KR102486186B1 - System, method and program for providing automated financial management cloud service using web scraping and artificial intelligence - Google Patents

Abstract

The present invention relates to a method for providing a financial management cloud service, which is performed by a system for providing a financial management cloud service. The method of the present invention comprises the steps of: allowing a scraping server to automatically scrape inflow and outflow of user funds from a user terminal; allowing a subject classification server to automatically classify the scraped inflow and outflow of user fund as an accounting account subject by using artificial intelligence (AI) model; allowing a time series analysis server to analyze the inflow and outflow of funds caused by a user automatically classified as the accounting account subject with data for time series, based on deposit/withdrawal transaction details of bank accounts, which correspond to the inflow and outflow of funds resulting from the user extracted by the scraping server, and to perform linking and processing; and allowing a finance analysis server to analyze and predict fund flow, based on information on a result of analysis by time series, which is analyzed with data for time series. Accordingly, a service for combining finance and accounting can be provided in real time.

Description

Automated financial management cloud service provision system, method and program using web scraping and artificial intelligence

The present disclosure relates to a financial management cloud service providing system, method and program. More particularly, the present disclosure relates to a system, method, and program for providing automated financial management cloud services using web scraping and artificial intelligence.

Currently, the basic principle in Korea's corporate accounting standards is based on accrual accounting. Accrual accounting refers to recording the financial effects of transactions, events, and circumstances on a business in the period in which the transaction occurs, rather than in the period in which cash is received or paid. Because accrual accounting is recorded at the time a transaction occurs, separate from receipt or payment of cash, records related to operating activities do not usually coincide with cash inflows and outflows.

From the perspective of operating a business in line with rapidly changing market conditions, cash flow is an important criterion for establishing a mid- to long-term fund management plan and is the result of corporate activities.

Therefore, from the point of view of running a business, a service that analyzes transactions or events that can be said to be the cause of corporate activities and the result, cash inflow and outflow, is required. There is a need to cross-validate the cause and effect of transactions.

In recent years, with the development of online, most transactions are changing from face-to-face cash transactions to non-face-to-face credit transactions, thereby creating an environment in which most transaction activities, such as fund inflows and outflows, can be collected online in real time.

However, the current financial management service is not linked to effective financial management, which can be called the final purpose of business activities, and has not been able to provide a service in which accounting and finance are converged in real time.

Republic of Korea Patent Registration No. 10-1972895 (2019.04.26. notice)

The embodiment disclosed in this disclosure collects the user's fund inflow and outflow behavior (transaction behavior) through scraping, automatically classifies the collected fund inflow and outflow behavior (transaction behavior) into an account, associates it with bank account deposit and withdrawal details, Its purpose is to analyze time-series data and provide real-time services that combine accounting and finance.

In addition, an object of the embodiments disclosed in the present disclosure is to provide a method capable of automatically improving account subject classification performance by vectorizing transaction behavior and account subjects using an artificial intelligence model.

In addition, the embodiment disclosed in the present disclosure aims to provide a service in which accounting and finance are converged by converting account-classified transaction activities into time-series data through linkage with deposit and withdrawal details of a bank account, and analyzing them.

In addition, an object of the embodiments disclosed in the present disclosure is to provide something that can improve the accuracy of processing in automatically classifying the inflow and outflow of funds into accounting account items.

In addition, the embodiment disclosed in the present disclosure analyzes and provides real-time analysis of the inflow and outflow of funds processed through automatic journalism as time-series data based on transaction details of bank accounts, thereby providing that accounting processing can be linked to financial management. There is a purpose.

The problems to be solved by the present disclosure are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the description below.

A financial management cloud service providing method according to an aspect of the present disclosure for achieving the above-described technical problem is performed by a financial management cloud service providing system that provides a financial management cloud service, by a scraping server, a user automatically scraping the user's fund inflow and outflow behavior of the terminal; automatically classifying, by an account subject classification server, the scraped user fund inflow and outflow behavior into accounting account subjects using the artificial intelligence model; Based on the deposit and withdrawal transaction details of the bank account corresponding to the fund inflow and outflow behavior of the user extracted from the scraping server, Analyzing and linking processing with time series data; and analyzing and predicting, by a financial analysis server, the flow of funds based on time-series analysis result information analyzed as the time-series data.

In addition, in the automatic classification process, when the account subject classification server automatically classifies into accounting account subjects using the artificial intelligence model, the accounting account subject of the transaction details corresponding to the user's fund inflow and outflow behavior It may be characterized in that it is encoded as a vector through neural network learning and automatically classified.

In addition, the automatic classification process may be characterized in that at least one of a customer mutual vector, a transaction time vector, a periodicity vector, a customer type vector, and a business type vector is used as an input vector of the artificial intelligence model.

In addition, the automatic classification process may be characterized in that at least one of a customer mutual vector, a transaction time vector, a periodicity vector, a customer type vector, and a business type vector is used as an input vector of the artificial intelligence model.

In addition, in the step of analyzing and linking the time series data, the time series analysis server calculates a cost value according to the linking process when linking the causative fund inflow and outflow behavior with the resultant fund inflow and outflow behavior , It may be characterized by selecting and linking the minimum cost value among the calculated cost values.

In addition, in the step of analyzing and linking processing with the time series data, the cost value of adding or dividing the fund inflow/output behavior by the time series analysis server, the causative fund inflow/output behavior and the resulting bank account deposit/withdrawal It may be characterized in that the cost value according to the time difference between the transaction details and the cost value required for customer matching are sequentially used and linked.

In addition, in the step of analyzing and predicting the flow of funds, when analyzing and predicting the flow of funds by the financial analysis server, an autoregressive cumulative moving average model, a recurrent neural network, and a long-short term model for analyzing the time series data It may be characterized in that at least one of the gate circulation networks is used.

In addition, a financial management cloud service providing system according to another aspect of the present disclosure is a financial management cloud service providing system for providing a financial management cloud service, comprising: a scraping server that automatically scrapes a user's fund inflow and outflow behavior of a user's terminal; an account subject classification server that automatically classifies the scraped user fund inflow and outflow behavior into accounting account subjects using the artificial intelligence model; Based on the deposit and withdrawal transaction history of the bank account corresponding to the fund inflow and outflow behavior that is the result of the user extracted from the scraping server, the fund inflow and outflow behavior that is the cause of the user automatically classified as the accounting account subject is analyzed and linked with time series data. Time series analysis server processing; and a financial analysis server that analyzes and predicts a flow of funds based on time-series analysis result information analyzed with the time-series data.

In addition, the time series analysis server, when linking the causal fund inflow and outflow behavior and the resulting fund inflow and outflow behavior, calculates a cost value according to the linking process, selects the minimum cost value among the calculated cost values, It can be characterized as being connected.

In addition, when the time series analysis server links the causal fund inflow and outflow behavior with the resultant fund inflow and outflow behavior, the cost value for adding the fund inflow and outflow behavior or dividing the fund inflow and outflow behavior, the causative fund inflow and outflow behavior, and It may be characterized in that the cost value according to the difference in time between the deposit and withdrawal transaction details of the resulting bank account and the cost value required for customer matching are sequentially used and linked.

In addition to this, a computer program stored in a computer readable recording medium for execution to implement the present disclosure may be further provided.

In addition to this, a computer readable recording medium recording a computer program for executing a method for implementing the present disclosure may be further provided.

According to the above-mentioned problem solving means of the present disclosure, the user's fund inflow and outflow behavior (transaction behavior) is collected through scraping, the collected fund inflow and outflow behavior (transaction behavior) is automatically classified into accounts, and the bank account's deposit and withdrawal history and It provides the effect of providing a service in which accounting and finance are converged in real time by linking processing and analyzing with time series data.

In addition, according to the above-described problem solving means of the present disclosure, by vectorizing transaction behavior and account subjects using an artificial intelligence model, an effect of automatically improving account subject classification performance is provided.

In addition, according to the above-described problem solving means of the present disclosure, the account classification transaction activity is converted into time series data through linkage with the deposit and withdrawal details of the bank account, and through its analysis, the effect of providing a service in which accounting and finance are converged to provide.

In addition, according to the above-described problem solving means of the present disclosure, an effect of improving processing accuracy in automatically classifying the inflow and outflow of funds into accounting account subjects is provided.

In addition, according to the above-mentioned problem solving means of the present disclosure, the flow of inflow and outflow of funds processed in automatic journalism is analyzed and provided in real time as time series data based on transaction details of bank accounts, so that accounting can be linked to financial management provide effect.

The effects of the present disclosure are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description below.

1 is a diagram showing a financial management cloud service providing system of the present disclosure.
2 is a diagram showing data of the account subject classification server of FIG. 1;
3 is a diagram for explaining a process of learning an artificial intelligence model using account subject classification data of the artificial intelligence model of FIG. 2 .
FIG. 4 is a diagram for explaining a process of creating a context in the context generator of FIG. 3 .
5 is a diagram showing the configuration of an artificial intelligence model for generating an artificial intelligence model in the artificial intelligence model generator of the account subject classification server of FIG. 3 .
6 is a diagram for explaining a process of classifying account subjects using an artificial intelligence model for classifying account subjects in the account subject classification server of FIG. 3 .
7 is a diagram showing the configuration of the time series analysis server of FIG. 1 .
8 is a diagram for explaining a process of outputting time series analysis result information from the time series analysis server of FIG. 7 .
9 is a flowchart illustrating a financial management cloud service providing method of the financial management cloud service providing system of the present disclosure.

Like reference numbers designate like elements throughout this disclosure. The present disclosure does not describe all elements of the embodiments, and general content or overlapping content between the embodiments in the technical field to which the present disclosure belongs is omitted. The term 'unit, module, member, or block' used in the specification may be implemented as software or hardware, and according to embodiments, a plurality of 'units, modules, members, or blocks' may be implemented as one component, It is also possible that one 'part, module, member, block' includes a plurality of components.

Throughout the specification, when a part is said to be "connected" to another part, this includes not only the case of being directly connected but also the case of being indirectly connected, and indirect connection includes being connected through a wireless communication network. do.

In addition, when a certain component is said to "include", this means that it may further include other components without excluding other components unless otherwise stated.

Throughout the specification, when a member is said to be located “on” another member, this includes not only a case where a member is in contact with another member, but also a case where another member exists between the two members.

Terms such as first and second are used to distinguish one component from another, and the components are not limited by the aforementioned terms.

Expressions in the singular number include plural expressions unless the context clearly dictates otherwise.

In each step, the identification code is used for convenience of description, and the identification code does not explain the order of each step, and each step may be performed in a different order from the specified order unless a specific order is clearly described in context. there is.

Hereinafter, the working principle and embodiments of the present disclosure will be described with reference to the accompanying drawings.

In this specification, the 'financial management cloud service providing system according to the present disclosure' includes all of various systems capable of providing results to users by performing calculation processing. For example, a financial management cloud service providing system according to the present disclosure may include a computer, a server, and a portable terminal, or may be in any one form.

Here, the computer may include, for example, a laptop, a desktop, a laptop, a tablet PC, a slate PC, and the like equipped with a web browser.

A portable terminal is, for example, a wireless communication system that ensures portability and mobility, and includes a personal communication system (PCS), a global system for mobile communications (GSM), a personal digital cellular (PDC), a personal handyphone system (PHS), and a PDA ( Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), WiBro (Wireless Broadband Internet) terminal, smart phone This includes all kinds of handheld-based wireless communication systems such as watches, rings, bracelets, anklets, necklaces, glasses, contact lenses, or wearable systems such as head-mounted-devices (HMDs). can

The financial management cloud service providing system according to the present disclosure accumulates information on the user's fund inflow and outflow behavior (transaction behavior) through automated web scraping, and classifies the account subject using the accumulated information using an artificial intelligence model, The results of the performed account subject classification can be managed and analyzed as time series data capable of real-time fund flow analysis, and can be provided to predict future fund flows.

This financial management cloud service provision system collects the user's fund inflow and outflow behavior (transaction behavior) through scraping, automatically classifies the collected fund inflow and outflow behavior (transaction behavior) into an account, and processes it in conjunction with the deposit and withdrawal details of the bank account and by analyzing time-series data, it is possible to provide services in which accounting and finance are converged in real time.

In addition, the financial management cloud service provision system can improve the performance of automatically classifying account subjects by vectorizing transaction behavior and account subjects using an artificial intelligence model.

In addition, the financial management cloud service providing system converts account-classified transaction activities into time-series data by linking them with bank account deposit and withdrawal details, and through analysis thereof, it is possible to provide a service in which accounting and finance are converged.

In addition, the financial management cloud service provision system can improve processing accuracy in automatically classifying the inflow and outflow of funds into accounting account subjects.

In addition, the financial management cloud service providing system analyzes and provides real-time analysis of the inflow and outflow of automatically processed funds as time-series data based on transaction details of bank accounts, so that accounting processing can be linked to financial management.

Hereinafter, a financial management cloud service providing system for providing an automated financial management cloud service using web scraping and artificial intelligence will be examined in detail.

1 is a diagram showing a financial management cloud service providing system of the present disclosure.

Referring to FIG. 1, the financial management cloud service providing system 100 includes a web server 120, a scraping server 130, an account subject classification server 140, a time series analysis server 150, and a financial analysis server 160. can include At this time, each server may be connected to the network and communicate. Also, each server may be configured in the form of a module or service.

First, the web server 120 may receive information necessary for automatic scraping from the user's terminal 110 through subscription and consent of a user (business operator or corporation) for automatic scraping of the scraping server 130 . The web server 120 may provide a web-based financial management cloud service to the user's terminal 110 .

The user's terminal 110 is a time series analysis server that links and processes the information processed from the account subject classification server 140 through the scraping information of the scraping server 130 and the deposit and withdrawal transaction history of the user's bank account as time series data. The result processed through (150), the result of analysis of money flow and the prediction result, which are the processing result of the financial analysis server 160, may be displayed on the screen. For example, the user's terminal 110 may be a web terminal.

The scraping server 130 may automatically scrape a user's fund inflow and outflow behavior (hereinafter referred to as transaction behavior) of the user's terminal 110 . The account subject classification server 140 may automatically classify the user fund inflow/output behavior scraped by the scraping server 130 into accounting account subjects using an artificial intelligence model.

2 is a diagram showing data of the account subject classification server of FIG. 1;

Referring to FIG. 2 , the account subject classification server data 200 for classifying the transaction behavior of the account subject classification server 140 into account subjects may include common data 210 and user data 220 .

The common data 210 may include operator information data 211 , estimated information data 212 , and confusion information data 213 . The business information data 211 may include information such as business name or corporate name, corporate registration number or business registration number, type of business, type of business, phone number, address, representative name, etc., and estimated information data 212 is a business name or information such as keywords (eg 'gas station', 'haejangguk', 'department store', 'cafe', etc.) or franchise names that can be used to estimate the type of business or handling product. The confusion information data 213 may include information on cases where the same object is marked differently (eg, CU ^TM , CU, BGF Retail, BG Retail, etc.).

The user data 220 may include user information data 221, transaction behavior data 222, and account subject classification data 223 of the artificial intelligence model.

The user information data 221 may include information such as the user's ID, company name or corporate name, corporate registration number or business registration number, type of business, type of business, phone number, address, and representative name. The transaction activity data 222 is information on transaction behavior collected from the scraping server 130, and includes bank transaction details, electronic tax invoice issuance details, electronic invoice issuance details, card purchase details, card sales details, cash receipt purchase details, cash Information such as receipt sales details may be included. For example, bank transaction details may include information such as account number, transaction date, transaction time, client, recipient, brief, and transaction point, and card purchase details include number, date of use, time of use, amount of use, merchant address, It may include information such as the affiliate store name, affiliate business number, and affiliate business type. Items of information collected according to the attribute classification of each transaction activity may be different.

3 is a diagram for explaining a process of learning an artificial intelligence model using account subject classification data of the artificial intelligence model of FIG. 2 .

Referring to FIG. 3, the account subject classification data 223 of the artificial intelligence model is the learning result of the artificial intelligence model for accounting subject classification, and includes context information 350, model information 360, and model learning result during model learning. information 370.

The process of learning the artificial intelligence model using the account subject classification data of the artificial intelligence model is as follows.

First, in the process of learning the artificial intelligence model, the user's terminal 110 receives the accumulated past transaction data 310 and creates a context for learning the artificial intelligence model through the context generator 320 to classify the account subject. can create

Here, the context may include information such as user information, which is the subject of the transaction that affects the classification of the account, information on the customer who is the transaction target, the time of occurrence of the transaction, the periodicity of the time of occurrence of the transaction, and the nature of the transaction. There is, and the above information may be provided in the form of numerical values or strings or words.

In addition, the past transaction data 310 may be data transmitted from the user's terminal 110 through the web server 120 as a set of past transaction activities classified by account subject, and in the process of providing the financial management cloud service, the user's It may be accumulated data of transaction activities classified by account subject identified from the terminal 110 . In addition, the past transaction data 310 may be data accumulated from terminals 110 of other users of the same type of industry or type of business.

FIG. 4 is a diagram for explaining a process of creating a context in the context generator of FIG. 3 .

Referring to FIG. 4 , the context generator 320 may generate data for learning an artificial intelligence model for classifying account subjects. Data for learning the artificial intelligence model for classifying account subjects can be configured in the form of 'set of context + account subject' for one transaction activity.

First, the context creation unit 320 may input the received transaction behavior (S410), and business information for the corresponding transaction from the business information data 211 constituting the common data 210 for the input transaction behavior. can be searched (420). If there is a search result (Yes in S420), the context creation unit 320 may generate information such as type of business, type of business, product handled, etc. of the retrieved business information as a context (S450).

If there is no search result (No in S420), the context generator 320 goes through a process of arranging confused information based on the confusing information information data 213 constituting the common data 210 (S430). , Based on the estimated information data 212 constituting the common data 210, information capable of estimating the characteristics of the business partner is extracted (S440), and a context may be additionally created based on the extracted information (S450). .

The context generator 320 may create an account subject classification context as a set of a context through an estimated keyword, a context through a customer search result, a transaction time, and periodicity of the transaction time (S460). The context generated by the context generator 320 may be separately stored as context information 350 and managed. Here, the context information 350 may be composed of contexts generated for one transaction activity and account subjects for the transaction activity.

5 is a diagram showing the configuration of an artificial intelligence model for generating an artificial intelligence model in the artificial intelligence model generator of the account subject classification server of FIG. 3 .

Referring to FIG. 5 , the artificial intelligence model generator 330 is an input layer of the artificial intelligence model 500 for classifying account subjects based on contexts generated through the context generator 320 and learning data composed of account subjects. The size of the layer 510 and the output layer 530 may be determined, and the configuration of the hidden layer 520 may be determined according to the size of each input layer 510 and the output layer 530. When the account subject classification server 140 automatically classifies into accounting account subjects using the artificial intelligence model 500 for classifying account subjects, the accounting account subject of the transaction details corresponding to the user's fund inflow and outflow behavior is vector through neural network learning. It can be automatically classified by encoding with . At least one of a customer mutual vector, a transaction time vector, a periodicity vector, a customer type vector, and a business type vector may be used as an input vector of the artificial intelligence model 500 for classifying account subjects.

The artificial intelligence model generator 330 generates an artificial intelligence model 500 for classifying account subjects according to the determined size and configuration of each of the layers 510 and 530 and derives model information 360. can do.

As shown in FIG. 3 , the artificial intelligence model learning unit 340 may derive model learning result information 370 by learning the learning data generated by the artificial intelligence model generating unit 330 . The account subject classification data 223 of the artificial intelligence model generated through the learning process of the artificial intelligence model 500 for account subject classification may be separately managed according to the nature of the transaction for each terminal 110 of each user. For example, the nature of the transaction may include information such as card purchase, card sales, electronic tax invoice, electronic invoice, purchase of cash receipt, sales of cash receipt, bank transaction, and the like.

The learning process of the artificial intelligence model 500 for classifying account subjects is not only learned for each individual user's terminal 110, but also past transaction data of users' terminals 110 clustered by various characteristics such as industry, business type, and transaction goods It is learned based on, and according to the cluster, account subject classification data 223 of the artificial intelligence model can be generated and used to classify account subject for the transaction behavior of the terminal 110 of the new user.

6 is a diagram for explaining a process of classifying account subjects using an artificial intelligence model for classifying account subjects in the account subject classification server of FIG. 3 .

Referring to FIG. 6 , the account subject classification server 140 may classify the account subject for the transaction behavior of the user's terminal 110 collected from the scraping server 130 through the web server 120 .

Artificial intelligence model generation unit 630 is a character of the transaction behavior of the user's terminal 110 from the user data 220, for example, card purchase, card sales, electronic tax invoice, electronic invoice, cash receipt purchase, cash receipt Account subject classification data 223 of the artificial intelligence model classified by sales, bank transactions, etc. may be generated, and the learned artificial intelligence model may be loaded.

The context creation unit 620 may create a context for classifying the account subject through the process of generating the context of FIG. 4 from the transaction behavior data 610 collected from the scraping server 130, and the account classification result information 650 ) may be derived through the account classification artificial intelligence model 640 generated by the artificial intelligence model generator 630. The derived account classification result information 650 may be one or more account subjects arranged in descending order based on the probability that the transaction is a specific account subject.

7 is a diagram showing the configuration of the time series analysis server of FIG. 1 , and FIG. 8 is a diagram for explaining a process of outputting time series analysis result information from the time series analysis server of FIG. 7 .

As shown in FIGS. 7 and 8 , the time series analysis server 150 is a transaction activity 810 classified as an account subject from the account subject classification server 140, such as card purchase and sales, electronic tax accountant, electronic invoice, cash Transaction activities such as receipt purchase and sales and bank transaction details 820 are input through the input unit 710, and through the transaction activity connection unit 720, the transaction activities that are the cause of business activities and the transaction activities that are the result can be linked. It can (830). The output unit 730 may output time-series analysis result information 840 that is the result.

The time series analysis server 150 corresponds to the fund inflow and outflow behavior (transaction) that is the result of the user extracted from the scrapping server 130, which is the cause of the user automatically classified as an accounting account subject. Based on the details of deposit and withdrawal transactions in the bank account, it can be analyzed and processed as time series data. The time series analysis server 150 may calculate a cost value according to the linking process, select a minimum cost value from among the calculated cost values, and link them when linking the cause and the resultant fund flow. .

Here, the process of transaction activity linkage 830 may be performed through the processes of customer matching 831 , transaction amount matching 832 , and matching cost calculation 833 . Account matching 831 is to match whether the account of the transaction act 810 classified as the cause and the account of the resulting bank transaction details are the same account. This is a process that must be accompanied by different customers. For example, it may occur that the transaction contract is signed with company A, but the bank transaction is conducted in the name of the representative of company A or a related person.

Transaction amount matching 832 is matching based on the amount of transaction activity that is the cause and the deposit and withdrawal amount of the resulting bank transaction details. Usually, one cause may appear as multiple results, and multiple causes may result in one It is a process that is performed because it can appear as a result of

In the transaction amount matching 832 between the transaction activity that is the cause and the details of the transaction that is the result, the amount of the transaction that occurred at time t with respect to the customer j that is the cause is A(j,t), and the resulting bank If C(j', t) is the deposit and withdrawal amount of the bank transaction details that occurred at time t for the customer j' matched with the customer j in the transaction details, the matching is the sum or partial sum of the amount of the transaction Matching cases such as the sum or partial sum of deposits and withdrawals of bank transaction details can be expressed as the following [Equation 1].

[Equation 1]

Here, the matching cost calculation 833 is a method for deriving the optimal matching by calculating the cost value required for each matching when a plurality of matching occurs. ) or the cost value of dividing the act of inflow and outflow of funds (trading activity), the cost value according to the difference in time between the causative act of inflow and outflow of money (transaction activity) and the resultant bank account deposit and withdrawal transaction details, necessary for matching with business partners In the form of a sum of the cost values multiplied by each weight, it can be expressed in the form of the cost function of [Equation 2] below.

[Equation 2]

는 매칭에 필요한 비용값,

은 다수의 거래 행위를 더하거나 하나의 거래 행위를 나눔에 따른 비용값,

은 원인이 되는 거래 행위와 결과가 되는 은행 계좌의 입출금 거래 내역간의 시간상의 차이에 따른 비용값,

은 거래처 매칭에 필요한 비용값,

,

,

는 매칭에 필요한 가중치값일 수 있다.From here,

is the cost value required for matching,

is the cost of adding multiple trading activities or dividing one trading activity,

is the cost value according to the difference in time between the transaction activity that is the cause and the transaction history of the bank account that is the result,

is the cost value required for customer matching,

,

,

may be a weight value required for matching.

The time series analysis server 150 may perform matching between the causative transaction activity and the deposit and withdrawal transaction details of the bank account through a process of selecting the minimum cost result for each matching result as the optimal matching, and the artificial intelligence model However, a cost function can be used for machine learning, and other types of costs may be added to the cost item for matching.

The time series analysis server 150 may transmit the time series analysis result information 840 analyzed as time series data through a matching process between the causative transaction and the deposit and withdrawal transaction details of the bank account to the web server 120, and the web server ( 120) can provide analysis results to the user through the user's terminal 110 in an efficient manner as needed. The time series analysis server 150 may transmit the time series analysis result information 840 to the financial analysis server 160 .

The financial analysis server 160 may analyze and predict the flow of funds based on time-series analysis result information analyzed as time-series data. The financial analysis server 160 may use at least one of an autoregressive cumulative moving average model and a recurrent neural network, a long-short period model, and a gate recurrent network for analyzing time series data when analyzing and predicting a flow of funds. The financial analysis server 160 converts the time series data processed into time series data by the time series analysis server 150 into an artificial intelligence neural network model using at least one of an autoregressive cumulative moving average model, a recurrent neural network, a long-short period model, and a gate recurrent network. The flow of money can be predicted. The financial analysis server 160 may analyze the current state of funds (used funds, funds to be used, funds received, funds received, etc.) and transmit the result to the web server 120 .

9 is a flowchart illustrating a financial management cloud service providing method of the financial management cloud service providing system of the present disclosure.

Referring to FIG. 9 , the financial management cloud service providing method includes accessing the web server 120 providing the financial management cloud service (S910), automatically scraping the user's fund inflow and outflow behavior (S920), accounting account It may include automatically classifying into subjects (S930), analyzing and linking with time series data (S940), and analyzing and predicting the flow of funds (S950).

In the step of accessing the web server 120 providing the financial management cloud service, the user uses the web browser of the terminal 110 connected to the Internet such as a smartphone, PC, or tablet to the web server providing the financial management cloud service. (120) can be accessed (S910).

The web server 120 subscribes the user's terminal 110 to the financial management cloud service and receives information necessary for automatically scraping the user's transaction behavior in the scraping server 130 from the user's terminal 110, A web-based financial management cloud service can be provided with the consent of the use of information. The web server 120 may transmit the information provided by the utilization agreement from the user's terminal 110 to the scraping server 130 and store the user's information in an encrypted form in a separate storage device or DB server. The web server 120 may set information such as a period, scraping interval, and schedule for automatically scraping the transaction of the user's terminal 110, and may transmit the set information to the scraping server 130.

In the step of automatically scraping the user's fund inflow and outflow behavior, the user's fund inflow and outflow behavior (transaction behavior) of the user's terminal 110 may be automatically scraped by the scraping server 130 (S920).

The scraping server 130 uses the information provided from the user's terminal 110 to transfer the deposit and withdrawal transaction details of the user's bank account from the server 180 of the credit card company, financial institutions such as banks, and external organizations such as the National Tax Service to the web server. Scraping can be performed automatically according to the scraping period, interval, schedule, etc. set in (120), and the scraped result information can be transmitted to the web server 120. Here, the scraped result information may be transmitted to the web server 120 in an encrypted form.

The scraping server 130 transfers information such as bank transaction details, electronic tax invoice issuance details, electronic invoice issuance details, card purchase details, card sales details, cash receipt purchase details, and cash receipt sales details of the user's terminal 110 to the outside. You can scrape from your institution's server 180.

The scraping server 130 may be built in the form of a server within the service 100, and may be a service of a third party providing an external scraping service.

The web server 120 collects bank transaction details, electronic tax invoice issuance details, electronic invoice issuance details, card purchase details, card sales details, and cash receipt purchases, which are transaction activities of the user's terminal 110 received from the scraping server 130. Information such as details, sales details of cash receipts, etc. can be stored in a separate storage device and can be stored through a separate DB server. In addition, the web server 120 collects electronic tax invoice issuance details, electronic invoice issuance details, card purchase details, card sales details, cash receipt purchase details, and cash Information such as receipt sales details may be transmitted to the account subject classification server 140 for account subject classification.

In the step of automatically classifying into accounting account subjects, the account subject classification server 140 may automatically classify scrapped user fund inflow and outflow behavior (transaction behavior) into accounting account subjects using the artificial intelligence model 500 ( S930).

The account subject classification server 140 is a transaction activity of the user's terminal 110 received from the web server 120, e-tax invoice issuance details, electronic invoice issuance details, card purchase details, card sales details, cash receipt purchase details, Information such as cash receipt sales details may be classified by using the learned artificial intelligence model 500, and the result may be transmitted to the web server 120. The web server 120 transmits the account subject classification result information received from the account subject classification server 140 and the bank transaction details of the user's terminal 110 collected from the scraping server 130 to the time series analysis server 150. can Here, the transmitted data may be encrypted data.

When the account subject classification server 140 automatically classifies into accounting account subjects using the artificial intelligence model 500, the accounting account subject of the transaction details corresponding to the user's fund inflow and outflow behavior (transaction behavior) is vectored through neural network learning. It can be automatically classified by encoding with . The input vector of the artificial intelligence model 500 may use at least one of a trader mutual vector, a trade time vector, a periodicity vector, a trader type vector, and a trader industry vector.

In the step of analyzing and linking processing with time series data, by the time series analysis server 150, the user's fund flow behavior (transaction behavior) that causes the user to be automatically classified as an accounting account subject extracted from the scrapping server 130. Based on the deposit/withdrawal transaction history of the bank account corresponding to the resultant fund inflow/output activity (transaction activity), time-series data may be analyzed and linked processing (S940).

The time series analysis server 150 may calculate a cost value according to the linking process, select a minimum cost value from among the calculated cost values, and link them when linking the cause and the resultant fund flow. . The time series analysis server 150 calculates the cost value of adding or dividing the fund inflow and outflow behavior (transaction behavior), the time difference between the causative fund inflow and outflow behavior (transaction behavior) and the resultant deposit and withdrawal transaction details of the bank account The cost function of [Equation 2] can be used in the form of a sum of a cost value according to and a cost value required for business partner matching multiplied by each weight value.

The time series analysis server 150 may perform matching between the causative transaction activity and the bank account deposit/withdrawal transaction details through a process of selecting the minimum cost result for each matching result as the optimal matching.

The time series analysis server 150 may transmit the time series analysis result information 840 analyzed as time series data through a matching process between the causative transaction and the deposit and withdrawal transaction details of the bank account to the web server 120, and the web server ( 120) can provide analysis results to the user through the user's terminal 110 in an efficient manner as needed. The time series analysis server 150 may transmit the time series analysis result information 840 to the financial analysis server 160 .

In the step of analyzing and predicting the flow of funds, the flow of funds may be analyzed and predicted based on the time series analysis result information 840 analyzed as time series data by the financial analysis server 160 (S950).

The financial analysis server 160 may use at least one of an autoregressive cumulative moving average model and a recurrent neural network, a long-short period model, and a gate recurrent network for analyzing time series data when analyzing and predicting a flow of funds. The financial analysis server 160 converts the time series data processed into time series data by the time series analysis server 150 into an artificial intelligence neural network model using at least one of an autoregressive cumulative moving average model, a recurrent neural network, a long-short period model, and a gate recurrent network. The flow of money can be predicted. The financial analysis server 160 may analyze the current state of funds (used funds, funds to be used, funds received, funds received, etc.) and transmit the result to the web server 120 . The web server 120 may provide a financial analysis result to the user through the user's terminal 110 in an efficient manner as needed. The web server 120 integrates the result information processed by the scraping server 130, the account subject classification server 140, the time series analysis server 150, and the financial analysis server 160, and the web server 110 of the user's terminal It can be provided in an efficient format through a browser or application.

At least one component may be added or deleted corresponding to the performance of the components shown in FIG. 1 . In addition, it will be easily understood by those skilled in the art that the mutual positions of the components may be changed corresponding to the performance or structure of the system.

Meanwhile, the disclosed embodiments may be implemented in the form of a recording medium storing instructions executable by a computer. Instructions may be stored in the form of program codes, and when executed by a processor, create program modules to perform operations of the disclosed embodiments. The recording medium may be implemented as a computer-readable recording medium.

Computer-readable recording media include all types of recording media in which instructions that can be decoded by a computer are stored. For example, there may be read only memory (ROM), random access memory (RAM), magnetic tape, magnetic disk, flash memory, optical data storage system, and the like.

As above, the disclosed embodiments have been described with reference to the accompanying drawings. Those skilled in the art to which the present disclosure pertains will understand that the present disclosure may be implemented in a form different from the disclosed embodiments without changing the technical spirit or essential features of the present disclosure. The disclosed embodiments are illustrative and should not be construed as limiting.

100: Financial management cloud service delivery system
120: web server 130: scraping server
140: account subject classification server 150: time series analysis server
160: financial analysis server

Claims (10)

A method performed by a financial management cloud service providing system providing a financial management cloud service,
automatically scraping, by the scraping server, the user's fund inflow and outflow behavior of the user's terminal;
automatically classifying, by an account subject classification server, the scraped user fund inflow and outflow behavior into accounting account subjects using an artificial intelligence model;
Based on the deposit and withdrawal transaction details of the bank account corresponding to the fund inflow and outflow behavior of the user extracted from the scraping server, Analyzing and linking processing with time series data; and
Analyzing and predicting, by a financial analysis server, the flow of funds based on time-series analysis result information analyzed with the time-series data;
The step of analyzing and linking processing with the time series data,
By the time series analysis server, when linking the causal fund inflow and outflow behavior with the resultant fund inflow and outflow behavior, a cost value according to the linking process is calculated, and a minimum cost value is selected from among the calculated cost values to be linked. A financial management cloud service providing method comprising the steps of: According to claim 1,
The automatic classification step,
When the account subject classification server automatically classifies into accounting account subjects using the artificial intelligence model, the accounting account subject of the transaction details corresponding to the user's fund inflow and outflow behavior is encoded into a vector through neural network learning and automatically classified A financial management cloud service providing method comprising the steps of: According to claim 2,
The automatic classification step,
A method for providing a financial management cloud service, characterized in that using at least one of a customer mutual vector, a transaction time vector, a periodicity vector, a customer type vector, and a customer industry vector as an input vector of the artificial intelligence model. delete According to claim 1,
The step of analyzing and linking processing with the time series data,
By the time series analysis server, the cost value of adding or dividing the fund inflow and outflow behavior, the cost value according to the time difference between the causative fund inflow and outflow behavior and the resulting bank account deposit and withdrawal transaction details, and customer matching A method of providing a financial management cloud service, characterized in that by sequentially using and linking necessary cost values. According to claim 1,
Analyzing and predicting the flow of money,
When the financial analysis server analyzes and predicts the money flow, at least one of an autoregressive cumulative moving average model and a circular neural network, a long-short period model and a gate circular network for analyzing the time series data is used. How to provide managed cloud services. In the financial management cloud service providing system for providing a financial management cloud service,
A scraping server that automatically scrapes the user's fund inflow and outflow of the user's terminal;
an account subject classification server that automatically classifies the scraped user fund inflow and outflow behavior into accounting account subjects using an artificial intelligence model;
Based on the deposit and withdrawal transaction history of the bank account corresponding to the fund inflow and outflow behavior that is the result of the user extracted from the scraping server, the fund inflow and outflow behavior that is the cause of the user automatically classified as the accounting account subject is analyzed and linked with time series data. Time series analysis server processing; and
A financial analysis server that analyzes and predicts a flow of funds based on time-series analysis result information analyzed with the time-series data;
The time series analysis server,
Financial management characterized in that when linking the cause of the fund inflow and outflow and the resultant fund inflow and outflow, calculating the cost value according to the linking process, and selecting and linking the minimum cost value among the calculated cost values Cloud service delivery system. delete According to claim 7,
The time series analysis server,
When linking the above-mentioned causal act of fund inflow and outflow with the above-mentioned resultant act of fund inflow and outflow, the cost of adding the act of inflow and outflow of funds or dividing the act of inflow and outflow of funds, the causal act of inflow and outflow of funds and the resultant deposit and withdrawal transaction of the bank account A financial management cloud service providing system characterized by sequentially using and linking cost values according to time differences between details and cost values necessary for customer matching. A program that is combined with a computer that is hardware and stored in a computer-readable recording medium to execute the financial management cloud service providing method of any one of claims 1 to 3, 5 and 6.

Images