KR102234130B1 - An apparatus and method for managing transaction information providing automatic matching between accounts receivables and deposit information - Google Patents

Abstract

A transaction management apparatus according to an embodiment of the present invention includes: a transaction management service providing unit that provides a management service of a transaction information database for transaction detail information collected from a network; And a matching analysis processing unit that performs matching processing between deposit and withdrawal information corresponding to the new transaction information and sales information according to the learning-based matching analysis processing for new transaction information in the transaction information database, wherein the transaction management service providing unit Based on the matching result, an automated matching service between sales receivables and deposit and withdrawal information is provided.

Description

TECHNICAL FIELD An apparatus and method for managing transaction information providing automatic matching between accounts receivables and deposit information}

The present invention relates to a method and apparatus for managing accounts receivable. More specifically, the present invention relates to a transaction information management method and apparatus for providing an automated matching service between learning-based accounts receivable and deposit and withdrawal information.

Recently, as the pace of economic development has accelerated, the number of orders and transactions between companies is also showing a remarkable rate of increase. It has a pattern.

When a business relationship is established, a seller who sells goods and services normally generates accounts receivable through a series of billing processes targeting the purchaser, and proceeds with the payment confirmation process from the purchaser. . When billing occurs, accounts receivable in the state of receivables are formed, and cash, credit transactions, drafts, and B2B loans (e.g., loans secured by accounts receivable, loans for corporate purchases, etc.) ), account transfer is processed according to the use of a credit card, etc., payment is completed through such payment details confirmation and matching, and tax calculation processing is performed.

In general, companies manage various data such as purchases, sales, and salaries mainly by recording them in physical or electronic books, etc., and calculating sales performance, operating profit, total salary, net profit, and tax calculation based on confirmation of deposit details. The calculation of accounting information such as is made through such book management.

In such book management, there is an aspect that can be automated by computer processing such as the ERP management system for the calculation of simple accounting information, but in the case of checking the payment details corresponding to accounts receivable, the deposit method, deposit method, or deposit cycle for each company. All of these are different, and the name of the depositor is frequently changed, so it is very difficult to automate, and there is a problem in that manual work is unconditionally required, and human resources are required accordingly.

In addition, there are difficulties in manually handling the matching management between deposit details and accounts receivable. Transaction information including deposit details may have different bank field names for each bank, and the company due to various representations of the company type. Name irregularities exist, foreign companies' Korean notation methods are different, and the input text format may have different character sets and encoding types according to the computer policy of the recipient bank, and the exception is the official bankbook of the remittance company. There is a situation where there is a deposit, in which a plurality of accounts receivables are merged, there is a case of prepayment, or even a case where the beneficial owner makes a lump sum deposit for the transaction price of various business sites with different representatives.

In this way, in the transaction information management, especially the deposit confirmation and matching process for the management of accounts receivables, due to its irregularity, a problem arises that is complex and difficult to judge. Not only is it difficult to automate, but it is also difficult to actually handle it manually. There is a problem that takes a lot of time.

The present invention is proposed to solve the above-described conventional problem, and by performing a learning process based on a case-based pattern matching process for transaction information data including deposit details, existing informal deposit details and accounts receivable An object of the present invention is to provide a transaction information management method and apparatus capable of enabling automated classification determination for interrelationship matching and effectively shortening the processing time.

In order to achieve the above object, an apparatus according to a preferred embodiment of the present invention includes: a transaction management service providing unit for providing a management service of a transaction information database for transaction details information collected from a network; And a matching analysis processing unit that performs matching processing between deposit and withdrawal information corresponding to the new transaction information and sales information according to the learning-based matching analysis processing for new transaction information in the transaction information database, wherein the transaction management service providing unit Based on the matching result, an automated matching service between sales receivables and deposit and withdrawal information is provided.

In addition, a method according to an embodiment of the present invention is a transaction management method, comprising: providing a transaction information database management service for transaction details information collected from a network; Performing matching processing between deposit and withdrawal information corresponding to the new transaction information and sales information according to a learning-based matching analysis processing for new transaction information in the transaction information database; And providing an automated matching service between accounts receivable and deposit/withdrawal information based on the matching result.

Meanwhile, the method according to the embodiment of the present invention as described above may be implemented with a program for executing the method on a computer and a recording medium in which the program is recorded.

According to the present invention having such a configuration, it is possible to construct a learning pattern database that facilitates classification and determination of the deposit details by performing learning processing using a case-based pattern matching process corresponding to the deposit details data.

And, according to an embodiment of the present invention, by creating a pattern mapping table by using deposit history data, account receivable information, and customer logic learning information obtained from transaction information for a predetermined period of time stored in the initial learning pattern database construction. In addition, the automated classification system can be converted into a database without the need for individual user manual input procedures, so processing efficiency and system compatibility can be improved.

In addition, according to an embodiment of the present invention, by providing a verification process based on receivable detail information corresponding to a certain period of time for a pattern database, an automated verification process is provided without a separate user's manual verification, and is generated accordingly. The accuracy of the mapping table of the learned pattern database can be improved.

In addition, according to an embodiment of the present invention, by providing an automated payment confirmation process based on an initial learning pattern database, and by providing a learning pattern update according to a user confirmation input, the pattern matching accuracy is further improved according to the accumulation of the user's service use. Can be.

Therefore, it is possible to determine the automatic classification of the relationship between the existing informal deposit history and the relationship matching between the accounts receivable, even without a separate user input, and the automation between the receivables and deposit and withdrawal information that can effectively shorten the manual processing time. It is possible to provide a transaction information management method and apparatus for providing a matched service.

1 is a conceptual diagram schematically showing an entire system according to an embodiment of the present invention.
2 is a flowchart illustrating an overall system operation according to an embodiment of the present invention.
3 is a more detailed block diagram of a matching analysis processing unit according to an embodiment of the present invention.
4 is a diagram for explaining transaction information classified according to an embodiment of the present invention.
5 is a diagram illustrating an example of a matrix generated by a matrix generator.
6 is a diagram showing an example of a matrix in which scores are counted by a score counting unit.
7 and 8 are examples of screens output through a user interface unit.
9 and 10 are flowcharts illustrating a method of updating a pattern DB according to an embodiment of the present invention.

In the present invention, various modifications may be made and various embodiments may be provided, and specific embodiments will be illustrated in the drawings and described in detail.

However, this is not intended to limit the present invention to a specific embodiment, it should be understood to include all changes, equivalents, and substitutes included in the spirit and scope of the present invention.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In the present application, terms such as "comprise" or "have" are intended to designate the presence of features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, but one or more other features. It is to be understood that the presence or addition of elements or numbers, steps, actions, components, parts, or combinations thereof does not preclude in advance.

Unless otherwise defined, all terms used herein including technical or scientific terms have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related technology, and should not be interpreted as an ideal or excessively formal meaning unless explicitly defined in the present application. Does not.

Hereinafter, preferred embodiments of the present invention will be described in more detail with reference to the accompanying drawings. In describing the present invention, in order to facilitate an overall understanding, the same reference numerals are used for the same elements in the drawings, and duplicate descriptions for the same elements are omitted.

1 is a conceptual diagram schematically showing an entire system according to an embodiment of the present invention.

Referring to FIG. 1, the entire system according to an embodiment of the present invention may constitute a transaction management system, a user interface unit 10 for providing an automated matching service between accounts receivable and deposit and withdrawal information, and a transaction management service system. Study 20, learning-based matching analysis processing unit 30, pattern database 39, may include a transaction information database 29 and communication unit 50.

Here, the components constituting each system are each implemented as a modularized processor and electrically connected to form a single transaction management device, or partially combined and configured as a modularized processor as needed, and the remaining components are internal network It can also be implemented as a system that is connected through.

To this end, the communication unit 50 may process access and data transmission/reception with the internal network or financial service network, and the network may be a local area network (LAN), a wide area network (WAN), or value-added network. Implemented as a wired network such as a value added network (VAN), or a mobile radio communication network, satellite communication network, Bluetooth, Wibro (Wireless Broadband Internet), HSDPA (High Speed Downlink Packet Access), LTE (Long Term Evolution) can also be implemented as a wireless network.

For example, the system according to an embodiment of the present invention outputs transaction information processed by the transaction management service providing unit 20 through the user interface unit 10, and receives a user input corresponding to the output transaction information. It may be implemented as a transaction management device that updates the transaction information database 29.

In addition, in the system according to an embodiment of the present invention, the transaction management service providing unit 20, the learning-based matching analysis processing unit 30, the pattern database 39, and the transaction information database 29 constitute one transaction management device. , The user interface unit 10 may provide processing of user input/output information corresponding to a separate user terminal connected to the transaction management device.

As such, the system according to an embodiment of the present invention may provide a transaction management process and related services. The transaction management service may basically include information storage and update processing corresponding to the transaction information database 29. The transaction information may include, for example, transaction details information, and the transaction details information may include various information for transaction management, such as customer information, sales receivable information, and deposit details information.

In particular, the transaction management device according to an embodiment of the present invention can collect and manage transaction details information in real time in providing a transaction management service, and use this to generate account receivable information, and deposit and withdrawal details information corresponding thereto. When is collected, a result confirmation interface through an automated matching process is provided through the user interface unit 10, thereby providing an automated matching service between accounts receivable and deposit and withdrawal information.

Accordingly, the user can check the automated matching information between the accounts receivable and deposit and withdrawal information output through the user interface unit 10, and input the corresponding user confirmation information so that the update of the transaction information database 29 is processed. can do.

In order to provide such a matching service, the transaction management apparatus according to an embodiment of the present invention may perform a learning-based pattern analysis processing corresponding to the transaction information database 29 through the learning-based matching analysis processing unit 30. .

According to an embodiment of the present invention, the learning-based matching analysis processing unit 30 may perform an initial analysis process of the transaction information database 29 for processing a matching service, and perform an initial pattern database construction according to the initial analysis process. Thus, a pattern mapping table may be created and stored in the pattern database 39.

Here, the learning-based matching analysis processing unit 30 may collect transaction learning information of the transaction information database 29 for learning-based processing. Here, the transaction learning information may include learning information extracted from transaction details information of the transaction information database 29 accumulated for a certain period of time, and may include structured learning information for learning-based processing.

In addition, the learning-based matching analysis processing unit 30 may process matching analysis corresponding to the structured learning information through various learning algorithms, and by storing mapping table information formed according to the matching analysis in the pattern database 39 , You can build an initial pattern database.

In particular, in an embodiment of the present invention, the learning process of the learning-based matching analysis processing unit 30 is based on an automated artificial intelligence learning process and a case based reasoning (CBR) processing. It may be exemplified to perform analysis processing of the transaction information database 29.

In this case, the learning-based matching analysis processing unit 30 can obtain a matching case between the previously matched accounts receivable and deposit and withdrawal details information by structuring the transaction details information, and through the similarity function processing according to the data pattern analysis between each case. , Grouping processing between similar cases can be performed. In addition, the learning-based matching analysis processing unit 30 may build an initial pattern database by building a pattern mapping table for each case corresponding to the grouping process in the pattern database 39.

Accordingly, the pattern database 39 may include mapping rule information between accounts receivable corresponding to the deposit/withdrawal data pattern, and may include mapping table information grouped on a case-based basis.

For example, the learning case information obtained from the transaction information database 29 includes customer information classified in advance according to company information, brief pattern information of deposit and withdrawal details mapped in response to the customer information, deposit date information, deposit cycle information, It may include information on the deviation of the deposit day, information on the transaction passbook, information on the transaction means (cash, draft, bond-backed loan, etc.), and transaction code information. In addition, the learning-based matching analysis processing unit 30 performs a grouping process according to the accumulation and similarity analysis processing of the learning case information to provide an initial pattern database 39 including data pattern information learned through the pattern database 39. You can build it.

In addition, according to an embodiment of the present invention, the initial pattern database 39 can improve the accuracy by performing a verification process corresponding to the initial pattern database based on the separately input receivable information (or receivable information). have.

To this end, based on the initial pattern database 39, the learning-based matching analysis processing unit 30 automatically matches the transaction information database 29 for a certain period of time, and the actual unaccepted details are The pattern database can be verified by checking whether it matches exactly.

For example, the learning-based matching analysis processing unit 30 may obtain actual receivable information according to the information on issuance of tax invoices versus sales receivables obtained from the transaction information database 29 for the current time, and actual receivable information and transactions Matching accuracy can be verified by comparing the receivable information determined according to the pattern matching analysis of the information database 29.

The learning-based matching analysis processing unit 30 can predict the matching accuracy for the pattern database 39 through such verification processing, and the user performs an analysis of the transaction information database 29 for more accurate database construction. Alternatively, a separate user setting input may be input through the user interface unit 10.

For example, tuning setting information according to a user input corresponding to the initial pattern database 39 may be input through the user interface unit 10, and the pattern database 39 update processing is performed according to the input tuning setting information. Can be. The tuning setting information may include adjustment information of a pattern mapping table, similarity function parameter change information, or case information adjustment information.

Meanwhile, the transaction management service provider 20 may provide a mapping, storage, and management service for accounts receivable in the transaction information database 29.

In particular, the transaction management service providing unit 20 may provide a receivable matching service corresponding to the deposit and withdrawal details between each customer received from the financial service network through the communication unit 50, and for this purpose, a learning-based matching analysis processing unit A transaction information mapping process and output service based on the pattern database 39 using 30 may be provided through the user interface unit 10.

To this end, the transaction management service providing unit 20, when new deposit and withdrawal details information is received through the communication unit 50 or the user interface unit 10, stores it in the transaction information database 29, and the learning-based matching analysis processing unit ( 30) to perform the receivable matching process based on the pattern database 39, and the matching result information may be received and output through the user interface unit 10.

In addition, the transaction management service providing unit 20 may update the transaction information database 29 by receiving a user confirmation input corresponding to the matching result information through the user interface unit 10. For example, the user may input a matching complete button or, if necessary, input a confirmation input corresponding to a user confirmation request corresponding to a matching result having a partial similarity of less than a certain value through the user interface unit 10.

On the other hand, the transaction management service provider 20 may update the transaction information database 29 based on the matched deposit and withdrawal history information, and the updated transaction information database 29 through the learning-based matching analysis processing unit 30 By re-performing the learning process, it is possible to process the pattern database 39 to be updated.

Accordingly, the learning-based matching analysis processing unit 30 may perform additional learning processing to improve the accuracy of the pattern database 39 by using repetitive transaction management service processing and user confirmation input information. Accordingly, as the usage records of the user's transaction management service are accumulated, the accuracy of the matching analysis can be continuously improved.

In addition, the learning-based matching analysis processing unit 30 according to an embodiment of the present invention may determine change information according to data pattern analysis corresponding to a newly input matching case, and when the change information is greater than a predetermined value, abnormal pattern information May be provided to the transaction management service providing unit 20. The transaction management service providing unit 20 may output an abnormal pattern information notification through the user interface unit 10 to process an abnormality notification of pattern analysis.

For example, the learning-based matching analysis processing unit 30, the payment cycle delay pattern corresponding to the data pattern for each customer, change of the deposit date, change of the depositor name, change of the deposit bank, occurrence of divided payments, omission of deposit, or other monitoring information. By analyzing in real time and outputting notification information corresponding thereto through the user interface unit 10, it is possible to provide more real-time convenience for deposit and withdrawal and account receivable management.

2 is a flowchart illustrating an overall system operation according to an embodiment of the present invention.

Referring to FIG. 2, the operation of the entire system according to an embodiment of the present invention may be described as an operation of the transaction management device. First, the transaction management device includes transaction learning information accumulated for a certain period from the transaction information database 29. Is collected (S101).

Then, the transaction management apparatus performs case reasoning learning (CBR)-based similarity pattern analysis on the accumulated transaction learning information through the learning-based matching analysis processing unit 30 (S103).

Thereafter, the transaction management apparatus builds an initial pattern database according to the pattern analysis result through the pattern database 39 (S105), and performs verification and adjustment of the initial pattern database based on the receivable information (107).

Here, the verification and adjustment process may include a process of comparing actual accounts receivable for a certain period of time and calculating accuracy corresponding to the automated matching of the tax invoice and deposit details of the transaction information database 29 for a certain period of time.

In addition, the transaction management apparatus applies tuning setting information according to a user input input through the user interface unit 10 to construct the pattern database 39 (S109).

For example, the tuning setting information may include setting and filtering information of the pattern database 39 for tuning a matching rate, such as by industry type, transaction type, and exception case.

Then, when the pattern database 39 is built, the transaction management device provides a receivable matching service for the new deposit and withdrawal details of the transaction information database through the transaction management service providing unit 20 (S111).

Here, the transaction management service providing unit 20 may selectively provide a user confirmation request corresponding to a matching result having a similarity of less than or equal to a predetermined value (S113), and receive user confirmation information corresponding thereto.

For example, similarity information may be classified into strong matching, weak matching, non-matching, etc. In case of weak matching, user confirmation information may be requested. The user may input the matched deposit and withdrawal details and confirmation information corresponding to the sales receivable information through the user interface unit 10. The user identification information may be, for example, any one of matched, mismatched, or unknown.

In addition, the transaction management device may process the update of the transaction information database 29 according to the matching completion input (S115), and through the learning-based matching analysis processing unit 30, a new transaction information database 29 is updated. It is possible to additionally perform a learning process based on the matching transaction information and the user confirmation information (S117).

Accordingly, an update of the pattern database 39 according to the learning result may be processed (S119), and additional accuracy of the pattern database 39 may be improved through repeated execution of such matching processing.

Hereinafter, a pattern DB update process of the matching analysis processing unit according to an embodiment of the present invention will be described in more detail.

3 is a block diagram illustrating a matching analysis processing unit according to an embodiment of the present invention in more detail, and FIG. 4 is a diagram illustrating transaction information classified according to an embodiment of the present invention.

Referring to FIG. 3, as described above, the matching analysis processing unit 30 generates an initial learning-based pattern database (DB) 39 according to the learning-based analysis processing for the transaction information database 29 for a certain period of time. can do.

For example, in order to establish the initial pattern DB 39, the matching analysis processing unit 30 stores sales information (accepted information) of the past one year from the transaction information database 29 based on the time of inquiry of each customer's electronic tax invoice. Including) and the bank inquiry time of each customer, the deposit and withdrawal information for the past one year is scraped and collected and analyzed, and the initial pattern DB 39 can be constructed accordingly.

In addition, when the pattern DB 39 is constructed, the matching analysis processing unit 30 may process the update of the pattern DB 39 linked with the update of new deposit and withdrawal details information and the mapping information of the transaction information database 29.

To this end, the matching analysis processing unit 30 may collect new mapping case information from the transaction information database 29. More specifically, the matching analysis processing unit 30 may collect sales information after the previous inquiry point of each customer's electronic tax invoice and deposit and withdrawal information after the previous inquiry point of the bank of each customer, and the currently collected sales information and Trade receivable information among the previously collected sales information can be classified by business number, and the currently collected deposit and withdrawal information and non-matching deposit and withdrawal information among the previously collected deposit and withdrawal information can be classified by bank deposit and withdrawal group key.

In addition, the matching analysis processing unit 30 may process the update of the pattern DB 39 based on the new mapping case information. For example, the matching analysis processing unit 30 performs individual transaction matching with the electronic tax invoice corresponding to the corresponding business number for the bank deposit and withdrawal group key that is the same as the bank deposit and withdrawal group key registered in the previously established pattern DB 39. Through the pattern DB update process, learning evaluation and verification of bank deposit/withdrawal group keys that are not registered in the structured pattern DB can be performed to update the pattern DB.

For such processing, the matching analysis processing unit 30 includes a collection unit 31, an extraction unit 32, a classification unit 33, a pattern DB registration determination unit 34, a matrix generation unit 35, and score calculation. A unit 36, a verification unit 37, and a pattern DB update unit 38 may be included.

First, the collection unit 31 collects new sales information and deposit/withdrawal information for updating the pattern DB.

To this end, the collection unit 31 is a sales information collection unit 31a that scrapes and collects new sales information (including unreceived information) after the point of inquiry before the electronic tax invoice of each customer from the network connected through the communication unit 50. ), and a deposit and withdrawal information collection unit 31b that scrapes and collects new deposit and withdrawal information after a bank transfer inquiry point of each customer.

In addition, the collection unit 31 includes sales information from an institution that manages electronic tax invoices (not shown) and an institution that manages bank deposit and withdrawal information (not shown). Deposit and withdrawal information after the previous inquiry point can be scraped. In addition, the scraped transaction information may be stored and updated in the transaction information database 29.

For example, if the time point of inquiry before the electronic tax invoice is July 25, sales information after the point of inquiry before the electronic tax invoice may mean sales information after July 25.

Meanwhile, if, for example, the customer's bank transfer inquiry point is July 25, the deposit and withdrawal information after the bank transfer inquiry point of the customer may refer to the deposit and withdrawal information after July 25.

Here, the transaction information may include sales information, and the sales information may include business information and item information (eg, sales amount, tax-related information (eg, VAT), detailed item information on the electronic tax invoice). For example, there may be a business number, business name, representative name, etc. Item information may include sales amount, tax-related information (eg, VAT), detailed item information (eg, individual item name, individual item unit price, individual item quantity), etc. have.

Deposit and withdrawal information includes transaction date and time, deposit bank, deposit account number, brief (e.g., Internet withdrawal transfer), client/recipient (e.g., customer name (i.e., business name)), withdrawal amount, deposit amount, balance, withdrawal account memo, and processing It may include a point (for example, the name of a corresponding branch such as Gaeponam), and the like.

In addition, the extraction unit 32 extracts case information for updating the pattern DB from the collected information.

To this end, the extracting unit 32 includes a trade receivable information extracting unit 32a that extracts trade receivable information from the previously collected electronic tax invoice sales information, and a mismatching non-matching deposit and withdrawal information from the previously collected bank deposit and withdrawal information. It includes a deposit and withdrawal information extraction unit (32b).

Here, the previously collected electronic tax invoice sales information may include sales information for a certain period of time that was collected until the point of inquiry before the electronic tax invoice, and the previously collected bank deposit and withdrawal information is a certain period in which the customer's bank transfer information was collected. It may include deposit and withdrawal information during the period.

The previously collected electronic tax invoice sales information may include sales information for which payment has been confirmed and sales information for which payment has not been confirmed, wherein the sales information for which payment has not been confirmed is receivable, and may correspond to information on sales receivables. .

In addition, the mismatched deposit and withdrawal information may mean deposit and withdrawal information that is not matched between the deposit and withdrawal from among the previously collected bank deposit and withdrawal information.

The classification unit 33 performs grouping and classification processing on the extracted case information.

To this end, the classification unit 33 is a classification unit 33a for each business number that classifies the sales information currently collected by the sales information collection unit 31a and the account receivable information extracted from the account receivable information extraction unit 32a on the basis of the business number. ), and bank deposit/withdrawal group key classification unit (33b) for classifying the deposit and withdrawal information currently collected by the deposit and withdrawal information collection unit 31b and the non-matching deposit and withdrawal information extracted by the non-matching deposit and withdrawal information extraction unit 32b based on the bank deposit and withdrawal group key. Includes.

Here, the bank deposit/withdrawal group key is for management convenience, and can be said to be a group key that has the representativeness of deposit and withdrawal grouping. For example, deposit bank, deposit account number, brief, client/recipient, withdrawal account memo, processing store, etc. It may be a value obtained by a hash function (eg, MD5) using information.

Here, the classification operation of the classification unit 33 will be further described with reference to FIG. 4. In FIG. 4, it is assumed that the time of inquiry before the electronic tax invoice is July 25, and July 26 is assumed to be the current inquiry time. In this case, the sales information 15 up to the point of inquiry before the electronic tax invoice may be composed of sales information 12 and account receivable information 14 for which payment has been confirmed. In FIG. 4, reference numeral 16 denotes sales information between a previous inquiry point and a current inquiry point of the electronic tax invoice.

Accordingly, the classification unit 33a by business number uses the sales receivable information 14 extracted by the sales receivable information extraction unit 32a and the sales information 16 currently collected by the sales information collection unit 31a, based on the business number. Can be classified.

For example, if the total number of sales receivable information 14 extracted by the sales receivable information extracting unit 32a and the sales information 16 currently collected by the sales information collecting unit 31a were 10, they would have been classified based on business number. Cases can be classified into several groups (eg, four groups are assumed). In this case, the first group may have a unique business number and 3 pieces of information may be included, the second group may have a unique business number and include 3 pieces of information, and the third group may have a unique business number and 3 One piece of information may be included, and the fourth group may have a unique business number and one piece of information may be included. Here, the above-described four groups and the number of information included in each group are only an example, and are not limited thereto.

Meanwhile, FIG. 4 may also be applied to the bank deposit/withdrawal group key classification unit 33b. In this case, July 25 may be the time point for the customer's bank transfer inquiry, and July 26 may be the current time for bank inquiry. In this case, reference numeral 14 may be mismatched deposit and withdrawal information, and reference numeral 16 may be currently collected deposit and withdrawal information. 4, the bank deposit/withdrawal group key classification unit 33b includes non-matching deposit and withdrawal information 14 extracted from the non-matching deposit and withdrawal information extraction unit 32b and deposit and withdrawal information 16 currently collected by the deposit and withdrawal information collection unit 31b. ) Can be classified based on the bank deposit/withdrawal group key.

For example, if the number of non-matching deposit and withdrawal information 14 extracted by the non-matching deposit and withdrawal information extracting unit 32b and the deposit and withdrawal information 16 currently collected by the deposit and withdrawal information collection unit 31b is a total of 10, these are the bank deposit and withdrawal group keys. When classified by criteria, it can be classified into several groups (eg, five groups are assumed). In this case, the first group may have a unique bank deposit/withdrawal group key and 3 pieces of information may be included, the second group may have a unique bank deposit and withdrawal group key and 2 pieces of information may be included, and the third group may have a unique bank deposit/withdrawal group key. With a deposit and withdrawal group key, two pieces of information can be included, the fourth group has a unique bank deposit and withdrawal group key and two pieces of information are included, and the fifth group has a unique bank deposit and withdrawal group key and one piece of information is included. I can. Here, the above-described five groups and the number of information included in each group are only one example, and are not limited thereto.

On the other hand, the pattern DB registration determination unit 34 determines whether the same bank deposit/withdrawal group key as the currently classified bank deposit/withdrawal group key is registered in the structured pattern DB 39 by the bank deposit/withdrawal group key classification unit 33b. do.

If the bank deposit and withdrawal group key identical to the currently classified bank deposit and withdrawal group key is registered in the structured pattern DB 39, the pattern DB registration determination unit 34 is a business number that exists in pairs with the bank deposit and withdrawal group key. An update command (including a bank deposit/withdrawal group key) to perform an individual transaction details matching operation with the electronic tax invoice of the user may be transmitted to the pattern DB update unit 38.

Conversely, if the bank deposit/withdrawal group key identical to the currently classified bank deposit/withdrawal group key is not registered in the structured pattern DB 39, the pattern DB registration determination unit 34 uses the currently classified bank deposit/withdrawal group key as a residual item. Can be marked.

In addition, the matrix generation unit 35 includes account receivable information 14 classified by business number for one or more bank deposit/withdrawal group keys marked as residual items by the pattern DB registration determination unit 34 and the currently collected sales information. Using (16), a learning pattern matrix is generated.

Here, the matrix may be configured in the form of a lookup table, and case-based learning pattern information may be calculated and inserted. The matrix generator 35 may generate a matrix composed of a matrix as shown in FIG. 5. The matrix of FIG. 5 lists business names (e.g., business A, business B, business C, business D) corresponding to each business number in a row (horizontal), and deposit/withdrawal bank names corresponding to each bank deposit/withdrawal group key. (Deposit and withdrawal A, deposit and withdrawal B, deposit and withdrawal C, deposit and withdrawal D, deposit and withdrawal E) can be listed in columns (vertical).

In FIG. 5, each business name has a unique business number, and each business name may include one or more sales information and/or trade receivable information.

In FIG. 5, each deposit and withdrawal bank name has a unique bank deposit and withdrawal group key, and each deposit and withdrawal bank name may include one or more deposit and withdrawal information and/or non-matching deposit and withdrawal information.

Referring to FIG. 5, each business name and each deposit/withdrawal bank name may be defined as a factor, and the score inclusion unit 36 is the correlation between the factors in the matrix generated by the matrix generation unit 35. Can be analyzed and the corresponding score can be entered into the matrix.

Accordingly, the score calculation unit 36 analyzes the correlation between the information included in each business number and the information included in each bank deposit/withdrawal group key in analyzing the relationship between the factors in the matrix, but includes in each business number. It is possible to analyze text similarity, amount and date similarity, and amount similarity between the obtained information and the information included in each bank deposit/withdrawal group key.

In the case of text similarity analysis, it is analyzed whether the text of the operator and the text of the deposit/withdrawal bank match each other. Here, the text of the business operator may be the business name (the name of the customer) in the sales information of the electronic tax invoice, and the text of the deposit/withdrawal bank may be the text of the brief column (that is, the name of the depositor) in the deposit and withdrawal information of the bank.

In the case of the similarity analysis of the amount and date, it is determined whether the amount and date of the business operator coincide with the amount and date of the deposit and withdrawal bank. In the case of analysis of the similarity of the amount and date, only the amount of deposit after issuance of the electronic tax invoice is analyzed in accordance with the rules for deposit after issuance of the electronic tax invoice.

In the case of the similarity analysis of the amount, it is determined whether the sum of single or multiple amounts between the information included in the business number and the information included in the bank deposit/withdrawal group key is determined, and regardless of the date, it is simply determined whether or not the amount is identical.

In the score inputting unit 36, when the text, the amount, and the date all match each other, the score “80” may be input, and when the amount and the date coincide with each other, the score “60” You can enter ", and if only the amount matches each other, you can enter "30". In this way, the score will be input differently depending on whether one of the text, the amount, and the date coincide, two coincide, or all three coincide. That is, if the text, the amount, and the date are all analyzed as similar (or the same) according to the text similarity, the amount and date similarity, and the amount similarity analysis, the score incorporation unit 36 calculates the score of the highest value (eg, 80) If it is analyzed that the amount and date are similar (or the same) according to the amount and date similarity analysis, enter the score of the median value (e.g., 60), and only the amount is considered to be similar (or the same) according to the amount similarity analysis. Once analyzed, you can enter the score of the lowest value (eg, 30).

In other words, the score inclusion unit 36 is based on whether one of the text similarity, the amount and date similarity, and the amount similarity analysis is similar (or the same), two are similar (or the same), or all three are similar (or the same). Differential scores can be entered at the intersection of information and deposit and withdrawal bank information.

Accordingly, as illustrated in FIG. 5, the score calculation unit 36 may analyze the correlation between factors in the matrix and input a score corresponding thereto into the matrix. The higher the matching reliability (ie, relevance), the higher the score may be input.

In this way, when a matrix in which scores are included as shown in FIG. 5 is generated by the score incorporation unit 36, it is possible to determine which of information for each business number and information for each bank deposit/withdrawal group key matches each other.

The verification unit 37 optimizes the matrix by limiting the score to a value greater than zero, and sequentially performs matching verification based on bank deposit and withdrawal and matching verification based on the electronic tax invoice for the optimized matrix. Handle it.

Then, the verification unit 37, in verifying based on the bank deposit and withdrawal, first checks whether a single business operator exists for the bank deposit and withdrawal. In the case of Fig. 5, there is a single business operator for deposit and withdrawal C, deposit and withdrawal D, and deposit and withdrawal E, respectively. That is, for deposit and withdrawal C, there is a business operator C, for deposit and withdrawal D, there is a business C, and for deposit and withdrawal E, there is a business D. In this case, the verification unit 37 considers that the deposit and withdrawal C and the business operator C are clearly matched with each other (ie, the information of the bank deposit and withdrawal group key corresponding to the deposit and withdrawal C and the information of the business number corresponding to the business C). The matching level to be newly registered in the pattern DB 39 is temporarily set as "system recommendation", and the deposit/withdrawal D and the business operator C consider that they are clearly matched, and the information of the bank deposit/withdrawal group key corresponding to the deposit/withdrawal D and For the information of the business number corresponding to the business operator C), the matching level to be newly registered in the pattern DB 39 is temporarily set as "system recommendation", and the deposit/withdrawal E and the business operator D are regarded as obviously matching each other, and these (ie, For the information of the bank deposit/withdrawal group key corresponding to the deposit/withdrawal E and the business number corresponding to the business D), a matching level to be newly registered in the pattern DB 39 is temporarily set as “system recommendation”.

After checking the existence of single business operators described above, it is checked whether there are multiple business operators based on bank deposit and withdrawal. In the case of FIG. 5, it is checked that there is a business operator A and a business operator B for deposit and withdrawal A, and it is checked that there are business operators A and B for deposit and withdrawal B. In this case, the verification unit 37 regards the information of the deposit and withdrawal A as matching the information of the operator A or the information of the operator B, and sets the matching grade to be newly registered in the pattern DB 39 for them as "recommended". , The information of the deposit/withdrawal B is regarded as matching with the information of the operator A or the information of the operator B, and the matching grade to be newly registered in the pattern DB 39 for them is set as “recommended”.

In addition, the verification unit 37 verifies only information for which the matching grade is temporarily set as "system recommendation" in verifying based on the electronic tax invoice. First, it checks whether there is a single bank deposit or withdrawal. In the case of FIG. 5, it can be checked that there is a single deposit and withdrawal E for the operator D. Accordingly, the information included in the operator D and the information included in the deposit/withdrawal E are individually matched, and if all of them are matched, the matching level for them is determined as "system recommendation". If the matching failure is not matched at all, the grade is determined as "recommended". Here, in individually matching the information included in the corresponding business operator D and the information included in the deposit and withdrawal E, the electronic tax invoice and the bank deposit and withdrawal are matched in reverse order. By performing such matching, if the remaining portion of the electronic tax invoice occurs, verification fails, and in this case, the grade is determined as "recommended" due to the matching failure. The balance of bank deposits and withdrawals is assumed to be the sales before the time of collection.

After checking the existence of single bank deposits and withdrawals described above, it is checked whether there are multiple bank deposits and withdrawals based on the electronic tax invoice. In the case of FIG. 5, it can be checked that there is a deposit and withdrawal C and a deposit and withdrawal D for the operator C. Accordingly, the details of the operator C and the details of the deposit and withdrawal C are individually matched, and the detailed details of the operator C and the details of the deposit and withdrawal D are individually matched and reported, and the details of deposit and withdrawal C and deposit and withdrawal D are merged. By matching the details of the operator C and the operator C individually, it is checked whether the operator C matches the deposit and withdrawal C, the deposit and withdrawal D, and the deposit and withdrawal C + deposit and withdrawal D. If the details of the operator C are completely matched only with the details of the deposit and withdrawal C, the matching level of the operator C and the deposit and withdrawal C is determined as "system recommendation". If the details of the operator C are completely matched only with the details of the deposit and withdrawal D, the matching grade of the operator C and the deposit and withdrawal D is determined as "system recommendation". If the details of the operator C completely match the details of the deposit and withdrawal C and the deposit and withdrawal D, the matching grade of the operator C and the deposit and withdrawal C + deposit and withdrawal D is determined as "system recommendation". If the operator C does not completely match any of the deposit and withdrawal C, deposit and withdrawal D, and deposit and withdrawal C + deposit and withdrawal D, the matching failure is "recommended" (in this case, the (Meaning that it can match one).

Accordingly, the verification unit 37 may perform matching verification between information for each business number and information for each bank deposit/withdrawal group key based on the matrix in which the score is included, and determine a matching grade.

Meanwhile, among the above-described matching grades, the'system recommendation' grade may be referred to as a first grade, and among the above-described matching grades, the'recommendation' grade may be referred to as a second grade.

Meanwhile, the pattern DB update unit 38 stores information belonging to the bank deposit/withdrawal group key (that is, the same key as the previously registered bank deposit/withdrawal group key) included in the update command from the pattern DB registration determination unit 34 For the deposit/withdrawal group key, the transaction details are individually matched with the information belonging to the business number that exists as a pair. Accordingly, the matching grade (pre-grade) of the corresponding pattern information in the pattern DB 39 may be updated or maintained as it is.

In addition, the pattern DB update unit 38 may update the pattern DB 39 based on the verification result of the verification unit 37. In this case, the pattern DB 39 is updated based on information classified based on one or more bank deposit/withdrawal group keys and business numbers marked as residual items.

The pattern DB 39 includes pattern mapping table information using (bank deposit/withdrawal group key, business number) as a basic pair.

More specifically, the pattern DB 39 may include pattern information including a (bank deposit/withdrawal group key, business number), a matching grade (pre-grade), and a pre-registration reason for the (bank deposit/withdrawal group key, business number). That is, each pattern information in the pattern DB 39 may be composed of {(bank deposit/withdrawal group key, business number), matching grade (pre-grade), pre-registration reason}. Here, the business number is the business number of the electronic tax invoice, and may be replaced with a business name or a representative name. The prior grade may include user confirmation, system recommendation, and recommendation. The reason for pre-registration may be referred to as the reason for registration or may be a combination of text, amount, and date.

A predetermined number of pattern information is previously registered in the pattern DB 39. And, in the pattern DB 39, newly created pattern information based on sales information after the time of inquiry before the electronic tax invoice, deposit and withdrawal information after the time of inquiry before the bank, and previously collected trade receivable information and previously collected non-matching deposit and withdrawal information. May be registered or previously registered pattern information may be updated.

As the pattern DB update unit 38 updates the pattern DB 39 in this way, the matching analysis processing unit 30 may transmit the updated matching details to the transaction management service providing unit 20.

The transaction management service providing unit 20 may provide a user interface as shown in FIG. 7 or 8 through the user interface unit 10 and may receive a matching confirmation input and a user confirmation input through the user interface.

For example, a matching details screen including a recommended matching tab as shown in FIG. 7 may be provided through the user interface unit 10, and the user interface unit 10 has a pre-level of pattern information in the pattern DB 39 In the case of system recommendation, the matching result of transaction details may be output through the recommendation matching tab. The bank transaction details on the left of the recommended matching tab screen and the sales details of the electronic tax invoice on the right may indicate matching result information corresponding to strong matching. That is, in the case of the recommended matching tab, only transaction details when the matching reliability (eg, score) is greater than or equal to a threshold (eg, 50) may be output. 7 shows only the case where the bank transaction details on the left and the sales details of the electronic tax invoice on the right are strongly matched with each other in the text, the amount, and the date. Accordingly, the user can key-in the matching cancellation or matching confirmation for each transaction after confirming the bank transaction details on the left and the sales details of the electronic tax invoice on the right on the matching details screen of FIG. 7. In the case of FIG. 7, since the bank transaction details on the left and the sales details of the electronic tax invoice on the right are strongly matched with each other in the text, the amount, and the date, the user can easily check and key-in the matching confirmation for all transaction details.

On the other hand, the handwritten matching tab screen of the matching details of FIG. 8 shows the matching results of transaction details when the prior grade of the pattern information in the pattern DB 39 is recommendation. It may be a case that the bank transaction details on the left side of the handwritten matching tab screen and the sales details in the electronic tax invoice on the right are weakly matched. That is, in the handwritten matching tab, only transaction details when the matching reliability (eg, score) is less than or equal to a threshold (eg, 50) may be displayed. In FIG. 8, when the user selects the inquiry 41 after keying the inquiry period 40 in the bank deposit details inquiry on the left, the bank transaction details 42 during the inquiry period are displayed. Subsequently, when the user selects, for example, a specific summary 43, the bill details inquiry result screen 44 on the right corresponding thereto may be displayed. That is, as shown in FIG. 8, in the handwriting matching tab, a case in which the bank transaction details on the left and the sales details of the electronic tax invoice on the right are weakly matched may be output in the text, the amount, and the date. In the case of FIG. 8, it is shown that there are three sales details of a bill that can be weakly matched to any one transaction details 43 according to a bank deposit details inquiry. Accordingly, the user can key-in the matching cancellation or matching confirmation for each transaction after meticulously checking the result of the bank deposit details inquiry on the left and the result of the bill details inquiry on the right on the screen of FIG. 8.

Meanwhile, a pattern DB update method according to an embodiment of the present invention will be described with reference to the flowcharts of FIGS. 9 and 10.

In order to update (update) the pattern DB, the matching analysis processing unit 30 firstly includes sales information (including unaccepted information) after the previous inquiry point of each customer's electronic tax invoice, and deposit and withdrawal information after the previous inquiry point of each customer's bank. It is collected by scraping (S10, S20).

Accordingly, the matching analysis processing unit 30 extracts trade receivable information from the previously collected electronic tax invoice sales information, and extracts non-matching deposit and withdrawal information from the previously collected bank deposit and withdrawal information.

And, the matching analysis processing unit 30 classifies the currently collected sales information and the extracted trade receivable information based on the business number, and classifies the currently collected deposit and withdrawal information and the extracted non-matching deposit and withdrawal information based on the bank deposit and withdrawal group key ( S30, S40).

Here, the bank deposit/withdrawal group key is for management convenience, and can be said to be a group key that has the representativeness of deposit and withdrawal grouping. For example, deposit bank, deposit account number, brief, client/recipient, withdrawal account memo, processing store, etc. It may be a value obtained by a hash function (eg, MD5) using information.

Thereafter, the matching analysis processing unit 30 determines whether a bank deposit/withdrawal group key identical to the currently classified bank deposit/withdrawal group key is registered in the structured pattern DB 39 (S50).

If the same bank deposit/withdrawal group key as the currently classified bank deposit/withdrawal group key is registered in the structured pattern DB 39, the matching analysis processing unit 30 performs the currently classified bank deposit/withdrawal group key (that is, the previously registered bank deposit/withdrawal group key). The information belonging to the same bank deposit and withdrawal group key as the group key) is individually matched with the information belonging to the business number that exists as a pair for the corresponding bank deposit and withdrawal group key. Accordingly, the matching grade (pre-grade) of the corresponding pattern information in the pattern DB 39 may be updated or maintained as it is (S60).

Conversely, if the bank deposit/withdrawal group key identical to the currently classified bank deposit/withdrawal group key is not registered in the structured pattern DB 39, the matching analysis processing unit 30 generates new pattern information (that is, the pattern DB 39 ) To update) the currently classified bank deposit/withdrawal group key is marked as a residual item (S70).

Subsequently, when the matching analysis processing unit 30 has finished determining whether to register the pattern DB for all bank deposit/withdrawal group keys (“Yes” in S80), the at least one bank deposit/withdrawal group key marked as a residual item is classified based on the business number. A matrix is created using the accounts receivable information and the currently collected sales information.

Subsequently, the matching analysis processing unit 30 optimizes the matrix by limiting the score value to a value greater than zero, and for the optimized matrix, matching verification based on bank deposit and withdrawal and matching verification based on electronic tax invoice Perform sequentially (S100).

After the verification is completed, the matching analysis processing unit 30 finally updates the pattern DB 39 based on the verification result (S110). In this case, the pattern DB 39 is updated based on information classified based on one or more bank deposit/withdrawal group keys and business numbers marked as residual items.

In addition, the above-described method according to an embodiment of the present invention may be implemented as a computer-readable code on a computer-readable recording medium. The computer-readable recording medium includes all types of recording devices that store data that can be read by a computer system. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tapes, floppy disks, and optical data storage devices. Further, the computer-readable recording medium is distributed over a computer system connected through a network, so that computer-readable codes can be stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for implementing the method can be easily inferred by programmers in the art to which the present invention pertains.

As described above, an optimal embodiment has been disclosed in the drawings and specifications. Although specific terms have been used herein, these are only used for the purpose of describing the present invention, and are not used to limit the meaning or the scope of the present invention described in the claims. Therefore, those of ordinary skill in the art will understand that various modifications and equivalent other embodiments are possible therefrom. Therefore, the true technical scope of the present invention should be determined by the technical spirit of the appended claims.

Claims (18)

In the transaction management device,
A transaction management service providing unit that provides a management service of a transaction information database for transaction detail information collected from a network; And
A matching analysis processing unit for performing matching processing between deposit and withdrawal information corresponding to the new transaction information and sales information according to the learning-based matching analysis processing for new transaction information in the transaction information database,
The transaction management service provider provides an automated matching service between sales receivables and deposit and withdrawal information based on the matching processing result,
The matching analysis processing unit
Based on the case-based reasoning learning processing of the transaction information database, an initial pattern database is constructed by performing matching processing between deposit and withdrawal information corresponding to transaction learning information for a certain period of time and sales information,
By comparing the actual unaccounted details information separately input in correspondence with the predetermined period of time and the receivable mapping details of the automated matching analysis processing result corresponding to the initial pattern database, the matching analysis accuracy verification result corresponding to the initial pattern database is obtained. Print it out,
At least one of adjustment information of a pattern mapping table, similarity function parameter change information, or case information adjustment information set in response to case-based reasoning learning of the initial pattern database, according to tuning setting information according to user input corresponding to the verification result output Is updated to build a pattern database for learning-based matching analysis processing,
The transaction management service providing unit receives user confirmation information corresponding to the matching processing result and transmits it to the matching analysis processing unit,
The matching analysis processing unit updates the pattern database based on matching information of the new transaction information determined according to the user confirmation information,
The matching analysis processing unit,
A collection unit for collecting, from the transaction information database, electronic tax invoice sales information for a certain period of time based on a time of inquiry of a customer's electronic tax invoice and bank deposit and withdrawal information of a certain period of the past based on a time of inquiry of the customer's bank; And
An extraction unit for extracting learning data for learning the case-based reasoning of the initial pattern database from the sales information and the bank deposit and withdrawal information collected by the collection unit; further comprising,
The sales information includes business information and item information on the electronic tax invoice,
The extraction unit
A deposit confirmation sales information extracting unit for extracting sales information from the electronic tax invoice sales information collected by the collection unit on the basis of unaccepted information;
A deposit confirmation deposit and withdrawal information extracting unit for extracting bank deposit and withdrawal information determined based on an unaccepted time point from the bank deposit and withdrawal information collected by the collection unit;
A deposit-determined sales information classification unit for classifying the deposit-confirmed e-tax invoice sales information extracted by the deposit-determined sales information extracting unit based on business information of each e-tax invoice; And
A deposit confirmed deposit and withdrawal information classification unit for classifying the deposit confirmed bank deposit and withdrawal information extracted from the deposit confirmed deposit and withdrawal information extracting unit based on a bank deposit and withdrawal group key,
The matching analysis processing unit,
A lookup table generated by mapping the deposit confirmed sales information based on the business information and the deposit confirmed sales information based on the bank deposit and withdrawal group key classified by the deposit confirmed sales information classification unit and the deposit confirmed deposit and withdrawal information classification unit. ) Further comprising a pattern DB generator for generating the initial pattern database for the case-based reasoning learning using a learning pattern matrix in the form of)
Transaction management device. delete delete The method of claim 1,
The collection unit,
From the transaction information database, collecting sales information after the point of inquiry before the electronic tax invoice of the customer scrapped through the network and deposit and withdrawal information after the point of inquiry before the bank of the customer
Transaction management device. The method of claim 4,
The matching analysis processing unit
Classification of the collected sales information and trade receivable information in the previously collected sales information based on business number, and the non-matching deposit and withdrawal information in the collected deposit and withdrawal information and previously collected bank deposit and withdrawal information based on the bank deposit and withdrawal group key More inclusive of wealth
Transaction management device. The method of claim 5,
The matching analysis processing unit
If the bank deposit/withdrawal group key of the classified information is not registered in the pattern database, further comprising a pattern DB registration determination unit marking the bank deposit/withdrawal group key as a residual item
Transaction management device. The method of claim 6,
The matching analysis processing unit
In the learning pattern matrix generated based on the account receivable information classified based on the business number for the marked bank deposit and withdrawal group key, analyze the association between the account receivable information classified based on the business number and the bank deposit and withdrawal group key Further comprising a score inclusion unit for inputting a score according to the
In analyzing the association, the score incorporation unit performs an analysis between factors within a matrix for at least one of text similarity, amount and date similarity, and money similarity, and learns the score calculated according to the analysis between the factors within the matrix. Into the pattern matrix
Transaction management device. The method of claim 7,
The matching analysis processing unit
Further comprising a verification unit that performs matching verification between the account receivable information classified based on the business number and the bank deposit/withdrawal group key based on the matrix in which the score is included.
Transaction management device. The method of claim 8,
The matching analysis processing unit
Further comprising a pattern DB update unit for updating the pattern database based on the verification result in the verification unit
Transaction management device. In the transaction management method of the transaction management device,
Providing, by a management service provider, a management service of a transaction information database for transaction detail information collected from a network;
Performing, by a matching analysis processing unit, a matching process between deposit and withdrawal information corresponding to the new transaction information and sales information according to a learning-based matching analysis processing for new transaction information in the transaction information database; And
The management service providing unit comprises the step of providing an automated matching service between the sales receivable and deposit/withdrawal information based on the matching processing result,
The step of performing the matching process,
Establishing an initial pattern database by performing, by the matching analysis processing unit, a matching process between deposit and withdrawal information corresponding to the transaction learning information for a certain period of time and sales information based on the case-based reasoning learning processing of the transaction information database;
The matching analysis processing unit compares the actual unaccounted details information separately input corresponding to the predetermined period with the unreceived receivable mapping details of the automated matching analysis processing result corresponding to the initial pattern database, and corresponds to the initial pattern database. Outputting a matching analysis accuracy verification result; And
The matching analysis processing unit, according to tuning setting information according to user input corresponding to the verification result output, adjustment information of a pattern mapping table set in response to case-based reasoning learning of the initial pattern database, similarity function parameter change information or case Comprising the step of updating at least one of the information adjustment information to build a pattern database for learning-based matching analysis processing,
Receiving, by a transaction management service providing unit, user confirmation information corresponding to the matching processing result and transmitting it to the matching analysis processing unit; And
The matching analysis processing unit further comprises updating the pattern database based on matching information of the new transaction information determined according to the user confirmation information,
The matching analysis processing unit,
A collection unit for collecting, from the transaction information database, electronic tax invoice sales information for a certain period of time based on a time of inquiry of a customer's electronic tax invoice and bank deposit and withdrawal information of a certain period of the past based on a time of inquiry of the customer's bank; And
An extraction unit for extracting learning data for learning the case-based reasoning of the initial pattern database from the sales information and the bank deposit and withdrawal information collected by the collection unit; further comprising,
The sales information includes business information and item information on the electronic tax invoice,
The extraction unit
A deposit confirmation sales information extracting unit for extracting sales information from the electronic tax invoice sales information collected by the collection unit on the basis of unaccepted information;
A deposit confirmation deposit and withdrawal information extracting unit for extracting bank deposit and withdrawal information determined based on an unaccepted time point from the bank deposit and withdrawal information collected by the collection unit;
A deposit-determined sales information classification unit for classifying the deposit-confirmed e-tax invoice sales information extracted by the deposit-determined sales information extracting unit based on business information of each e-tax invoice; And
A deposit confirmed deposit and withdrawal information classification unit for classifying the deposit confirmed bank deposit and withdrawal information extracted from the deposit confirmed deposit and withdrawal information extracting unit based on a bank deposit and withdrawal group key,
The matching analysis processing unit,
A lookup table generated by mapping the deposit confirmed sales information based on the business information and the deposit confirmed sales information based on the bank deposit and withdrawal group key classified by the deposit confirmed sales information classification unit and the deposit confirmed deposit and withdrawal information classification unit. ) Further comprising a pattern DB generator for generating the initial pattern database for the case-based reasoning learning using a learning pattern matrix in the form of)
How to manage transactions. delete delete The method of claim 10,
The step of performing the matching process,
The matching analysis processing unit further comprises collecting, from the transaction information database, sales information after a previous inquiry point of the customer's electronic tax invoice scraped through the network and deposit and withdrawal information after the customer's bank previous inquiry point.
How to manage transactions. The method of claim 13,
The step of performing the matching process,
The matching analysis processing unit classifies the collected sales information and the sales receivable information in the previously collected sales information based on the business number, and uses the collected deposit and withdrawal information and the non-matching deposit and withdrawal information in the previously collected bank deposit and withdrawal information to a bank deposit and withdrawal group. Further comprising the step of classifying based on the key
How to manage transactions. The method of claim 14,
The step of performing the matching process,
The matching analysis processing unit further comprises marking a corresponding bank deposit/withdrawal group key as a residual item if the bank deposit/withdrawal group key of the classified information is not registered in the pattern database.
How to manage transactions. The method of claim 15,
The step of performing the matching process,
In the learning pattern matrix generated based on account receivable information classified based on the business number for the marked bank deposit/withdrawal group key, the matching analysis processing unit includes account receivable information classified based on the business number and the bank deposit/withdrawal. Further comprising the step of inputting a score according to the analysis of the association between the group keys,
In the step of inputting, in analyzing the correlation, analysis between factors within a matrix for at least one of text similarity, amount and date similarity, and money similarity, and the score calculated according to the analysis between the factors within the matrix Including the step of incorporating the learning pattern matrix
How to manage transactions. The method of claim 16,
The step of performing the matching process,
The verification unit further comprising the step of performing matching verification between the account receivable information classified based on the business number and the bank deposit/withdrawal group key based on the matrix in which the score is included.
How to manage transactions. The method of claim 17,
The step of performing the matching process,
The matching analysis processing unit further comprising the step of updating the pattern database based on the verification result by the verification unit.
How to manage transactions.

Images