2: Extract the tax registration information and the electronic ledger invoice information from the tax big data cloud platform through an ETL tool, so as to process the extracted tax registration information and the electronic ledger information to form a related analysis tax data source; Step 3: According to the unified social credit code of the initial tax-related enterprise, respectively find in the related analysis tax data source the initial tax-related enterprise relationship person corresponding to the initial tax-related enterprise, the result set of the initial tax-related enterprise's invoicing MAC code, the result set of the initial tax-related enterprise's invoicing IP address, and the result set of the initial tax-related enterprise's invoicing main board serial number; Step 4: According to the relationship person of the initial tax-related enterprise, find in the related analysis tax data source the related person's tax-related associated enterprise associated with any relationship person in the relationship person of the initial tax-related enterprise, and use all the related person's tax-related associated enterprises to form a related person's tax-related associated enterprise information set; According to the result set of the initial tax-related enterprise's invoicing MAC code, find in the related analysis tax data source the invoicing MAC code tax-related associated enterprise with the same invoicing MAC code as any invoicing MAC code of the initial tax-related enterprise in the result set of the initial tax-related enterprise's invoicing MAC code, and use all the invoicing MAC code tax-related associated enterprises to form an invoicing MAC code tax-related associated enterprise information set; According to the result set of the initial tax-related enterprise's invoicing IP address, find in the related analysis tax data source the invoicing IP address tax-related associated enterprise with the same invoicing IP address as any invoicing IP address of the initial tax-related enterprise in the result set of the initial tax-related enterprise's invoicing IP address, and use all the invoicing IP address tax-related associated enterprises to form an invoicing IP address tax-related associated enterprise information set; According to the result set of the initial main board serial number, find in the related analysis tax data source the main board serial number tax-related associated enterprise with the same main board serial number as any main board serial number of the initial main board serial number in the result set of the initial main board serial number, and use all the main board serial number tax-related associated enterprises to form a main board serial number tax-related associated enterprise information set; Step 5: Confirm the tax-related suspected enterprises as the tax-related associated enterprises that are simultaneously included in the related person's tax-related associated enterprise information set, the invoicing MAC code tax-related associated enterprise information set, the invoicing IP address tax-related associated enterprise information set, and the main board serial number tax-related associated enterprise information set.

2. The correlation analysis and mining method for tax-related data according to claim 1, characterized in that, In Step 2, the related analysis tax data source includes tax registration personnel information, enterprise invoicing MAC code information, enterprise invoicing IP address information, and enterprise invoicing main board number information; Among them, the tax registration personnel information includes the unified social credit code, the name of the personnel, the ID number of the personnel, and the identity of the personnel. The identity of the personnel includes legal person, ticket receiver, and / or investor; the enterprise invoicing MAC code information includes the unified social credit code and the invoicing MAC code; the enterprise invoicing IP address information includes the unified social credit code and the invoicing IP address; the enterprise invoicing mainboard number information includes the unified social credit code and the invoicing mainboard serial number.

3. The correlation analysis and mining method for tax-related data according to claim 2, characterized in that, In step 4, when obtaining the information set of the related person's tax-related associated enterprises, search for the ID card information of the initial tax-related enterprise of the related person in the associated analysis tax data source, so as to obtain the ID card information set according to all the ID card information; Use any ID card information in the ID card information set to search for the related person's tax-related associated enterprises in the associated analysis tax data source. The related person's tax-related associated enterprises found by using all the ID card information in the ID card information set can form the information set of the related person's tax-related associated enterprises.

4. The correlation analysis and mining method for tax-related data according to any one of claims 1 to 3, characterized in that, The initial tax-related enterprises that can be determined according to the obtained initial tax-related enterprises are one or more.

Images