KR20240013034A - Machine learning based digital asset custody system, and digital asset management method thereof - Google Patents

Abstract

A machine learning-based digital asset custody system and its digital asset management method are disclosed. The digital asset management method of the digital asset custody system according to an embodiment of the present invention includes receiving a digital asset management request from a customer server, and in response to the digital asset management request, creating a digital asset wallet for digital asset management. Step, recording the results of creating a digital asset wallet on the blockchain network, providing a digital asset custody service using the digital asset wallet, based on the collected digital asset transaction data, processing of digital assets based on machine learning. Predicting price increase rate and accuracy by period; And it includes the step of operating digital assets entrusted by the customer based on machine learning-based predicted results. According to the present invention, digital assets can be entrusted and effectively stored and managed through a custody system.

Description

Machine learning-based digital asset custody system and its digital asset management method {MACHINE LEARNING BASED DIGITAL ASSET CUSTODY SYSTEM, AND DIGITAL ASSET MANAGEMENT METHOD THEREOF}

The present invention relates to a digital asset custody system and its digital asset management method. More specifically, it relates to entrusted digital assets, storing and managing digital assets, operating customer digital assets based on machine learning, and digital asset management. It relates to a machine learning-based digital asset custody system that forms a portfolio and acts as an agent for customer digital asset investment and its digital asset management method.

Recently, by using blockchain technology, it has become possible to implement a distributed ledger that allows trust between participants without an intermediary and cannot be forged or altered, and digital assets such as cryptocurrency have emerged. A new asset management service targeting these digital assets, the Digital Asset Custody Service, has emerged. Digital asset custody service is an asset storage and management service targeting blockchain-based digital assets, and refers to an infrastructure business in the digital asset market similar to a bank's custody business.

In order to stabilize and develop the digital asset market, various institutional and technological supplements are required, and in this process, custody services are expected to play the role of industrial infrastructure and gateway. In particular, through the activation of custody services, the incorporation of digital assets into the institutional system becomes easier, trust in the digital asset market can be supplemented and strengthened, technical entry barriers for general investors are resolved, and internal control over digital asset transactions is improved. It is expected that strengthened and transparent disclosure functions will be implemented.

The technical problem to be solved through embodiments of the present invention is a machine learning-based digital asset custody system and its digital asset management method that effectively stores and manages entrusted digital assets using a machine learning-based artificial intelligence model. It is provided.

In addition, the present invention utilizes various artificial intelligence models to accurately predict the price increase rate and accuracy by period for each exchange, forming a digital asset portfolio appropriate for the customer's investment tendency, and a digital asset custody system that can act on behalf of the customer's asset investment. It is intended to provide a method for managing digital assets.

In addition, one object of the present invention is to provide a digital asset custody system in which users can participate as a consensus node for verification and a digital asset management method thereof.

In addition, the purpose of the present invention is to provide a digital asset custody system and a digital asset management method thereof including an abnormal transaction detection system that can automatically determine whether a digital asset transaction is an abnormal transaction.

In addition, the present invention aims to provide a digital asset custody system and a digital asset management method thereof including an abnormal transaction detection system that can increase the detection rate for abnormal transactions through a plurality of artificial intelligence models.

In addition, the present invention provides a digital asset custody system including an abnormal transaction detection system that can deliver a suspicious transaction notification to the internal transaction analysis team when a digital asset transaction is determined to be an abnormal transaction, and a digital asset management method thereof. The purpose is to

In addition, the present invention can take action in accordance with internal regulations and related systems for transactions that are finally classified as abnormal transactions based on additional detailed analysis by the internal transaction analysis team or the final judgment of abnormal transactions by members of the internal monitoring and control organization. The purpose is to provide a digital asset custody system equipped with an abnormal transaction detection system and its digital asset management method.

In addition, the present invention provides an abnormal transaction detection system that can improve performance, such as increasing data processing speed or reducing required memory capacity, compared to conventional systems by eliminating the need to separately execute multiple programs when detecting abnormal transactions. The purpose is to provide a digital asset custody system and its digital asset management method.

In addition, the present invention is a digital asset custody system including a staking service provision system that can provide a deposit service for virtual assets traded through a verifier that can deposit only virtual assets of a pre-fixed value, and its digital asset custody system. The purpose is to provide an asset management method.

In addition, the present invention solves the problem that it is difficult to use services for virtual assets traded through a verifier that can only deposit virtual assets of a fixed value in advance unless all relatively expensive virtual assets of a fixed value are deposited. The purpose is to provide a digital asset custody system equipped with a staking service provision system and a digital asset management method thereof.

In addition, the present invention does not require separate execution of multiple programs when attempting to proceed with the deposit process for virtual assets traded through a verifier that can only deposit virtual assets of a pre-fixed value, thereby processing data compared to a conventional system. The purpose is to provide a digital asset custody system equipped with a staking service provision system that can improve performance, such as increasing speed or reducing required memory capacity, and its digital asset management method.

In addition, an object of the present invention is to provide a digital asset custody system and a digital asset management method thereof including a virtual asset custody and withdrawal payment system configured to strengthen security through double verification based on random number codes and token IDs. The purpose is to

In addition, one object of the present invention is to provide a digital asset purchase and storage system using a cash deposit/withdrawal account that is configured to strengthen security by performing double verification of the company based on company information based on a real name account. The purpose is to provide a custody system and its digital asset management method.

In addition, the technical problem that the present invention seeks to solve is to issue, store, and manage digital assets (or token securities), including coins, between companies and securities companies or exchanges by a custody company, thereby providing digital assets. This is to achieve transparency in trading and management of (or token securities).

In addition, the technical problem that the present invention seeks to solve is to manage the distribution and sale of digital assets (or token securities), including coins, by a custody company between a company and a securities company or exchange, thereby providing digital assets. This is to achieve transparency in trading and management of (or token securities).

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

A digital asset management method of a digital asset custody system according to an embodiment of the present invention to solve the above problems includes receiving a digital asset management request from a customer company server; In response to the digital asset management request, creating a digital asset wallet for digital asset management; and providing a digital asset custody service using the digital asset wallet.

In one embodiment, providing the digital asset custody service includes receiving digital assets from a customer; A step in which digital asset transaction data for each exchange is collected; Based on the collected digital asset transaction data, predicting the price increase rate and accuracy of digital assets by period based on machine learning; And it may include operating digital assets entrusted by the customer based on machine learning-based predicted results.

As an embodiment, the step of managing the entrusted digital assets may include analyzing the customer's investment propensity through a customer survey and configuring a digital asset portfolio based on the customer's investment propensity.

As an embodiment, the step of predicting the price increase rate and accuracy includes preprocessing the digital asset transaction data, extracting investment factors from the digital asset transaction data, and calculating machine learning based on the extracted investment factors. It may include performing time series data learning, extracting a prediction model based on a result of the time series data learning, and predicting investment factors of digital assets using the prediction model.

As an example, the investment factors may include forecast accuracy, price increase rate, price stability, or appropriate investment amount.

The step of predicting the price increase rate and accuracy of the digital asset by period is the step of generating a plurality of different artificial intelligence models based on different machine learning algorithms using training transaction data collected for each exchange. ; Inputting transaction data collected for each of the plurality of exchanges into the plurality of different artificial intelligence models, and obtaining the price increase rate and accuracy of digital assets by period from the output of the plurality of artificial intelligence models according to the transaction data. step; determining the weight corresponding to each artificial intelligence model based on the accuracy of each of the artificial intelligence models based on transaction data for learning; And combining the period-specific price increase rate and accuracy of digital assets output from each of the plurality of artificial intelligence models based on the weight set for each artificial intelligence model to finally predict the period-specific price increase rate and accuracy of the digital asset for each exchange. It may include steps.

The plurality of different artificial intelligence models include a light gradient boosting machine (LGBM) artificial intelligence algorithm, a recurrent neural network (RNN), a recursive recurrent artificial neural network based on long short term memory (LSTM), and a fully connected layer. Based on ANN (artificial neural network), KNN (K-Nearest Neighbors) algorithm, K-Means Clustering algorithm, PCA (Principal Component Analysis) algorithm, and support vector machine algorithm. It may include both SVM (Suport Vector Machine) artificial neural network and Auto Encoder artificial neural network.

The step of predicting the price increase rate and accuracy of the digital asset by period is the LGBM (Light Gradient Boosting Machine) artificial intelligence algorithm and the RNN (recurrent neural network), which is a recursive recurrent artificial neural network based on the Long Short Term Memory (LSTM). network), the fully connected layer-based ANN (artificial neural network), the KNN (K-Nearest Neighbors) algorithm, the K-Nearest Neighbors artificial neural network, the K-Means Clustering algorithm, the PCA ( Digital assets output from a plurality of different artificial intelligence models, including the Principal Component Analysis (SVM) algorithm, the Support Vector Machine (SVM) artificial neural network based on the support vector machine algorithm, and the Auto Encoder artificial neural network. It includes the step of finally predicting the price increase rate and accuracy by period of digital assets for each exchange by reflecting the weight determined based on the accuracy based on the training transaction data of each of the artificial intelligence models and combining them with the price increase rate and accuracy by period. can do.

According to an embodiment of the present invention, a computer-readable recording medium on which a program for executing a digital asset management method of the digital asset custody system is recorded is provided.

According to an embodiment of the present invention, a digital asset management request is received from a customer company server; In response to the digital asset management request, create a digital asset wallet for digital asset management; We provide digital asset custody services using the digital asset wallet, collect digital asset transaction data for each exchange, and predict the price increase rate and accuracy of digital assets by period based on machine learning based on the collected digital asset transaction data. A digital asset custody system is provided that operates digital assets entrusted to customers based on predicted results based on machine learning.

According to an embodiment of the present invention, digital assets can be entrusted through a custody system and effectively stored and managed, and customers' digital assets entrusted to the custody system can be operated based on machine learning.

In addition, according to an embodiment of the present invention, by accurately predicting the price increase rate and accuracy by period for each exchange using various artificial intelligence models, it is possible to construct a digital asset portfolio appropriate for the customer's investment tendency and act on behalf of the customer's asset investment.

Additionally, according to an embodiment of the present invention, a digital asset custody system and digital asset management method can be provided in which users can participate as a consensus node for verification.

In addition, according to an embodiment of the present invention, it is possible to automatically determine whether a transaction of digital assets occurring in a custody system configured to provide a service for managing assets is an abnormal transaction.

Additionally, according to an embodiment of the present invention, the detection rate for abnormal transactions can be increased through a plurality of artificial intelligence models.

Additionally, according to an embodiment of the present invention, if a digital asset transaction is determined to be an abnormal transaction, a notification of a suspicious transaction may be sent to the internal transaction analysis team.

In addition, according to an embodiment of the present invention, transactions that are finally classified as abnormal transactions based on additional detailed analysis by the internal transaction analysis team or final judgment of abnormal transactions by members of the internal monitoring and control organization are subject to internal regulations and related systems. Action can be taken.

In addition, according to an embodiment of the present invention, when detecting abnormal transactions, there is no need to separately execute multiple programs, thereby improving performance, such as increasing data processing speed or reducing required memory capacity, compared to conventional systems. .

Additionally, according to an embodiment of the present invention, it is possible to provide a deposit service for virtual assets traded through a verifier that can deposit only virtual assets of a pre-fixed value.

In addition, according to an embodiment of the present invention, unless all relatively expensive virtual assets of a fixed value are deposited, services for virtual assets traded through a verifier that can deposit only virtual assets of a fixed value in advance are used. This difficult problem can be solved.

In addition, according to an embodiment of the present invention, when attempting to proceed with the deposit process for virtual assets traded through a verifier that can only deposit virtual assets of a pre-fixed value, there is no need to separately execute multiple programs, thus eliminating the need for conventional Compared to the system, performance can be improved by increasing data processing speed or reducing required memory capacity.

In addition, according to an embodiment of the present invention, a digital asset custody system including a virtual asset custody and withdrawal payment system with enhanced security through double verification based on a random number code and token ID and a digital asset management method thereof are provided. It can be.

In addition, according to an embodiment of the present invention, a digital asset custody system comprising a virtual asset purchase and storage system using a cash deposit/withdrawal account with enhanced security by performing double verification of the company's assets through two real-name accounts and its digital asset management method may be provided.

In addition, according to an embodiment of the present invention, a custody company performs the issuance, storage, and management of digital assets, including coins, between a company and a securities company or exchange, thereby enabling transactions and transactions of digital assets (or token securities). Management transparency can be achieved.

In addition, according to an embodiment of the present invention, management related to the distribution and sale of digital assets (or token securities), including coins, is performed by a custody company between a company and a securities company or exchange, thereby enabling trading and trading of digital assets. Management transparency can be achieved.

Figure 1 is a configuration diagram of a token withdrawal transaction fee payment system using a smart contract constituting a digital asset custody system according to an embodiment of the present invention.
Figure 2 is a flow chart of a token withdrawal transaction fee payment method using a smart contract performed by a digital asset custody system according to an embodiment of the present invention.
FIG. 3 is a conceptual diagram for explaining step S100 of FIG. 2.
FIG. 4 is a conceptual diagram for explaining step S200 of FIG. 2.
FIG. 5 is a conceptual diagram for explaining step S300 of FIG. 2.
FIG. 6 is a conceptual diagram for explaining step S400 of FIG. 2.
Figure 7 is a configuration diagram of a trustee terminal constituting a digital asset custody system according to an embodiment of the present invention.
Figure 8 is a conceptual diagram of a digital asset custody system according to an embodiment of the present invention.
Figure 9 is a flowchart showing the process of creating an applicant account and an approver account in a digital asset management method according to an embodiment of the present invention.
Figure 10 is a flowchart showing the digital asset management application and approval process of the digital asset management method according to an embodiment of the present invention.
Figure 11 is a flowchart showing the digital asset management application and approval process of the digital asset management method according to another embodiment of the present invention.
Figure 12 is a diagram showing the overall outline of the custody system according to an embodiment of the present invention.
Figure 13 is a flowchart showing a method of managing the assets of a digital asset custody customer based on machine learning, according to an embodiment of the present invention.
Figure 14 is a diagram conceptually showing the steps in which digital assets are entrusted to a customer.
15 is a diagram illustrating an exemplary form of digital asset transaction data collected from an exchange.
FIG. 16 is a flowchart illustrating an embodiment of step S330 of FIG. 13 in more detail.
Figure 17 is a diagram conceptually showing the steps of analyzing investment propensity through customer surveys.
Figure 18 is a diagram conceptually showing the steps of configuring an appropriate portfolio based on the customer's investment tendency and acting as an agent for customer asset investment.
Figure 19 is a configuration diagram of a digital asset investment agency analysis system that constitutes a digital asset custody system according to an embodiment of the present invention.
Figure 20 is a conceptual diagram of an artificial intelligence unit constituting a digital asset custody system according to an embodiment of the present invention.
Figure 21 is a diagram for explaining an example of implementing a custody system according to an embodiment of the present invention with a computing device.
Figure 22 is a block diagram of a digital asset custody service providing system according to an embodiment of the present invention.
Figure 23 is a flowchart of a verification method of a digital asset custody service according to an embodiment of the present invention.
Figure 24 is a flowchart of a method for disclosing transaction records of a digital asset custody service according to an embodiment of the present invention.
Figure 25 is an example diagram to explain a method of recording transaction data using a blockchain network.
Figure 26 is an example diagram to explain a method of disclosing transaction records.
Figure 27 is a block diagram of a digital asset custody service providing system according to another embodiment of the present invention.
Figure 28 is a flowchart of a method of providing a digital asset custody service that provides a premium service using a non-fungible token according to an embodiment of the present invention.
Figure 29 is an example diagram to explain the issuance of a non-fungible token containing the right to use a service.
Figure 30 is a configuration diagram of an abnormal transaction detection system constituting a digital asset custody system according to an embodiment of the present invention.
Figure 31 is a table showing deposit and withdrawal data of customers using a digital asset (virtual asset) custody service according to an embodiment of the present invention.
Figure 32 is a diagram for explaining a machine learning-based abnormal transaction detection method performed by a digital asset custody system according to an embodiment of the present invention.
Figure 33 is a flowchart of a control method of an abnormal transaction detection system constituting a digital asset custody system according to an embodiment of the present invention.
Figure 34 is a flowchart of a method of taking action against a consignor terminal determined to be an abnormal transaction according to an embodiment of the present invention.
Figure 35 is a diagram for explaining asset staking in a staking service providing system that constitutes a digital asset custody system according to an embodiment of the present invention.
Figure 36 is a diagram illustrating a staking service providing system constituting a digital asset custody system according to an embodiment of the present invention.
FIG. 37 is a diagram illustrating generating verifier deposit data and registering a verifier by a digital asset custody system according to an embodiment of the present invention.
Figure 38 is a diagram for explaining staking to a payment account by a digital asset custody system according to an embodiment of the present invention.
Figure 39 is a diagram for explaining the transmission of assets to a customer by unstaking by a digital asset custody system according to an embodiment of the present invention.
Figure 40 is a flowchart of a control method of a staking service providing system constituting a digital asset custody system according to an embodiment of the present invention.
Figure 41 is a flowchart showing some examples of a virtual asset custody method based on a digital asset custody system according to an embodiment of the present invention.
Figure 42 is a flowchart showing another example of a virtual asset custody method based on a digital asset custody system according to an embodiment of the present invention.
Figure 43 is a flowchart showing an example of a virtual asset withdrawal method based on a digital asset custody system according to an embodiment of the present invention.
Figure 44 is a flowchart showing an example of an account creation method for purchasing and storing virtual assets performed by a digital asset custody system according to an embodiment of the present invention.
Figure 45 is a flowchart showing an example of a method of purchasing a virtual asset using a cash deposit/withdrawal account performed by a digital asset custody system according to an embodiment of the present invention.
Figure 46 is a flowchart showing an example of a method of selling virtual assets using a cash deposit/withdrawal account performed by a digital asset custody system according to an embodiment of the present invention.
Figure 47 is a configuration diagram of a digital asset (token security) issuance management system that constitutes a digital asset custody system according to an embodiment of the present invention.
Figure 48 is a flowchart of a control method of a digital asset (token security) issuance management system that constitutes a digital asset custody system according to an embodiment of the present invention.
FIG. 49 is a conceptual diagram for explaining steps (S8100, S8200) of FIG. 48.
Figure 50 is a conceptual diagram for explaining steps (S8200, S8300) of Figure 48.
FIG. 51 is a conceptual diagram for explaining step S8400 of FIG. 48.
Figure 52 is a configuration diagram of a digital asset (token security) distribution and trading management system that constitutes a digital asset custody system according to an embodiment of the present invention.
Figure 53 is a flowchart of a control method of a digital asset (token security) distribution and trading management system performed by a digital asset custody system according to an embodiment of the present invention.
FIG. 54 is a conceptual diagram for explaining steps (S9100 to S9400) of FIG. 53.
Figure 55 is a conceptual diagram to explain the process of verifying instructions and purchasing or selling digital assets (or token securities) by the digital asset custody system according to an embodiment of the present invention.

The advantages and features of the present invention and methods for achieving them will become clear by referring to the embodiments described in detail below along with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various different forms. The present embodiments are merely provided to ensure that the disclosure of the present invention is complete, and that the present invention is not limited to the embodiments disclosed below and may be implemented in various forms, and the present invention is not limited to the embodiments disclosed below. It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout the specification.

In this specification, when a part 'includes' a certain component, this means that it may further include other components rather than excluding other components unless specifically stated to the contrary. As used herein, '~unit' refers to a unit that processes at least one function or operation, and may mean, for example, software, FPGA, or hardware components. The functions provided in '~ part' may be performed separately by multiple components, or may be integrated with other additional components. '~ part' in this specification is not necessarily limited to software or hardware, and may be configured to be in an addressable storage medium, or may be configured to reproduce one or more processors. Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

The embodiments described in this specification are intended to clearly explain the spirit of the present invention to those skilled in the art to which the present invention pertains, and therefore the present invention is not limited to the embodiments described in this specification, and the present invention is not limited to the embodiments described in this specification. The scope should be construed to include modifications or variations that do not depart from the spirit of the present invention.

The terms used in this specification are general terms that are currently widely used as much as possible in consideration of their function in the present invention, but this may vary depending on the intention of those skilled in the art, precedents, or the emergence of new technology in the technical field to which the present invention pertains. You can. However, if a specific term is defined and used with an arbitrary meaning, the meaning of the term will be described separately. Therefore, the terms used in this specification should be interpreted based on the actual meaning of the term and the overall content of this specification, not just the name of the term.

The drawings attached to this specification are intended to easily explain the present invention, and the shapes shown in the drawings may be exaggerated as necessary to aid understanding of the present invention, so the present invention is not limited by the drawings. In this specification, if it is determined that a detailed description of a known configuration or function related to the present invention may obscure the gist of the present invention, the detailed description thereof will be omitted as necessary.

Figure 1 is a configuration diagram of a token withdrawal transaction fee payment system using a smart contract constituting a digital asset custody system according to an embodiment of the present invention. Referring to FIG. 1, the digital asset custody system 10 according to an embodiment of the present invention may include one or more trustee terminals 100, one or more trustee terminals 200, and a trustee DB 300.

The trustor terminal 100 may be a terminal related to a trustor that entrusts the storage and management of digital assets to the trustor terminal 200. For example, the consignor terminal 100 may be a terminal belonging to an organization (consignor) such as a company, institution, or organization that holds digital assets.

The consignor terminal 100 may be provided as a terminal such as a desktop PC, laptop, notebook, smartphone, or smart pad, but is not limited thereto. The consignor terminal 100 may include one or more manager terminals 110, one or more applicant terminals 120, and one or more approver terminals 130.

The manager terminal 110, the applicant terminal 120, the approver terminal 130, and the trustee terminal 200 are connected through a wired/wireless communication network and can communicate with each other. The manager terminal 110 may be a terminal that requests the trustee terminal 200 for an owner account, an applicant account, and an approver account to entrust the storage and management of digital assets held by a trustee such as a company to the trustee.

The applicant terminal 120 accesses the applicant account issued by the trustee terminal 200 in accordance with the digital asset storage and management request of the manager terminal 110 and applies for digital asset management such as storage, transfer, and withdrawal of digital assets. It may be a terminal used by the applicant.

In an embodiment of the present invention, the digital asset that the consignor entrusts to the trustee may include one or more virtual currencies selected from virtual currencies including Bitcoin, Ethereum, and Klaytn.

The approver terminal 130 may be a terminal used by an approver who accesses the approver account issued by the trustee terminal 200 and approves the digital asset management application applied by the applicant terminal 120.

The trustee terminal 200 is entrusted with the storage and management of digital assets from the trustee terminal 100 and can provide a custody service that stores and manages digital assets requested from the trustee terminal 100.

The trustee database (DB) 300 contains digital asset information entrusted from the consignor terminal 100, owner account information, applicant account information, approver account information, and digital asset wallet information (corresponding to a random number) storing the entrusted digital assets. Wallet address), etc. can be saved.

Figure 2 is a flowchart of a token withdrawal transaction fee payment method using a smart contract performed by a digital asset custody system according to an embodiment of the present invention. FIG. 3 is a conceptual diagram for explaining step S100 of FIG. 2. Referring to Figures 2 and 3, first, a smart contract 400 related to a digital asset custody system (hereinafter, may be abbreviated as 'custody system') corresponding to a trustee terminal is sent from one or more trustee terminals 100. Each is delegated authority to allow withdrawal of digital assets and approves withdrawal of tokens (ERC-20, ERC-721, ERC-1155, etc.) related to digital assets (e.g., virtual currencies such as Bitcoin, Ethereum, Klaytn, etc.) A receiving step may be performed (S100). At this time, the consignor can directly pay the transaction fee to the smart contract 400 through the consignor terminal 100.

FIG. 4 is a conceptual diagram for explaining step S200 of FIG. 2. Referring to FIGS. 2 and 4, when the custody system 10 receives a request for withdrawal of digital assets from the consignor terminal 100, the custody system 10 performs withdrawal-related processing of digital assets by the verification logic 262 of the custody system 10. Data can be verified (S200).

FIG. 5 is a conceptual diagram for explaining step S300 of FIG. 2. Referring to Figures 2 and 5, when the data related to the withdrawal of digital assets is verified, the custody system 10 processes the transaction fee related to the transaction of the digital asset through the payment account 12 of the custody system 10. By withdrawing the fee, the transaction fee can be paid to the smart contract 400 instead of the consignor terminal (S300).

FIG. 6 is a conceptual diagram for explaining step S400 of FIG. 2. Referring to FIGS. 2 and 6, the smart contract 400 receives transaction fees from the custody system 10 through the payment account 12 and delivers tokens related to the digital assets requested for withdrawal by the trustor terminal 100. Upon approval, the digital asset can be transferred to the recipient's account related to the recipient's terminal (500) (S400).

According to an embodiment of the present invention, from the perspective of a customer corresponding to a consignor terminal, digital assets can be delivered to the recipient more conveniently without having to consider the occurrence of transaction fees that vary depending on the type of virtual asset and transaction situation. In addition, from the custodian's perspective, they can attract customers by paying transaction fees, and can also receive a certain fee from customers or earn additional income such as advertising due to the increase in customers.

Figure 7 is a configuration diagram of a trustee terminal constituting a digital asset custody system according to an embodiment of the present invention. Figure 8 is a conceptual diagram of a digital asset custody system according to an embodiment of the present invention.

Referring to FIGS. 1, 7, and 8, the trustee terminal 200 includes an owner account creation unit 210, a digital asset wallet register 220, an applicant account creation unit 230, an approver account creation unit 240, It may include an approval request unit 250, an account verification unit 260, a digital asset verification unit 270, a digital asset management unit 280, and a control unit 290.

The owner account creation unit 210 may create an owner account in response to a request for creating an owner account from the consignor terminal 100 and share it with the consignor terminal 100 . The owner account is an account granted to the consignor, and the owner account created by the owner account creation unit 210 may be recorded in the trustee DB 300.

The digital asset wallet register 220 can register a digital asset wallet (multisig wallet) for storing digital assets requested from the trustor terminal 100. The digital asset wallet register 220 can register one or more digital asset wallets (multisig wallets) for each plurality of virtual currencies.

The digital asset wallet register 220 stores the first digital asset wallet (first multi-sig wallet) requested for registration by the trustor terminal 100 and the second digital asset wallet (second multi-sig wallet) verified through the digital asset history. Multiple digital asset wallets (multi-sig wallets) can be registered as whitelist digital asset wallets.

A digital asset wallet (multisig wallet) that can be selected by the applicant account among a plurality of digital asset wallets (multisig wallets) registered as whitelist digital asset wallets can be set individually for each applicant account.

The applicant account creation unit 230 may create an applicant account according to a request for creating an applicant account from the consignor terminal 100 related to the owner account and share it with the consignor terminal 100.

In an embodiment, the applicant account creation unit 230 may create an applicant account for each of a plurality of virtual currency and digital asset wallets. The applicant account created by the applicant account creation unit 230 may be recorded in the trustee DB 300.

The approver account creation unit 240 may create a plurality of approver accounts according to a request for creating an approver account from the consignor terminal 100 related to the owner account and share them with the consignor terminal 100 .

The approver account creation unit 240 may create a plurality of approver accounts for each virtual currency and digital asset wallet. A plurality of approver accounts created by the approver account creation unit 240 may be recorded in the trustee DB 300.

The approval request unit 250 may receive a digital asset management application through an applicant account and request approval for the digital asset management application from multiple approver accounts.

The account verification unit 260 may perform verification of the digital asset management application depending on whether the digital asset management application is approved by a plurality of approver accounts.

The digital asset confirmation unit 270 can check digital assets stored in a digital asset wallet (multisig wallet) related to the digital asset management application.

The digital asset management unit 280 may perform digital asset management according to the digital asset management application upon completion of verification and confirmation of the account verification unit 260 and the digital asset verification unit 270.

The control unit 290 may control each component of the trustee terminal 200 to provide a service for managing digital assets of the trustee terminal.

According to the embodiment shown in FIGS. 7 and 8, the risk of theft and hacking of digital assets can be reduced by distributing accounts for management of digital assets into applicant accounts and approver accounts.

The approver terminal 130 may be given one approver account for each owner account, but a plurality of approver accounts may be given to the owner account. Additionally, the approver account may be set differently depending on the applicant account.

When multiple approver accounts are granted to the owner account, when the applicant terminal 120 of the consignor terminal 100 accesses the applicant account and applies for storage and management of digital assets, the trustee terminal 200 You can decide which of the approver accounts to request approval for storage and management of digital assets.

In one embodiment, the trustee terminal 200 selects all of the plurality of approver accounts, or randomly selects two or more approver accounts from the plurality of approver accounts, and uses the selected plurality of approver accounts to store and manage digital assets. You can request approval.

The trustee terminal 200 receives a digital asset management application (withdrawal application) from the withdrawal applicant through the applicant account by the approval request unit 250 and requests approval for the digital asset management application (withdrawal application) from a plurality of approver accounts. (S1).

The trustee terminal 200 can perform verification of the digital asset management application depending on whether the digital asset management application is approved by a plurality of approver accounts (withdrawal approval authority accounts) by the account verification unit 260 (S2). .

The distributed storage unit of the trustee DB (300) can distribute and store a number of distributed keys (20) using Shamir's Key Distribution method to sign the multi-sig wallet (40) used for digital asset management. The plurality of distribution keys 20 may be distributed and stored in the database or storage of the trustee DB 300, respectively.

The account verification unit 260 of the trustee terminal 200 may approve a digital asset management application upon completion of approval from a set number of approver accounts among a plurality of approver accounts (S2). In an embodiment, the account verification unit 260 may approve a digital asset management application upon completion of approval from more than half of the approver accounts among the plurality of approver accounts.

The digital asset management unit 280 of the trustee terminal 200 may generate a master key 30 by combining at least two of the plurality of distribution keys 20 (S3). In an embodiment, the digital asset management unit 280 may generate the master key 30 by combining more than half of the distribution keys 20 among the plurality of distribution keys 20.

Multiple distribution keys 20 may be set to correspond to multiple approver accounts, but may also be distributed and stored independently of the approver accounts. When the multiple distribution keys 20 are set to correspond to multiple approver accounts, the master key 30 can be generated by the distributed keys corresponding to the approver accounts that approved the digital asset management application among the multiple approver accounts. there is.

When multiple distribution keys (20) are set regardless of the approver's account, two or more distribution keys among the multiple distribution keys (20) are selected and combined according to preset rules or at random to generate the master key (30), or the entire distribution key (30) is set. The master key 30 can also be generated by combining all of the distributed keys 20.

Finally, upon completion of verification by the account verification unit 260, the trustee terminal 200 can perform digital asset management (withdrawal processing, etc.) according to the digital asset management application by the digital asset management unit 280.

The digital asset management unit 280 of the trustee terminal 200 performs digital asset management (withdrawal processing) by signing the multisig wallet 40 with the master key 30 generated by combining a plurality of distribution keys 20. (S4, S5).

For example, when three or more approver accounts are set for an owner account, the trustee terminal 200 randomly selects two or more approver accounts among the three or more approver accounts and stores digital assets with the selected approver accounts. Approval for management may be requested.

Applicants cannot know which of three or more approvers will be requested to approve storage and management of digital assets. Therefore, since it is difficult for an applicant to fraudulently mock with all three or more approvers, theft of digital assets through prior mocking can be prevented.

Additionally, the trustee terminal 200 may determine the number of approver accounts to request approval among a plurality of approver accounts based on application information for storage and management of digital assets received through the applicant account.

For example, if the size of the digital asset requested to be transferred or withdrawn exceeds the preset standard asset size, the trustee terminal 200 requests approval from two or three or more approver accounts and approval is obtained through all approver accounts. Processing such as transfer or withdrawal of the corresponding digital asset can only be performed.

On the other hand, if the size of the digital asset requested to be transferred or withdrawn from the applicant terminal 120 is less than or equal to the preset standard asset size, the trustee terminal 200 requests approval from one approver account and approves it through that one approver account. Once this happens, processing such as transfer or withdrawal of digital assets can be performed.

Figure 9 is a flowchart showing the process of creating an applicant account and an approver account in a digital asset management method according to an embodiment of the present invention. Referring to FIGS. 1, 7, 8, and 9, when the creation of an owner account is requested from the consignor terminal 100 (S11), the owner account creation unit 210 of the consignor terminal 200 creates an owner account to be issued to the consignor. You can create an account (S12). The owner account information generated by the owner account creation unit 210 may be shared with the consignor terminal 100 and recorded in the consignee DB 300 (S13, S14).

When creating an owner account, the digital asset wallet register 220 can register a digital asset wallet (multisig wallet) for storing the digital assets requested from the trustor terminal 100. The digital asset wallet register 220 can register one or more digital asset wallets (multisig wallets) for each plurality of virtual currencies.

The digital asset wallet register 220 includes a first digital asset wallet (first multi-sig wallet) requested for registration by the trustor terminal and a second digital asset wallet (second multi-sig wallet) verified through digital asset history. Digital asset wallets (multisig wallets) can be registered as whitelisted digital asset wallets.

A digital asset wallet (multisig wallet) that can be selected by the applicant account among a plurality of digital asset wallets (multisig wallets) registered as whitelist digital asset wallets can be set for each applicant account.

When the creation of an applicant account and an approver account is requested from the consignor terminal 100 (S15), the applicant account creation unit 230 and the approver account creation unit 240 of the consignor terminal 200 generate the consignor terminal 100 related to the owner account. )'s request for creating an applicant account and an approver account can be created (S16)

At this time, the applicant account creation unit 230 and the approver account creation unit 240 may create an applicant account and an approver account for each of a plurality of virtual currency and digital asset wallets, respectively. The applicant account and the approver account created by the applicant account creation unit 230 and the approver account creation unit 240 may be shared with the consignor terminal 100 and recorded in the trustee DB 300 (S17, S18).

Figure 10 is a flowchart showing the digital asset management application and approval process of the digital asset management method according to an embodiment of the present invention. Referring to FIGS. 1, 7, and 10, the approval request unit 250 may receive a digital asset management application through the applicant account and request approval for the digital asset management application from the approver account (S21).

When applying for digital asset management, the applicant can access the applicant account and receive approval for application through two or more authentications. Security can be improved through such multi-factor authentication. Multi-factor authentication can be determined by two or more of various authentication methods such as, for example, password authentication, ID authentication, facial recognition, and biometric recognition (fingerprint recognition). Meanwhile, the number or type of authentication methods corresponding to multifactor authentication may be set differently for each applicant account.

When the digital asset management application is approved through the approver account (S22), the account verification unit 260 performs verification of the digital asset management application depending on whether the digital asset management application is approved by the approver account and verifies the digital asset. Unit 270 can check the digital assets stored in the digital asset wallet related to the digital asset management application (S23).

Multiple approvers who have been requested to approve digital asset management can each access the approver account and approve digital asset management through two or more authentications. Security can be improved through such multi-factor authentication. Multi-factor authentication can be determined by two or more of various authentication methods, for example, password authentication, ID authentication, facial recognition, and biometric recognition (fingerprint recognition). Meanwhile, the number or type of authentication methods corresponding to multifactor authentication may be set differently for each approver account.

Upon completion of verification and confirmation by the account verification unit 260 and the digital asset verification unit 270, the digital asset management unit 280 performs digital asset management according to the digital asset management application, and stores information related to digital asset management in the trustee database. It can be recorded at (300) (S24, S25).

According to the embodiment of the present invention described above, the account for management of digital assets is distributed into an applicant account and an approver account, and the customer company's key information is distributed and stored into a plurality of distributed keys using Shamir's Key Distribution method to be distributed to the customer company. It can prevent internal misconduct and reduce the risk of digital asset theft and hacking.

Figure 11 is a flowchart showing the digital asset management application and approval process of the digital asset management method performed by the digital asset custody system according to an embodiment of the present invention. Referring to FIGS. 1 and 11 , the trustee terminal 200 may grant a plurality of approver accounts to each applicant account for the owner account (S32).

When the applicant terminal 120 of the consignor terminal 100 accesses the applicant account and applies for storage and management of digital assets, the trustee terminal 200 stores and manages the digital assets using one of the plurality of approver accounts. You can decide whether to request approval for etc.

In one embodiment, the trustee terminal 200 may determine the number of approver accounts to request approval among a plurality of approver accounts based on application information for storage and management of digital assets received through the applicant account (S34).

The trustee terminal 200 may randomly select one or more approver accounts from among a plurality of approver accounts as many as the number of approver accounts and request approval for storage and management of digital assets (S36).

If the size of the digital asset requested to be transferred or withdrawn exceeds the preset standard asset size, the trustee terminal 200 requests approval from two or three or more approver accounts and transfers the digital asset only if approval is obtained through all approver accounts. Processing such as transfer or withdrawal can be performed.

On the other hand, if the size of the digital asset requested to be transferred or withdrawn from the applicant terminal 120 is less than or equal to the preset standard asset size, the trustee terminal 200 requests approval from one approver account and approves it through that one approver account. Once this happens, processing such as transfer or withdrawal of digital assets can be performed.

For example, if the trustee terminal 200 determines to randomly select two approver accounts based on the size of the digital asset requested by the applicant account, it randomly selects two approver accounts among the three approver accounts and Approval for storage and management of digital assets can be requested through the approver account (S38).

Applicants cannot know which of the multiple approvers will be requested to approve storage and management of digital assets. Therefore, since it is difficult for the applicant to fraudulently mock all approvers, it is possible to prevent theft of digital assets from companies, etc. through prior mocking.

Figure 12 is a diagram showing an overview of the entire system environment surrounding the custody system according to an embodiment of the present invention. Referring to FIG. 12, the overall system environment may include a digital asset custody system, a customer server surrounding it, and a blockchain network.

First, a request for digital asset storage and withdrawal is made through the customer's server. The generated request is received by the web server. The web server requests the customer's identity (KYC, Know Your Customer) from the administrator server (Admin), and if there is no problem with the customer's identity, it sends a wallet creation request to the wallet operation server.

The wallet operation server creates a wallet for digital assets (or digital assets) in conjunction with the key storage system, and the public and private keys created during the wallet creation process are stored separately in the wallet operation server and the cold wallet room. As an example, a security policy in the ISMS may be applied to a cold wallet room to which the key storage system is accessible. In addition, the digital asset wallet creation results are linked to the mainnet and recorded and stored in the blockchain network.

Hereinafter, a digital asset management method using the digital asset custody system of FIG. 12 will be described with reference to various embodiments. Figure 13 is a flowchart showing a method of managing the assets of a digital asset custody customer based on machine learning, according to an embodiment of the present invention. The method of FIG. 13 may be performed by the custody system described in FIGS. 1 and 12. Therefore, if the performing entity is not specified in the following steps, it is assumed that the performing entity is the custody system.

In step S310, digital assets are entrusted to the customer. Hereinafter, digital assets may be used as terms such as virtual assets or customer assets. In step S320, digital asset transaction data for each exchange is collected. At the S330 stage, the price increase rate and accuracy of digital assets by period are predicted based on machine learning. In step S340, the customer's investment propensity is analyzed through a customer survey. At the S350 stage, a portfolio suitable for the customer's investment tendency is formed and investment agency of the customer's assets is performed.

Figure 14 is a diagram conceptually showing the steps in which digital assets are entrusted to a customer. To operate a digital asset custody service, the company operating the custody system entrusts digital assets from multiple customers. 15 is a diagram illustrating an exemplary form of digital asset transaction data collected from an exchange. Referring to Figure 15, digital asset transaction data may be collected daily, hourly, and minutely for each digital asset. As an example, the digital asset transaction data may include the digital asset transaction date and time, transaction start price, transaction end price, highest transaction price, lowest transaction price, and/or transaction volume.

FIG. 16 is a flowchart illustrating an embodiment of step S330 of FIG. 13 in more detail. In Figure 16, an embodiment of extracting investment factors such as prediction accuracy, growth rate, stability, and appropriate investment amount required for portfolio construction from digital asset transaction data and learning time series data using machine learning techniques to extract a prediction model is explained.

In step S331, digital asset transaction data is obtained. In step S332, digital asset transaction data is preprocessed. In step S333, investment factors are extracted from digital asset transaction data. As an example, the extracted investment factors may include forecast accuracy, price increase rate, price stability, or appropriate investment amount.

In step S334, machine learning-based time series data learning is performed based on the extracted investment factors. At this time, LGBM or LSTM can be used as the machine learning algorithm. In step S335, a prediction model is extracted based on the learning performance results. In step S336, investment factors for digital assets are predicted using the extracted prediction model.

Figure 17 is a diagram conceptually showing the steps of analyzing investment propensity through customer surveys. In Figure 17, the customer's investment propensity is classified into various types such as stable, neutral, and aggressive investment type through a customer survey, and the final investment propensity is determined by reflecting various factors such as the customer's age, income, investment experience and knowledge, and desired return. This resulting example is shown.

Figure 18 is a diagram conceptually showing the steps of configuring an appropriate portfolio based on the customer's investment tendency and acting as an agent for customer asset investment. In Figure 18, an embodiment is shown in which a digital asset portfolio is constructed by adjusting the rate of return and degree of diversification to suit the investment tendency based on the customer's investment tendency calculated previously, and digital asset investment is performed accordingly.

Referring to the example in Figure 18, if the customer's investment tendency is stability-seeking, the digital asset portfolio is composed of multiple digital assets of SOL, ETH, BTC, DOGE, and BORA based on that, and the investment proportion of each is 11%. Examples of settings such as , 22%, 51%, 9%, and 7% are shown. According to the embodiments of the present invention described above, digital assets can be entrusted and effectively stored and managed through a custody system.

Figure 19 is a configuration diagram of a digital asset investment agency analysis system that constitutes a digital asset custody system according to an embodiment of the present invention. Referring to FIG. 19, the digital asset custody system according to an embodiment of the present invention may include a digital asset investment agency analysis system 1000. The digital asset investment agency analysis system 1000 may include a transaction data analysis unit 1100, a customer investment propensity analysis unit 1200, and a digital asset investment agency unit 1300.

The digital asset investment agency analysis system (1000) collects digital asset transaction data for each exchange, predicts the price increase rate and accuracy of digital assets by period based on machine learning, analyzes customers' investment tendencies, and provides information tailored to the customer's investment tendencies. It may be a system that provides a service that constructs a portfolio and acts as an agent for customer digital asset investment. The digital asset investment agency analysis system 1000 may be provided on a server of a digital asset custody service provider, or may be implemented to link with the server of a digital asset custody service provider. Each component of the digital asset investment agency analysis system 1000 may be implemented with at least one processor.

The transaction data analysis unit 1100 can collect digital asset transaction data for each exchange and predict the price increase rate and accuracy of digital assets by period based on machine learning. The transaction data analysis unit 1100 analyzes transaction data collected for each exchange using a plurality of artificial intelligence models of the transaction data collection unit 1110 and the artificial intelligence unit 1140, which collects transaction data for each exchange. An analysis unit 1120, a machine learning unit 1130 that performs learning on each artificial intelligence model of the artificial intelligence unit 1140, and a plurality of different artificial intelligence models learned by the machine learning unit 1130. It may include an intelligence unit 1140.

The transaction data collection unit 1110 may be configured to collect transaction data related to digital assets from various exchange terminals. The transaction data collection unit 1110 may collect transaction data of digital assets for each exchange at a set period (for example, 1 day). The transaction data collection unit 1110 can transmit digital asset transaction data for each exchange to the exchange transaction data analysis unit 1120.

The exchange transaction data analysis unit 1120 can predict the price increase rate and accuracy of digital assets for each exchange using a plurality of different artificial intelligence models of the artificial intelligence unit 1140 based on digital asset transaction data for each exchange. That is, the exchange transaction data analysis unit 1120 inputs the information contained in the transaction data for each exchange into each artificial intelligence model, combines the information output from each artificial intelligence model, and calculates the price increase rate of digital assets for each exchange based on this. Accuracy can be predicted.

The machine learning unit 1130 is based on learning data including labeled transaction data for each exchange and digital asset price trend information, and sets the information in the transaction data as input variables and the price increase rate as the output variable to perform a machine learning method. You can learn artificial intelligence models through this. Each artificial intelligence model of the artificial intelligence unit 1140 learned by the machine learning unit 1130 can be used to predict the price increase rate and accuracy of digital assets based on subsequent transaction data.

Machine learning can mean using a model composed of multiple parameters and optimizing the parameters with given data. Machine learning can include supervised learning, unsupervised learning, and reinforcement learning, depending on the type of learning problem. Supervised learning is learning the mapping between input and output, and can be applied when input and output pairs are given as data. Unsupervised learning is applied when there is only input and no output, and can find regularities between inputs.

The machine learning unit 1130 can create artificial intelligence models in various ways. For example, the machine learning unit 1130 can learn features extracted from training data using a deep learning-based learning method. At this time, learning data is extracted from fields of past actual transaction data set for each exchange, or image analysis-based (e.g., CNN with several stages of convolutional layers) from statistical data, trend data, trend graphs, etc. of transaction data for each exchange. Various methods, such as extracting features using (Convolutional Neural Networks) structures, etc., can be used, but are not limited to this and methods not mentioned may also be used.

The learning method of the machine learning unit (1130) is a light gradient boosting machine (LGBM) artificial intelligence algorithm, a recurrent neural network (RNN) based on long short term memory (LSTM), and a fully connected layer. Based on ANN (artificial neural network), KNN (K-Nearest Neighbors), K-Nearest Neighbors, K-Means Clustering algorithm, PCA (Principal Component Analysis) algorithm, and support vector machine algorithm. This may be done through a machine learning algorithm including a Support Vector Machine (SVM) artificial neural network and/or an Auto Encoder artificial neural network.

The customer investment propensity analysis unit 1200 may analyze the customer's investment propensity through customer response data to a customer survey. The customer investment propensity analysis department (1200) reflects various factors such as the customer's age, income, investment experience and knowledge, and desired return through customer surveys and classifies the customer's investment propensity into various types such as stable, neutral, and aggressive investment types. can do.

The digital asset investment agency 1300 can perform customer asset investment agency by constructing an appropriate portfolio based on the analysis results of transaction data for each exchange and customer investment propensity. In other words, the digital asset investment agency 1300 can configure a digital asset portfolio by adjusting the rate of return and degree of diversification to suit the customer's investment propensity, based on the customer's investment propensity, and act as an agent for digital asset investment accordingly.

The transaction data analysis unit 1100 calculates the output value of each artificial intelligence model (1141-1145) based on the digital asset price increase rate and accuracy predicted by each artificial intelligence model (1141-1145) of the artificial intelligence unit (1140). By combining these, you can finally predict the price increase rate and accuracy of digital assets for each exchange. The exchange transaction data analysis unit 1120 may determine the weight corresponding to each artificial intelligence model based on the output prediction accuracy based on the training transaction data of each artificial intelligence model.

For example, if there are five artificial intelligence models, the exchange transaction data analysis unit 1120 sets the priorities of the five artificial intelligence models in the order of high output prediction accuracy based on the transaction data for learning, and determines the priority of each artificial intelligence model. Based on the ranking, the weight of each artificial intelligence model is determined, the weight determined for each artificial intelligence model is applied, and the price increase rate and accuracy of digital assets can be predicted based on the summed value.

Figure 20 is a conceptual diagram of an artificial intelligence unit constituting a digital asset custody system according to an embodiment of the present invention. As described with reference to FIGS. 19 and 20 , the machine learning unit 1130 may generate the first artificial intelligence model 1141 through the machine learning method of the LGBM (Light Gradient Boosting Machine) artificial intelligence algorithm. The LGBM artificial intelligence algorithm is a decision tree-based model that can be an algorithm that learns data at high speed through sampling that reduces data instances while maintaining accuracy based on the gradient of the data.

The machine learning unit 1130 can generate a second artificial intelligence model 1142 through a recursive recurrent artificial neural network (RNN) machine learning method based on long short term memory (LSTM). The LSTM artificial neural network learning method is a type of recurrent neural network (RNN) and can be a learning algorithm that uses short- and long-term memory to prevent the vanishing gradient problem, which was a problem with existing RNNs.

The machine learning unit 1130 can generate a third artificial intelligence model 1143 through a machine learning method of a machine learning algorithm based on KNN (K-Nearest Neighbors). The K-nearest neighbor-based machine learning algorithm may be an algorithm that selects the few points closest to the input feature vector and uses their labels.

The machine learning unit 1130 can generate the fourth artificial intelligence model 1144 through a machine learning method of a machine learning algorithm based on K-means clustering. A machine learning algorithm based on K-means clustering is a learning method that clusters a given input. It can be an algorithm that assumes that there are a total of K clusters and finds K centroids in the feature space. .

The machine learning unit 1130 can generate the fifth artificial intelligence model 1145 through a machine learning method of a principal component analysis (PCA)-based machine learning algorithm. A machine learning algorithm based on a principal component analysis model may be an algorithm that utilizes a technique to convert data from a high-dimensional space to a low-dimensional space while preserving the distribution of the original data as much as possible.

In addition, the machine learning unit 1130 can generate a sixth artificial intelligence model (not shown) through a machine learning method of an artificial neural network (ANN) machine learning algorithm based on a fully connected layer. ANN (artificial neural network) machine learning algorithm based on a fully connected layer is a method of learning an artificial neural network in which multiple hidden nodes are connected between multiple input nodes and multiple output nodes in a fully connected layer (FCL) structure. It may be an algorithm of .

The machine learning unit 1130 may additionally generate a seventh artificial intelligence model (not shown) through a machine learning method of a support vector machine (SVM)-based machine learning algorithm. A machine learning algorithm based on a support vector machine may be an algorithm that uses an optimal hyperplane that can distinguish between two categories of data given in the feature space.

In addition, the machine learning unit 1130 can generate an eighth artificial intelligence model (not shown) through a machine learning method of an auto encoder-based machine learning algorithm. An autoencoder-based machine learning algorithm may be an unsupervised learning algorithm that learns a function that makes the output value an approximation of the input value.

The exchange transaction data analysis unit 1120 uses the first artificial intelligence model (1141), the second artificial intelligence model (1142), the third artificial intelligence model (1143), and the fourth artificial intelligence model (1141) based on the transaction data for each exchange. 1144), the fifth artificial intelligence model 1145, and the sixth to eighth artificial intelligence models output, respectively, can obtain the price increase rate and accuracy prediction results of digital assets for each exchange.

The exchange transaction data analysis unit 1120 combines the output results predicted by the first to eighth artificial intelligence models according to the weight determined for each artificial intelligence model, and finally calculates the predicted price increase rate and accuracy of digital assets for each exchange. Can be printed. According to an embodiment of the present invention, the price increase rate and accuracy of digital assets for each exchange can be predicted with higher accuracy based on various artificial intelligence models, and accordingly, an exchange that matches (matches) the customer's investment type is recommended, We can recommend a digital asset investment portfolio suitable for customers.

Hereinafter, an exemplary computing device 500 in which a custody system and a digital asset management method thereof according to various embodiments of the present invention are implemented will be described with reference to FIG. 21. 21 is an exemplary hardware configuration diagram showing the computing device 500. As shown in FIG. 21, the computing device 500 loads one or more processors 510, a bus 550, a communication interface 570, and a computer program 591 performed by the processor 510. It may include a memory 530 that stores a computer program 591 and a storage 590 that stores a computer program 591. However, only components related to the embodiment of the present invention are shown in FIG. 21. Accordingly, a person skilled in the art to which the present invention pertains can see that other general-purpose components other than those shown in FIG. 21 may be further included.

The processor 510 controls the overall operation of each component of the computing device 500. The processor 510 is at least one of a Central Processing Unit (CPU), Micro Processor Unit (MPU), Micro Controller Unit (MCU), Graphic Processing Unit (GPU), or any type of processor well known in the art of the present invention. It can be configured to include. Additionally, the processor 510 may perform operations on at least one application or program to execute methods/operations according to various embodiments of the present invention. Computing device 500 may include one or more processors.

The memory 530 stores various data, commands and/or information. The memory 530 may load one or more programs 591 from the storage 590 to execute methods/operations according to various embodiments of the present invention. An example of the memory 530 may be RAM, but is not limited thereto.

Bus 550 provides communication functionality between components of computing device 500. The bus 550 may be implemented as various types of buses, such as an address bus, a data bus, and a control bus. The communication interface 570 supports wired and wireless Internet communication of the computing device 500. The communication interface 570 may support various communication methods other than Internet communication. To this end, the communication interface 570 may be configured to include a communication module well known in the technical field of the present invention.

Storage 590 may non-transitory store one or more computer programs 591. The storage 590 may be used in a non-volatile memory such as Read Only Memory (ROM), Erasable Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), flash memory, a hard disk, a removable disk, or in the technical field to which the present invention pertains. It may be configured to include any known type of computer-readable recording medium.

The computer program 591 may include one or more instructions implementing methods/operations according to various embodiments of the present invention. When the computer program 591 is loaded into the memory 530, the processor 510 can perform methods/operations according to various embodiments of the present invention by executing the one or more instructions.

Next, we explain the digital asset custody system in which users can participate as consensus nodes. Specifically, the present invention can provide a digital asset custody system that can have approval authority for verification based on the size of the digital asset entrusted by the user.

Figure 22 is a block diagram of a digital asset custody system (or digital asset custody service providing system) according to an embodiment of the present invention. Referring to Figure 22, the digital asset custody service providing system (hereinafter, may be abbreviated as 'system') according to an embodiment of the present invention includes a control unit 2100, an account creation unit 2200, and a consensus node management unit ( 2300), an account verification unit 2400, a digital asset management unit 2500, a transaction record storage unit 2600, and a transaction record disclosure unit 2700. Figure 22 shows seven components included in the system, but the components shown are not essential, and the system may have more or fewer components. Additionally, each component of the system may be physically included in one server, or may be a distributed server distributed for each function.

The control unit 2100 can oversee the operation of the system. Specifically, the control unit 2100 includes an account creation unit 2200, a consensus node management unit 2300, an account verification unit 2400, a digital asset management unit 2500, a transaction record storage unit 2600, and a transaction record disclosure unit 2700. You can execute the actions of each department by sending control commands to ). Unless otherwise specified below, the operation of the system may be interpreted as being performed under the control of the control unit 2100.

The account creation unit 2200 may receive digital asset management request data from a plurality of user terminals. The account creation unit 2200 may include a communication module capable of transmitting and receiving data. The plurality of user terminals may be terminals of customers who entrust digital asset management to the system. At this time, digital asset management request data may include user information, size of digital asset to be entrusted, entrustment period, user request, etc.

The account creation unit 2200 may create a plurality of user accounts corresponding to each user terminal based on digital asset management request data. For example, the account creation unit 2200 may create a first user account corresponding to the first user terminal and a second user account corresponding to the second user terminal.

Before the account creation unit 2200 creates a user account, the system may verify the identity of the user who transmitted the digital asset management request data (KYC, Know Your Customer). When the user's identity has been verified, the system may send a signal to the account creation unit 2200 to create an account.

The account creation unit 2200 can create a wallet that manages virtual assets along with a user account. The account creation unit 2200 can store the public key and private key created during the wallet creation process by dividing them into a storage unit and a cold wallet room of the system. At this time, the cold wallet room is accessible to the key storage system, and the security policy in the ISMS can be applied. The system can store the virtual asset wallet creation results in storage.

The consensus node management unit 2300 may distribute consensus authority shares for each of a plurality of user accounts based on the size of the digital asset included in the digital asset management request data. In the system of the present invention, custody customers can participate as consensus nodes in the blockchain network. Specifically, by participating as a consensus node, customers can evaluate the reliability of the entire system by verifying the data recorded in the block.

The present invention is characterized in that the stake for verification is determined according to the size of the customer's entrusted assets. The consensus node management unit 2300 may distribute consensus authority shares to each user account based on the ratio of the digital asset size of each user account to the total digital asset size of the plurality of user accounts.

For example, if the total digital asset size of multiple user accounts is 100 million won, the entrusted digital asset size of the first user account is 30 million won, and the entrusted digital asset size of the second user account is 10 million won, the consensus The node management unit 2300 may distribute a 30% consensus authority share to the first user account and a 10% consensus authority share to the second user account.

The system can charge fees or distribute interest generated by the blockchain network based on the consensus authority stake distributed to each user account. Users have a stake in verification based on their share of consensus rights distributed to them. In the example above, if the first user account completes approval for verification, 30% of the consensus approval stake may be added.

The account verification unit 2400 may perform verification of the digital asset management application according to the sum of the agreed approval shares of multiple user accounts. Specifically, the account verification unit 2400 may process the verification as completed when the sum of the agreed-upon shares is greater than or equal to the reference value. At this time, the reference value, which is the standard for verification completion processing, may be adjusted based on at least one of the type of digital asset, security level, total digital asset size, and user request.

For example, if the total digital asset size is 100 million won or less, the standard value may be 50%, and if the total digital asset size is 500 million won or more, the standard value may be 80%. Also, for example, if there is a user whose consensus authority share is more than 50%, the reference value may be set to a value exceeding 50%. Also, for example, if the security level is the highest level, the reference value may be set to 90% or higher.

When the account verification unit 2400 performs verification, the digital asset management unit 2500 may perform digital asset management for a plurality of user accounts based on digital asset management request data. Since the contents of the digital asset management unit 2500 may overlap with the contents of the digital asset management unit 280 of FIG. 7, detailed descriptions are omitted.

The transaction record storage unit 2600 can record information generated or processed by the account creation unit 2200, consensus node management unit 2300, account verification unit 2400, and digital asset management unit 2500 on the blockchain. there is. For example, the transaction record storage unit 2600 may store data on a plurality of user accounts and a plurality of wallets created by the account creation unit 2200. For example, the transaction record storage unit 2600 may store information about the consensus authority share distributed by the consensus node management unit 2300. For example, the transaction record storage unit 2600 may store data about digital asset management contents of the digital asset management unit 2500.

The transaction record storage unit 2600 can set sensitivity for the data it stores. Specifically, the transaction record storage unit 2600 can set a score or level related to security for the data it stores. For example, the transaction record storage unit 2600 may set the security level of the user's individual membership registration details, wallet management information, and activity details. Also, for example, the transaction record storage unit 2600 may set the security level of deposit/withdrawal details on the blockchain network to medium.

The transaction record disclosure unit 2700 may transmit data to the user terminal at the request of the user terminal. The transaction record disclosure unit 2700 may transmit encrypted or unencrypted data depending on which user terminal requested which information and the sensitivity set for the data. The transaction record disclosure unit 2700 may include a communication module capable of transmitting and receiving data.

Specifically, the transaction record disclosure unit 2700 may receive a transaction viewing signal related to the first user account from the first user terminal corresponding to the first user account. Since this is a case of viewing information related to one's account, the transaction record disclosure unit 2700 can transmit data related to the first user account to the first user terminal without encrypting it, regardless of sensitivity.

On the other hand, the transaction record disclosure unit 2700 may receive a transaction viewing signal related to the second user account from the first user terminal corresponding to the first user account. Since this is a case of viewing information related to another user's account, the transaction record disclosure unit 2700 can selectively transmit data depending on sensitivity.

Specifically, the transaction record disclosure unit 2700 may transmit first transaction disclosure data to the first user terminal. The first transaction disclosure data may include first data and second data. At this time, the first data may be data whose sensitivity is below the threshold among data related to the second user account. Also, at this time, the second data may be data whose sensitivity exceeds the threshold among data related to the second user account. Additionally, the second data may be encrypted data (for example, data using a two-way encryption method).

In the above explanation, an example in which sensitivity is set for data stored by the transaction record storage unit 2600 and the transaction record disclosure unit 2700 transmits data according to the sensitivity is explained. However, it is not limited to this, and an example in which the transaction record storage unit 2600 does not set the sensitivity but the transaction record disclosure unit 2700 sets the sensitivity is also possible. For example, the transaction record disclosure unit 2700 may set sensitivity to data stored in the transaction record storage unit 2600 according to security and transmit the data to the user terminal according to the sensitivity.

Figure 23 is a flowchart of a method for verifying digital asset custody services and managing digital assets according to an embodiment of the present invention. Referring to Figure 23, the verification method of the digital asset custody service according to an embodiment of the present invention includes receiving digital asset management request data (S2110), creating a plurality of user accounts (S2120), and establishing consensus authority. It may include a step of distributing shares (S2130), a step of performing verification of a digital asset management application (S2140), and a step of performing digital asset management (S2150). Figure 23 shows that steps S2110 to S2150 are performed sequentially, but the present invention is not limited to this, and some steps may be omitted, performed simultaneously, or new steps may be added.

The step of receiving digital asset management request data (S2110) may be a step in which the system receives a request for entrustment of digital assets from a plurality of user terminals. At this time, digital asset management request data may include user information, size of digital asset to be entrusted, entrustment period, user request, etc.

The step of creating a plurality of user accounts (S2120) may be a step in which the account creation unit 2200 creates user accounts corresponding to the user terminal based on digital asset management request data.

The step of distributing the consensus authority share (S2130) may be a step in which the consensus node management unit 2300 sets the consensus authority share for each user based on the size of the entrusted digital asset included in the digital asset management request data. The consensus node management unit 2300 may set the consensus authority share based on the ratio of the size of the entrusted digital assets of each user account to the total entrusted digital assets of the plurality of user accounts.

In the step of performing verification of the digital asset management application (S2140), the account verification unit 2400 manages the digital asset based on whether the sum of the shares approved for the consensus rights held by each user is more than the standard value. This may be the stage of approving verification of the application.

The step of performing digital asset management (S2150) may be a step of performing management of digital assets entrusted by the digital asset management unit 2500 when verification is performed in step S2140.

Figure 24 is a flowchart of a method for disclosing transaction records of a digital asset custody service according to an embodiment of the present invention. Referring to Figure 24, the method of disclosing transaction records of a digital asset custody service according to an embodiment of the present invention includes recording transaction data on a blockchain (S2210) and setting sensitivity to data (S2220). , It may include receiving a transaction viewing signal (S2230), selecting data and/or encrypting data (S2240), and transmitting data (S2250). Figure 24 shows steps S2210 to S2250 being performed sequentially, but the present invention is not limited to this, and some steps may be omitted, performed simultaneously, or new steps may be added.

The step of recording transaction data on the blockchain (S2210) may be a step in which the transaction record storage unit 2600 stores the data generated or processed in steps S2110 to S2150 of FIG. 23 in the storage of the system.

The step of setting sensitivity to data (S2220) may be a step in which the transaction record storage unit 2600 or the transaction record disclosure unit 2700 sets sensitivity according to the security level of the data.

For example, the transaction record storage unit 2600 or the transaction record disclosure unit 2700 may set the sensitivity to low since the deposit/withdrawal information on movement of virtual assets among the transaction records recorded in the blockchain is information that needs to be disclosed.

Also, for example, the transaction record storage unit 2600 or the transaction record disclosure unit 2700 may store user personal information, membership registration/change information, virtual asset wallet creation/modification/deletion information, and/or customer activity history information. Since this is information that requires security, the sensitivity can be set high.

The step of receiving a transaction viewing signal (S2230) may be a step in which the system receives a transaction viewing signal related to a transaction viewing request from the user terminal. At this time, the transaction viewing signal may include data information to be viewed.

The step of selecting and/or encrypting data (S2240) may be a step in which the transaction record disclosure unit 2700 selects data based on the transaction viewing signal received in step S2230. For example, if the transaction viewing signal is a signal requesting information about one's account, the transaction record disclosure unit 2700 may transmit all requested data to the user terminal.

However, for example, if the transaction viewing signal is a signal requesting information about another person's account, the transaction record disclosure unit 2700 transmits data with sensitivity lower than the threshold to the user terminal as is, but transmits data with sensitivity higher than the threshold. Data can be encrypted and transmitted (S2250). Alternatively, the transaction record disclosure unit 2700 may not transmit data with a sensitivity higher than the threshold and thus may not disclose the data.

Figure 25 is an example diagram to explain a method of recording transaction data using a blockchain network. Referring to Figure 25, the system can store transaction records that occur within the system on the blockchain network. Transaction records stored in each block can be encrypted and stored in hash form. Accordingly, the contents can only be confirmed through the parties and addresses that participated in the virtual asset transaction corresponding to the transaction record.

Figure 26 is an example diagram to explain a method of disclosing transaction records. Referring to FIG. 26, information that needs to be disclosed or information that is not sensitive can be set to be disclosed depending on the user's selection or the sensitivity of the data. On the other hand, information that the user does not want to disclose or that is sensitive can be set to private.

Among the transaction records recorded in the blockchain, deposit/withdrawal information on the movement of virtual assets on the blockchain is information that needs to be disclosed, so it can be made public. On the other hand, membership sign-up/change information, virtual asset wallet creation/change/deletion information, and/or customer activity history inquiry information that is part of the user's personal information or is sensitive information may be set as non-public.

The transaction record disclosure unit 2700 can transmit information set to public as is at the request of the user terminal. However, the transaction record disclosure unit 2700 may not transmit information set as non-public even if requested by the user terminal, or may encrypt and transmit this information upon request.

Next, we explain the digital asset custody system that provides premium services to users. Specifically, the present invention can provide a digital asset custody system that can provide premium services to users using non-fungible tokens.

Figure 27 is a block diagram of a digital asset custody service providing system according to another embodiment of the present invention. Referring to FIG. 27, the digital asset custody service providing system according to another embodiment of the present invention includes a control unit 3100, an account creation unit 3200, a consensus node management unit 3300, a token issuance unit 3400, and a token. It may include a processing unit 3500, a transaction record storage unit 3600, and a transaction record disclosure unit 3700. Figure 27 shows seven components included in the system, but the components shown are not essential, and the system may have more or fewer components. Additionally, each component of the system may be physically included in one server, or may be a distributed server distributed for each function.

The control unit 3100 can oversee the operation of the system. Specifically, the control unit 3100 includes an account creation unit 3200, a customer management unit 3250, a consensus node management unit 3300, a token issuance unit 3400, a token processing unit 3500, a transaction record storage unit 3600, and a transaction unit. The operations of each department can be executed by sending a control command to the record and disclosure unit 3700. In this embodiment, unless otherwise specified, the operation of the system can be interpreted as being performed under the control of the control unit 3100. Since the contents of the account creation unit 3200 overlap with the contents of the account creation unit 2200 of FIG. 22, detailed descriptions are omitted.

The customer management unit 3250 may determine the class of the user account based on the size of the digital asset included in the digital asset management request data. Additionally, the customer management unit 3250 may perform overall management functions so that customer services corresponding to the determined class can be provided to the user.

For example, the customer management department 3250 sets users with entrusted digital assets of 100 million won or less to the first class, users with more than 100 million won to 300 million won or less to the second class, and users with more than 300 million won to the third class. You can. At this time, the first class may correspond to the general level, the second class may correspond to the VIP level, and the third class may correspond to the VVIP level, but are not limited thereto. Since the contents of the consensus node manager 3300 overlap with the contents of the consensus node manager 2300 of FIG. 22, detailed descriptions are omitted.

When determining the class of the user account, the customer management unit 3250 may consider the consensus authority share of the user account set by the consensus node management unit 3300. For example, if the customer's entrusted assets are less than 100 million won, but the consensus share is more than 50%, the customer management unit 3250 may determine the user account to be a third class corresponding to VVIP.

The token issuing unit 3500 may check service details corresponding to the class based on the class of the user account determined by the customer management unit 3250. The token issuing unit 3500 may issue a token containing the right to use services available at the service provider based on the confirmed service details. For example, service providing organizations may include, but are not limited to, hotels, restaurants, lounges, golf courses, exercise centers, exhibition halls, or performance halls.

The token issuing unit 3500 may create a smart contract based on information about service use rights and issue a non-fungible token based on the created smart contract. At this time, the non-fungible token may be a multiple type, but is not limited to this.

For example, the token issuing unit 3500 may issue a first non-fungible token containing the right to use one night at a 5-star hotel. Also, for example, the token issuing unit 3500 may issue a second non-fungible token containing a one-year right to use Incheon International Airport Lounge A. Also, for example, the token issuing unit 3500 may issue a third non-fungible token containing the right to use a specific sports center for one year.

The token processing unit 3500 may obtain data regarding whether or not a token is used from a service providing organization. When the token processing unit 3500 obtains information on completion of use of a non-fungible token containing the right to use a service from a service provider, it can discard the non-fungible token. Additionally, the token processing unit 3500 may discard the token even if the token has not been used when the usage period of the non-fungible token has expired.

The transaction record storage unit 3600 can record data generated or processed by the account creation unit 3200, customer management unit 3250, token issuance unit 3400, and token processing unit 3500 on the blockchain. Since the contents of the transaction record storage unit 3600 overlap with the contents of the transaction record storage unit 2600 of FIG. 22, detailed descriptions are omitted. Since the contents of the transaction record disclosure unit 3700 overlap with the contents of the transaction record disclosure unit 2700 of FIG. 22, detailed descriptions are omitted.

Figure 28 is a flowchart of a method of providing a digital asset custody service that provides a premium service using a non-fungible token according to an embodiment of the present invention. Referring to Figure 28, the method of providing a digital asset custody service that provides a premium service using a non-fungible token according to an embodiment of the present invention includes receiving digital asset management request data (S3110) and creating a user account. step (S3120), determining the class of the user account (S3130), checking service details corresponding to the class (S3140), issuing a non-fungible token containing the right to use the service (S3150), and service When used, a step (S3160) of discarding the non-fungible token may be included. Figure 28 shows steps S3110 to S3150 being performed sequentially, but the present invention is not limited to this, and some steps may be omitted, performed simultaneously, or new steps may be added.

The step of receiving digital asset management request data (S3110) may overlap with the content of step S3110 of FIG. 23, so detailed description is omitted. The step of creating a user account (S3120) may be a step in which the account creation unit 3200 creates a user account corresponding to the user terminal based on digital asset management request data.

The step of determining the class of the user account (S3130) may be a step in which the customer management unit 3250 determines the class of the user account based on the size of the digital asset included in the digital asset management request data. The customer management unit 3250 may determine the class of the user account based on the consensus authority share distributed by the consensus node management unit 3300, as well as the digital asset size.

The step of checking the service details corresponding to the class (S3140) may be a step in which the token issuing unit 3400 checks whether any service usage rights are included in the class for token issuance. For example, a first class user can use a 5-star hotel, a second class user can use the first class right and a meal ticket for four at a specific restaurant, and a third class user can use the second class. You can use the rights and lounge access for 1 year.

The step of issuing a non-fungible token containing a service use right (S3150) may be a step in which the token issuing unit 3400 issues a non-fungible token containing a service use right based on the service details confirmed in step S3140. At this time, the token issuing unit 3400 may issue multiple non-fungible tokens when there are multiple service use rights. Alternatively, the token issuing unit 3400 may issue one non-fungible token even when there are multiple service use rights.

When the service is used, the step of discarding the non-fungible token (S3160) may be a step of discarding the used token when the token processing unit 3500 obtains information related to the completion of token use from the service provider. At this time, when one non-fungible token includes multiple service usage rights, the token processing unit 3500 can discard the non-fungible token only when all service usage rights have been used.

Figure 29 is an example diagram to explain the issuance of a non-fungible token containing the right to use a service. Referring to Figure 29, the custody system can create a smart contract based on information related to service use rights. The custody system can issue a non-fungible token containing the right to use the service based on the created smart contract. The custody system can discard non-fungible tokens when the service corresponding to the non-fungible token has been used.

Meanwhile, if a customer of a digital asset custody service uses the custody system 10 to transfer illegal funds or attempts to conduct an unfair transaction, it may be necessary to detect the customer's transaction and block the transaction. Figure 30 is a configuration diagram of an abnormal transaction detection system according to an embodiment of the present invention, and Figure 31 is a table showing deposit and withdrawal data of customers using the virtual asset custody service according to an embodiment of the present invention.

Referring to FIG. 30, the abnormal transaction detection system 4000 may include an analyst terminal 4100 and an operator terminal 4200. The abnormal transaction detection system 4000 may be a system that determines whether a digital asset transaction occurring in the custody system 10, which is configured to provide a service for managing digital assets, is an abnormal transaction. The abnormal transaction detection system 4000 may be installed on the server of a digital asset custody service provider, but the abnormal transaction detection system 4000 is not necessarily limited to being installed on the server of the service provider.

The analyst terminal 4100 may be a terminal configured to determine whether a transaction subject to inspection is an abnormal transaction. The analyst terminal 4100 may include a deposit/withdrawal data reception unit 4110, an abnormal transaction analysis unit 4120, a machine learning unit, and a machine learning unit 4140.

The deposit/withdrawal data receiving unit 4110 may be configured to receive deposit/withdrawal data of a transaction to be inspected from the custody system 10. The transaction subject to inspection may be a transaction that is subject to determination as to whether or not it is an abnormal transaction among transactions occurring in the custody system 10. The deposit/withdrawal data receiving unit 4110 may transmit the deposit/withdrawal data of the transaction subject to inspection to the abnormal transaction analysis unit 4120.

Referring to Figure 31, you can check the information included in the customer deposit/withdrawal related data generated according to the operation of the virtual asset custody service. For example, deposit/withdrawal data may include information such as the date and time the transaction occurred through the custody system 10, the ID of the transaction customer, the address name, the country and region where the transaction occurred, and the balance after the transaction occurred. However, information included in deposit/withdrawal data is not limited to the information shown.

The deposit/withdrawal data receiving unit 4110 may receive information included in the deposit/withdrawal data of the transaction to be inspected from the custody system 10 and transmit it to the abnormal transaction analysis unit 4120.

Referring to Figures 30 and 31, the abnormal transaction analysis unit 4120 can use an artificial intelligence model based on deposit/withdrawal data to determine whether the transaction to be inspected is an abnormal transaction subject to transaction suspension. In other words, the abnormal transaction analysis unit 4120 inputs the information contained in the deposit/withdrawal data of the transaction to be inspected into the artificial intelligence model, and can determine whether the transaction to be inspected is an abnormal transaction based on the information output by the artificial intelligence model. .

The machine learning unit 4130 sets the deposit/withdrawal data of normal transactions for learning and the deposit/withdrawal data of abnormal transactions for learning as input variables, and sets normal status information corresponding to normal transactions for learning and normal status information corresponding to abnormal transactions for learning as output variables. Thus, an artificial intelligence model can be created through machine learning methods.

Information included in the deposit/withdrawal data of normal transactions for learning purposes may include information about the transaction amount without any abnormality, information about the customer, information about the country or region, etc. Normal status information corresponding to a normal transaction for learning may include information to the effect that the normal transaction for learning was a normal transaction.

Information included in the deposit/withdrawal data of abnormal transactions for learning purposes may include information about the transaction amount, information about the customer, and information about the country or region that may be problematic. The normality information corresponding to an abnormal transaction for learning purposes may include information to the effect that the abnormal transaction for learning purposes was an abnormal transaction or a transaction subject to regulation.

The machine learning department (4130) deals with transactions of digital assets executed by consignors who have had reasons for transaction suspension in the past, transactions of digital assets by withdrawal targets who have had reasons for transaction suspension in the past, abnormal cross-trades, abnormally high-priced transactions, and abnormally low-priced transactions. An artificial intelligence model can be created through machine learning by setting at least one abnormal transaction for learning as an input variable and setting the normal status information corresponding to the abnormal transaction for learning as an output variable.

Machine learning can mean using a model composed of multiple parameters and optimizing the parameters with given data. Machine learning can include supervised learning, unsupervised learning, and reinforcement learning, depending on the type of learning problem. Supervised learning is learning the mapping between input and output, and can be applied when input and output pairs are given as data. Unsupervised learning is applied when there is only input and no output, and can find regularities between inputs.

The machine learning unit 4130 can create artificial intelligence models in various ways. For example, the machine learning unit 4130 can learn features extracted from information included in abnormal transactions for learning or information included in normal transactions for learning using a deep learning-based learning method. At this time, a CNN (Convolutional Neural Networks) structure that stacks several stages of convolutional layers can be used to learn how to extract features from the information included in abnormal transactions for learning or the information included in normal transactions for learning. However, the learning method of the machine learning unit 4130 is not necessarily limited to a method using the CNN structure. For example, the learning methods of the machine learning unit 4130 include artificial neural network (ANN), recurrent neural network (RNN), K-Nearest Neighbors (KNN), K-Means Clustering, Principal Component Analysis (PCA), and SVM ( This may be done through a machine learning algorithm including Suport Vector Machine or Auto Encoder.

The taker terminal 4200 may be a manager who manages transactions occurring in the custody system 10 or a taker terminal that takes action. The operator terminal 4200 may be configured to determine whether to suspend transactions for the consignor terminal 100 related to the transaction to be inspected.

Referring to FIG. 30, if the transaction to be inspected is determined to be an abnormal transaction, the analyst terminal 4100 may transmit a warning signal to the operator terminal 4200. When the operator terminal 4200 receives a warning signal, it can output information indicating the occurrence of an abnormal transaction and which transaction is an abnormal transaction to an output unit such as a display of the operator terminal 4200. An administrator or operator using the operator terminal 4200 can take action on transactions determined to be abnormal transactions based on this information. For example, the manager or operator can closely analyze whether the abnormal transaction detected through the operator terminal 4200 is normal. Additionally, when there are multiple managers or takers, decision-making by the internal monitoring and control organization may be performed through the taker terminals 4200 used by each manager or taker.

When receiving a warning signal, the operator terminal 4200 can transmit a transaction stop signal to the custody system 10 to stop the transaction subject to inspection. When the custody system 10 receives a transaction stop signal, it can stop the transaction subject to inspection corresponding to the transaction stop signal. In addition, when the custody system 10 receives a transaction stop signal, it can also stop or sanction other transactions by the trustor terminal 100 involved in the transaction subject to inspection.

The operator terminal 4200 may include an input unit 4210. The input unit 4210 may receive input information from the manager. The input unit 4210 may receive input information through a button provided on the operator terminal 4200 that receives physical force from the manager or operator, or by receiving input information through a touch input method on the display. It is not limited to this.

Meanwhile, the transaction subject to inspection, which was determined to be an abnormal transaction by the analyst terminal 4100, may be confirmed as a normal transaction rather than an abnormal transaction as a result of detailed analysis by the manager or operator. In addition, the transaction subject to inspection, which was determined to be an abnormal transaction by the analyst terminal 4100, may be confirmed as a normal transaction by the decision-making of the internal monitoring and control organization through the operator terminals 4200. At this time, the input unit 4210 of the operator terminal 4200 may receive context confirmation information to the effect that the transaction to be inspected has been confirmed to be a normal transaction from the manager or the operators.

After transmitting the transaction stop signal, when the input unit 4210 receives normal confirmation information, the operator terminal 4200 transmits a transaction resumption signal to the custody system 10 to cancel the suspension of the transaction to be inspected. can do.

When the custody system 10 receives a transaction resumption signal, it can resume the progress of the transaction to be inspected that was stopped by the transaction stop signal. In addition, when the custody system 10 receives a transaction resumption signal, it can also resume processing of other transactions of the consignor terminal 100 involved in the transaction subject to inspection.

Meanwhile, the transaction subject to inspection, which has been determined to be an abnormal transaction by the analyst terminal 4100, may also be confirmed as an abnormal transaction based on a detailed analysis result by the manager or operator. In addition, the transaction subject to inspection, which was determined to be an abnormal transaction by the analyst terminal 4100, may be confirmed as a normal transaction by the decision-making of the internal monitoring and control organization through the operator terminals 4200. At this time, the input unit 4210 of the operator terminal 4200 may receive context confirmation information from the manager or operators to the effect that the transaction to be inspected has been confirmed to be an abnormal transaction.

After transmitting a transaction stop signal, the operator terminal 4200, when the input unit 4210 receives abnormality confirmation information to the effect that the transaction to be inspected has been confirmed to be an abnormal transaction, instructs the operator terminal 4200 to take action in accordance with laws and regulations regarding the transaction to be inspected. A signal to this effect can be transmitted to the custody system 10.

When the custody system 10 receives a signal to proceed with action in accordance with laws and regulations for the transaction subject to inspection, it can proceed with a preset process in accordance with the law for the transaction to be inspected that was stopped by the transaction stop signal. In addition, when the custody system 10 receives a signal to proceed with action in accordance with laws and regulations for the transaction subject to inspection, the custody system 10 performs preset actions in accordance with the law for other transactions of the consignor terminal 100 involved in the transaction subject to inspection. The process can proceed.

Figure 32 is a diagram for explaining a machine learning-based abnormal transaction detection method according to an embodiment of the present invention. Referring to FIG. 32, the abnormal transaction detection system 4000 inputs the virtual asset deposit/withdrawal data of the transaction to be inspected into a plurality of artificial intelligence models, and detects the transaction to be inspected as abnormal based on the information output by the plurality of artificial intelligence models. It can be classified as a transaction. Referring to FIGS. 30 and 32 , the machine learning unit 4130 may generate a plurality of different artificial intelligence models based on different and separate machine learning algorithms.

The abnormal transaction analysis unit 4120 can obtain a result of determining whether the transaction to be inspected is abnormal, which is output by a plurality of artificial intelligence models based on deposit and withdrawal data. The abnormal transaction analysis unit 4120 can determine whether the transaction subject to inspection is an abnormal transaction based on the number of artificial intelligence models that determined the transaction to be inspected as an abnormal transaction.

Specifically, the abnormal transaction analysis unit 4120 may classify the transaction to be inspected as an abnormal transaction if the number of artificial intelligence models that determine the transaction to be inspected as an abnormal transaction is more than a preset number. For example, if the preset number is 3 and the number of artificial intelligence models that classified the transaction to be inspected as an abnormal transaction among the five artificial intelligence models is 4, the abnormal transaction analysis unit 4120 classifies the transaction to be inspected as an abnormal transaction. Can be classified.

The abnormal transaction analysis unit 4120 may determine a weight corresponding to each artificial intelligence model based on the detection rate of abnormal transactions for training of each artificial intelligence model. For example, if there are five artificial intelligence models, the corresponding weights can be determined for each artificial intelligence model in the order of the highest detection rate for abnormal transactions for learning: 5, 4, 3, 2, and 1.

The abnormal transaction analysis unit 4120 can determine whether the transaction subject to inspection is an abnormal transaction based on the weight of the artificial intelligence model that judges the transaction to be inspected as an abnormal transaction and the weight of the artificial intelligence model that determines the transaction to be inspected as a normal transaction. there is.

Specifically, the abnormal transaction analysis unit 4120 may classify the transaction to be inspected as an abnormal transaction if the combined weight of the artificial intelligence models that determined the transaction to be inspected to be an abnormal transaction is greater than or equal to a preset value. For example, if the preset value is 11 and the weights of the artificial intelligence models that classified the transaction to be inspected as an abnormal transaction among the artificial intelligence models are 5, 4, and 3, the combined value is 12, which is more than the preset value, so it is abnormal. The transaction analysis unit 4120 may classify the transaction subject to inspection as an abnormal transaction.

The machine learning unit 4130 may generate the first artificial intelligence model 4141 through a machine learning method of a machine learning algorithm based on a distance-based outlier detection model (K-Nearest Neighbors; KNN). A machine learning algorithm based on an outlier detection model may be an algorithm that selects the few points closest to the input feature vector and uses their labels.

The machine learning unit 4130 can generate a second artificial intelligence model 4142 through a machine learning method of a machine learning algorithm based on a cluster-based outlier detection model (K-Means Clustering). A machine learning algorithm based on a cluster-based outlier detection model is a learning method that clusters a given input. It can be an algorithm that assumes that there are a total of K clusters and finds K centroids in the feature space. .

The machine learning unit 4130 can generate a third artificial intelligence model 4143 through a machine learning method of a machine learning algorithm based on a principal component analysis (PCA) model. A machine learning algorithm based on a principal component analysis model may be an algorithm that utilizes a technique to convert data from a high-dimensional space to a low-dimensional space while preserving the distribution of the original data as much as possible.

The machine learning unit 4130 can generate the fourth artificial intelligence model 4144 through a machine learning method of a machine learning algorithm based on a support vector machine (SVM). A machine learning algorithm based on a support vector machine may be an algorithm that uses an optimal hyperplane that can distinguish between two categories of data given in the feature space.

The machine learning unit 4130 can generate the fifth artificial intelligence model 4145 through a machine learning method of an auto encoder-based machine learning algorithm. The autoencoder-based machine learning algorithm is an unsupervised learning algorithm that learns a function that makes the output value an approximation of the input value.

The abnormal transaction analysis unit 4120 analyzes the first artificial intelligence model (4141), the second artificial intelligence model (4142), the third artificial intelligence model (4143), and the fourth artificial intelligence model (4144) based on deposit and withdrawal data. , it is possible to obtain the result of determining whether the transaction subject to inspection is an abnormal transaction output by the fifth artificial intelligence model (4145).

The abnormal transaction analysis unit 4120 can determine whether the transaction subject to inspection is an abnormal transaction based on the weight of the artificial intelligence model that judges the transaction to be inspected as an abnormal transaction and the weight of the artificial intelligence model that determines the transaction to be inspected as a normal transaction. there is.

Figure 33 is a flowchart of a control method of an abnormal transaction detection system according to an embodiment of the present invention. Referring to FIG. 33, the deposit/withdrawal data receiving unit 4110 may receive deposit/withdrawal data of the transaction to be inspected from the custody system 10 (S4001).

The abnormal transaction analysis unit 4120 can use an artificial intelligence model based on deposit/withdrawal data to determine whether the transaction to be inspected is an abnormal transaction subject to transaction suspension (S4002). If the analyst terminal 4100 determines that the transaction to be inspected is an abnormal transaction, the analyst terminal 4100 may transmit a warning signal to the operator terminal 4200 (S4003).

When the operator terminal 4200 receives a warning signal, it can output information indicating the occurrence of an abnormal transaction and which transaction is an abnormal transaction to an output such as a display of the operator terminal 4200 (S4004). An administrator or operator using the operator terminal 4200 can take action on transactions determined to be abnormal transactions based on this information.

Figure 34 is a flowchart of a method of taking action against the consignor terminal 100 that is determined to be an abnormal transaction according to an embodiment of the present invention. Referring to FIG. 34, if the transaction to be inspected is determined to be an abnormal transaction, the analyst terminal 4100 may transmit a warning signal to the operator terminal 4200. At this time, upon receiving the warning signal, the operator terminal 4200 may transmit a transaction stop signal to the custody system 10 to stop the transaction subject to inspection (S4101). When the custody system 10 receives a transaction stop signal, it can stop the transaction subject to inspection corresponding to the transaction stop signal. In addition, when the custody system 10 receives a transaction stop signal, it can also stop or sanction other transactions by the trustor terminal 100 involved in the transaction subject to inspection.

When the operator terminal 4200 receives a warning signal, it can output information indicating the occurrence of an abnormal transaction and which transaction is an abnormal transaction to an output such as a display of the operator terminal 4200 (S4102).

An administrator or operator using the operator terminal 4200 can take action on transactions determined to be abnormal transactions based on this information. For example, the manager or operator can closely analyze whether the abnormal transaction detected through the operator terminal 4200 is normal. Additionally, when there are multiple managers or takers, decision-making by the internal monitoring and control organization may be performed through the taker terminals 4200 used by each manager or taker. The input unit 4210 can receive the results of detailed analysis or the decision-making results of the internal monitoring and control organization from the manager or operator (S4103).

The operator terminal 4200 may receive information to the effect that the transaction to be inspected has been confirmed as an abnormal transaction or information to the effect that it has been confirmed to be a normal transaction based on the results of detailed analysis or the decision-making results of the internal monitoring and control organization (S4104 ).

After transmitting a transaction stop signal, the operator terminal 4200 receives abnormality confirmation information to the effect that the input unit 4210 has confirmed that the transaction to be inspected is an abnormal transaction ('Yes' in S4104), and then responds to the transaction to be inspected. A signal to the effect of taking action in accordance with the law can be transmitted to the custody system 10. At this time, when the custody system 10 receives a signal to proceed with action in accordance with the law for the transaction subject to inspection, it can proceed with a preset process in accordance with the law for the transaction to be inspected that was stopped by the transaction stop signal. (S4105). In addition, when the custody system 10 receives a signal to proceed with action in accordance with laws and regulations for the transaction subject to inspection, the custody system 10 performs preset actions in accordance with the law for other transactions of the consignor terminal 100 involved in the transaction subject to inspection. The process can proceed.

After transmitting the transaction stop signal, the operator terminal 4200 stops the transaction to be inspected when the input unit 4210 receives normal confirmation information to the effect that the transaction to be inspected has been confirmed to be a normal transaction ('No' in S4104). A transaction resumption signal to the effect of canceling can be transmitted to the custody system 10. At this time, when the custody system 10 receives the transaction resumption signal, it can resume the progress of the transaction to be inspected that was stopped by the transaction stop signal (S4106). In addition, when the custody system 10 receives a transaction resumption signal, it can also resume processing of other transactions of the consignor terminal 100 involved in the transaction subject to inspection.

The deposit/withdrawal data receiving unit 4110, abnormal transaction analysis unit 4120, machine learning unit 4130, and input unit 4210 may include any one processor among a plurality of processors included in the abnormal transaction detection system 4000. there is. Additionally, the social network post classification method according to the embodiments of the present invention described so far and the embodiments to be described in the future may be implemented in the form of a program that can be driven by a processor.

Here, the program may include program instructions, data files, and data structures, etc., singly or in combination. Programs may be designed and produced using machine code or high-level language code. The program may be specially designed to implement the above-described abnormal transaction detection method, or may be implemented using various functions or definitions known and available to those skilled in the art in the computer software field. A program for implementing the above-described abnormal transaction detection method may be recorded on a recording medium readable by a processor. At this time, the recording medium may be the memory 4140 provided in the abnormal transaction detection system 4000.

The memory 4140 can store programs that perform the operations described above and operations described later, and the memory 4140 can execute the stored programs. When there are a plurality of processors and memories 4140, they may be integrated into one chip or may be provided in physically separate locations. The memory 4140 may include volatile memory such as Static Random Access Memory (S-RAM) or Dynamic Random Access Memory (D-Lab) for temporarily storing data. In addition, the memory 4140 includes read only memory (ROM), erasable programmable read only memory (EPROM), and electrically erasable programmable read only memory (EEPROM) for long-term storage of control programs and control data. May include non-volatile memory.

The processor may include various logic circuits and operation circuits, process data according to a program provided from the memory 4140, and generate control signals according to the processing results. The processor and memory 4140 may be provided in the server.

Meanwhile, customers of digital asset custody services may wish to receive services for virtual assets traded through a verifier that can deposit only virtual assets of a pre-fixed value.

Ethereum is an example of a virtual asset traded through a verifier that can only deposit virtual assets of a pre-fixed value. Ethereum changed from proof-of-work to proof-of-stake due to the 2.0 upgrade, and proof-of-stake through staking allows you to receive a proof-of-stake fee instead of mining. At this time, in order to deposit Ethereum, a fixed value of 32 ETH must be deposited per validator. In other words, it may be difficult for customers to deposit Ethereum that is not divisible by 32 ETH, such as 10 ETH, through the verifier.

As such, there is a problem in that it is difficult for customers to use services for virtual assets traded through a verifier that can only deposit virtual assets of a fixed value in advance unless they deposit all relatively expensive virtual assets of a fixed value. Therefore, there is a need to provide a staking service to reduce the burden of operating a validator for customers who want to deposit a small amount of virtual assets.

Figure 35 is a diagram for explaining asset staking through a custody system in a staking service providing system according to an embodiment of the present invention. Figure 36 is a diagram illustrating a system for providing staking services through a custody system according to an embodiment of the present invention.

Referring to Figures 35 and 36, the staking service provided by the staking service providing system according to one embodiment can be divided into solo staking service and pool staking service.

For example, a solo staking service can be provided for customers who want to deposit Ethereum to stake the entire 32 ETH. In addition, a pool staking service can be provided in which multiple customers raise 32 ETH by depositing only a small amount and then stake it to share the interest.

The staking service provision system may be installed on the digital asset custody service provider's server, but the staking service provision system is not necessarily limited to being provided on the service provider's server.

The staking service providing system may include a custody system 10, a smart contract 400, and a staking contract 5100. The custody system 10 provides a service for managing virtual assets to be traded through a validator, and can deposit a pre-fixed value of virtual assets to be traded for each validator. The virtual asset to be traded may be a type of virtual asset in which only a pre-fixed value can be deposited per verifier, such as the Ethereum described above.

The smart contract 400 receives transaction fees from the custody system 10 through the payment account 12, and transfers the transaction target virtual asset corresponding to the consignor terminal 100 based on the signal received from the consignor terminal 100. transactions can be performed. The smart contract 400 may receive a signal from the consignor terminal 100 indicating that it wishes to deposit a virtual asset to be traded.

Based on the signal received by the smart contract 400 from the consignor terminal 100, the staking contract 5100 can deposit a virtual asset to be traded at a pre-fixed value for each verifier.

Figure 37 is a diagram for explaining generating verifier deposit data and registering a verifier according to an embodiment of the present invention. Referring to Figure 37, the staking service providing system generates validator deposit data through the node server before depositing with the staking contract 5100, and sends the validator to the node. It can be configured to prepare a verifier by registering with .

The trustor terminal 100 may transmit a staking request signal to the custody system 10 indicating that it wishes to use the staking service. When the custody system 10 receives a staking request signal from the trustor terminal 100, it can generate validator deposit data through the node of the virtual asset being traded.

The custody system 10 can register the verifier in the node of the virtual asset being traded based on the verifier deposit data. At this time, the custody system 10 may store the verifier deposit data in the custody database 5200. The custody database 5200 may be a database that stores verifier deposit data corresponding to the consignor terminal 100 of a customer using the staking service.

Verifier deposit data may include a keystore file and deposit data. When the custody system 10 receives a staking request signal from the trustor terminal 100, it can generate a keystore file and deposit data. Afterwards, the custody system 10 can transmit the keystore file to the node of the virtual asset being traded. At this time, the custody system 10 may transmit deposit data to the consignor terminal 100.

Referring to FIGS. 35, 36, and 37, when the study system receives a staking request signal from one of the consignor terminals 100, the value of the virtual asset to be traded for which the consignor terminal 100 requested deposit is calculated per verifier. It can be determined whether it corresponds to a pre-fixed value.

The custody system 10 generates verifier deposit data corresponding to the consignor terminal 100 when the value of the virtual asset to be traded for which one consignor terminal 100 has requested deposit corresponds to a pre-fixed value for each verifier. You can. At this time, the custody system 10 may register the verifier corresponding to the consignor terminal 100 to the node of the virtual asset to be traded.

The custody system 10 is a transaction object for which at least one other consignor terminal 100 has requested a deposit if the value of the virtual asset for which one consignor terminal 100 has requested a deposit is less than a pre-fixed value for each verifier. By combining the values of virtual assets, the trustor terminals 100 can be combined so that the sum of the values of virtual assets requested by a plurality of trustor terminals 100 is a pre-fixed value for each verifier.

At this time, the custody system 10 may generate a unique number corresponding to a plurality of trustor terminals 100 that are combined so that the sum of the values of the requested virtual assets for trading becomes a pre-fixed value for each verifier.

When a unique number corresponding to a plurality of combined consignor terminals 100 is generated, the custody system 10 generates verifier deposit data corresponding to one unique number and uses the verifier corresponding to one unique number as a transaction target. You can register on the virtual asset node.

In this way, the staking service provision system according to one embodiment is such that even if the value of the virtual asset to be traded for which one consignor terminal 100 has requested a deposit is less than the pre-fixed value for each verifier, the other consignor terminal 100 has requested a deposit. By allowing one verifier to correspond to a group that is combined with the virtual assets to be traded and grouped with a single unique number, a virtual asset deposit service can also be provided to customers who wish to deposit a virtual asset worth less than a pre-fixed value per verifier. You can.

Figure 38 is a diagram for explaining staking with a payment account according to an embodiment of the present invention. Referring to Figure 38, you can see the process of staking the assets deposited by the customer in the smart contract 400 along with the validator deposit data through the payment account.

When the custody system 10 receives a staking request signal from the consignor terminal 100, it verifies the transaction target virtual asset deposited by the consignor terminal 100 in the smart contract 400 through the payment account 12. By transmitting the deposit data to the smart contract 400, the assets deposited in the smart contract 400 can be staked together with the verifier deposit data.

Figure 39 is a diagram to explain transferring assets to a customer through unstaking according to an embodiment of the present invention.

Referring to Figure 39, all or some of the customers who used the staking service can request unstaking when they want to retrieve their deposits. When staking is executed, the customer's principal and interest can be deposited into the payment account 12. At this time, a fee can be collected from the interest and then the principal and interest can be delivered to each customer.

When the smart contract 400 receives the withdrawal request signal and deposit data from the consignor terminal 100, the staking contract 5100 deposits the principal and interest of the verifier corresponding to the consignor terminal 100 into the payment account 12. It can be configured to do so.

On the other hand, if the customer who wants to unstake is a customer who deposited a small amount of virtual assets for trading through a verifier corresponding to a unique number, a separate unstaking procedure may be required.

The staking contract 5100 generates a unique number corresponding to a plurality of combined trustor terminals 100, and then the smart contract 400 connects to any one of the trustor terminals 100 corresponding to the unique number. Upon receiving the withdrawal request signal and deposit data from 100, it may be configured to deposit the principal and interest of the verifier corresponding to the unique number into the payment account 12. At this time, the smart contract 400 may be configured to distribute the principal and interest deposited into the payment account 12 to the consignor terminals 100 corresponding to the unique number.

Figure 40 is a flowchart of a control method of a staking service providing system according to an embodiment of the present invention. Referring to FIG. 40, the smart contract 400 may receive a signal from the consignor terminal 100 indicating that it wishes to deposit a virtual asset to be traded (S5001).

When the custody system 10 receives a staking request signal from the trustor terminal 100, it can generate validator deposit data through the node of the virtual asset being traded. At this time, the custody system 10 can register the verifier in the node of the virtual asset to be traded based on the verifier deposit data (S5002).

When the custody system 10 receives a staking request signal from the consignor terminal 100, it verifies the transaction target virtual asset deposited by the consignor terminal 100 in the smart contract 400 through the payment account 12. By transmitting the deposit data to the smart contract 400, the assets deposited in the smart contract 400 can be staked together with the verifier deposit data (S5003).

When the smart contract 400 receives the withdrawal request signal and deposit data from the consignor terminal 100, the staking contract 5100 deposits the principal and interest of the verifier corresponding to the consignor terminal 100 into the payment account 12. It can be configured to do so.

At this time, if the staking contract 5100 generates a unique number corresponding to a plurality of combined trustor terminals 100, the smart contract 400 uses one of the trustor terminals 100 corresponding to the unique number. Upon receiving the withdrawal request signal and deposit data from the consignor terminal 100, the principal and interest of the verifier corresponding to the unique number may be deposited into the payment account 12. At this time, the smart contract 400 may be configured to distribute the principal and interest deposited into the payment account 12 to the consignor terminals 100 corresponding to the unique number.

Figure 41 is a flowchart showing some examples of a virtual asset custody method based on the custody system 10 according to an embodiment of the present invention. As shown, the method disclosed in FIG. 41 may be performed by the custody system 10, the custody system DB 6010, the smart contract 400, and the smart contract DB 6020. Meanwhile, in FIG. 41, the custody system 10 and the custody system DB 6010 are shown as separate configurations, but they are not limited thereto. For example, the custody system DB 6010 may be included in the custody system 10. Likewise, the smart contract DB 6020 may be included in the smart contract 400.

The custody system 10 can generate a plurality of random number codes (S6012). Here, the random number code may refer to any combination of characters including uppercase and/or lowercase alphabets. Additionally, the custody system 10 can generate a token ID. Here, the token ID may refer to any combination of numbers including one or more Arabian numbers. Furthermore, the custody system 10 can generate a root seed for a Merkle tree, which will be described later, by matching a random number code and a random token ID one-to-one.

The custody system 10 can acquire a Merkle tree through Merkle tree logic using the root seed (S6014). Then, the custody system 10 may store the root seed placed at the lowest level of the Merkle tree in the custody system DB 6010 (S6016). Additionally, the custody system 10 may transmit a request to store the root of the Merkle tree to the smart contract 400 so that the root placed at the highest level of the Merkle tree is stored in the smart contract 400 (S6018).

The smart contract 400 may store the root in the smart contract DB 6020 in response to receiving a Merkle tree root storage request from the custody system 10 (S6022).

Figure 42 is a flowchart showing another example of a virtual asset custody method based on the custody system 10 according to an embodiment of the present invention. As shown, the method disclosed in FIG. 42 may be performed by the custody system 10, the custody system DB 6010, the smart contract 400, and the smart contract DB 6020. Meanwhile, in FIG. 42, the custody system 10 and the custody system DB 6010 are shown as separate configurations, but are not limited thereto. For example, the custody system DB 6010 may be included in the custody system 10. Likewise, the smart contract DB 6020 may be included in the smart contract 400. Meanwhile, the method disclosed in FIG. 42 may be performed after the method disclosed in FIG. 41.

The custody system 10 may receive a request for custody of a certain amount of assets from the consignor terminal 100 (S6024). Then, the custody system 10 can verify the customer and confirm the validity of the asset (S6026). Specifically, the custody system 10 may check whether member identification information (eg, ID) associated with the consignor terminal 100 exists in the custody system DB 6010. If member identification information exists, the custody system may match the member identification information associated with the trustor terminal 100 to one target root seed among the plurality of root seeds. Then, the member identification information and the target root seed matched with the member identification information may be recorded in the custody system DB 6010 (S6028). Additionally, the custody system 10 may request the smart contract 400 to execute a deposit function in response to determining that a certain amount of assets are valid (S6032).

The smart contract 400 can execute a deposit function based on a certain amount of assets (S6034). Then, the smart contract 400 matches a random address for depositing virtual assets to the virtual asset deposit amount, member identification information, and member identification information obtained as a result of executing the deposit function, and creates a smart contract DB (6020) ) can be saved in (S6036).

Figure 43 is a flowchart showing an example of a virtual asset withdrawal method based on the custody system 10 according to an embodiment of the present invention. As shown, the method disclosed in FIG. 43 may be performed by the custody system 10, the custody system DB 6010, the smart contract 400, and the smart contract DB 6020. Meanwhile, in FIG. 43, the custody system 10 and the custody system DB 6010 are shown as separate configurations, but they are not limited thereto. For example, the custody system DB 6010 may be included in the custody system 10. Likewise, the smart contract DB 6020 may be included in the smart contract 400. Additionally, the consignor terminal 100 and the recipient terminal 6030 may be configured as separate user terminals as shown, but may also be configured as a single user terminal.

The custody system 10 may receive a withdrawal request from the consignor terminal 100 (S6038). Then, the custody system checks whether a target root seed matching the member identification information exists in the plurality of root seeds based on the withdrawal application (S6042), and in response to determining that the target root seed exists, A request to verify withdrawal availability can be sent to the smart contract 400 (S6044).

The smart contract 400 may obtain a result of determining whether the virtual asset deposit amount is more than the virtual asset withdrawal amount based on the withdrawal request through the address matched to the member identification information and transmit it to the custody system 10.

In response to receiving a result of determining whether the virtual asset deposit amount is more than the virtual asset withdrawal amount based on the withdrawal request from the smart contract 400, the custody system 10 requests the smart contract 400 to execute a deposit function. can be transmitted (S6048). Additionally or alternatively, the custody system 10, in response to receiving a result of determining whether the virtual asset deposit amount is more than the virtual asset withdrawal amount based on the withdrawal request from the smart contract 400, sends a plurality of root seeds. Using this, you can obtain a Merkle tree proof for the target root seed from the Merkle tree. Then, the custody system 10 may transmit a request for execution of a deposit function including a Merkle tree proof to the smart contract 400.

The smart contract 400 may execute the deposit function in response to a request to execute the deposit function (S6052). The smart contract 400 can obtain the virtual asset withdrawal amount associated with the withdrawal application as a result of executing the deposit function. Then, the smart contract 400 may transmit an amount of virtual assets corresponding to the virtual asset withdrawal amount to the recipient terminal 6030.

In response to a request to execute the deposit function, the smart contract 400 checks the validity of the target root seed based on the Merkle tree proof and the root (S6054), and in response to determining that the target root seed is valid, executes the deposit function. As a result of execution, you can obtain the virtual asset withdrawal amount associated with the withdrawal application. Then, the smart contract can transmit an amount of virtual assets corresponding to the obtained virtual asset withdrawal amount to the recipient terminal 6030 (S6056).

The smart contract 400 can issue a withdrawal confirmation to the consignor terminal 100 (S6058). Specifically, the smart contract 400 may generate another token ID that is different from the token ID included in the target root seed based on the target root seed. Here, the other token ID may refer to the token ID for issuing a withdrawal confirmation certificate. The smart contract 400 can generate a withdrawal confirmation based on another token ID and transmit it to the consignor terminal 100. In this case, other token IDs can be configured so that they are not regenerated through another withdrawal application that is different from the withdrawal application. In other words, other token IDs may be configured so that they cannot be reused. For example, another token ID may consist of a Non-Fungible Token (NFT). Accordingly, the withdrawal confirmation may include an image for NFT.

Figure 44 is a flowchart showing an example of an account creation method for purchasing and storing virtual assets according to an embodiment of the present invention. As shown, the method may be executed by at least one of an enterprise server 7010, a custody system 7020, and a bank server 7030. Referring to FIG. 44, the account creation method may begin with the step of concluding a real-name account contract between the custody system 7020 and the bank server 7030 (S7012).

The corporate server 7010 may request the custody system 7020 to create a wallet (S7014). Additionally, the corporate server 7010 may request the bank server 7030, which is linked with the custody system 7020, to create a first real name account (S7016). In response, the bank server 7030 may create a first real name account for the enterprise server 7010 (S7018). In other words, the corporate server 7010 can obtain the first real name account in advance through the bank server 7030. Additionally, in response, the custody system 7020 may check whether a bank server 7030 for the enterprise server 7010 exists in the bank server 7030 (S7022).

The custody system 7020 and/or the bank server 7030 may check the corporate information of the corporate server 7010 (S7024). Specifically, the custody system 7020 may generate the first company information based on the transaction history information of the company server 7010 through the second real name account. In this case, the first company information may be updated at predetermined intervals. Similarly, the bank server 7030 may generate second company information based on transaction history information of the company server 7010 through the second real name account. In this case as well, the second company information may be updated at predetermined intervals. Additionally or alternatively, the bank server 7030 may generate second company information based on information obtained by accessing a third real-name account of another bank server 7030 associated with the bank server 7030.

The custody system 7020 and/or the bank server 7030 may check the company information of the company server 7010 by comparing at least part of each of the first company information and the second company information. Additionally or alternatively, the custody system 7020 and/or the bank server 7030 may confirm the company information of the company server 7010 based on the results of analyzing each of the first company information and the second company information. . Specifically, the custody system 7020 may calculate the first risk level and the second risk level of the corporate server 7010 based on the first company information and the second company information, respectively. Then, the custody system 7020 determines the second blindness if the first risk level is included in the first predetermined critical range and the second risk level is included in the second predetermined critical range. The account can be assigned to the corporate server (7010). Here, the first risk level and the second risk level may be updated according to each update of the first company information and the second company information.

The custody system 7020 may create a wallet for the enterprise server 7010 in response to receiving a wallet creation request including first company information from the enterprise server 7010 (S7026). Additionally or alternatively, custody system 7020 may, in response to determining that a first real name account exists at bank server 7030, provide second enterprise information of enterprise server 7010 obtained from bank server 7030. And based on the first company information, a second real name account associated with the bank server 7030 may be granted to the company server 7010.

Figure 45 is a flowchart showing an example of a method of purchasing a virtual asset using a cash deposit/withdrawal account according to an embodiment of the present invention. As shown, the method may be executed by at least one of an enterprise server 7010, a custody system 7020, a bank server 7030, and a virtual asset exchange server 7040. Referring to FIG. 45, the method of purchasing a virtual asset may begin with a step (S7028) in which cash for purchasing a virtual asset is deposited from a corporate server. However, if there is cash in the wallet of the corporate server 7010 held by the custody system 7020, the virtual asset purchase method may begin with the step of receiving a request to purchase virtual embroidery from the corporate server.

In response to the deposit of cash for purchasing virtual assets from the corporate server 7010, the custody system 7020 may request the virtual asset exchange server 7040 to create a sub-wallet of the wallet. Then, the custody system 7020 can transfer a certain amount of virtual assets stored in the sub-wallet to the wallet by purchasing them through the virtual asset exchange server 7040.

The bank server 7030 may be configured to access at least one of the second real name account and the wallet in order to obtain information related to a certain amount of virtual assets. For example, the custody system 7020 provides the bank server 7030 with an API granted access to at least one of the second real name account and the wallet, and the bank server 7030 calls the API to provide the second real name account and the wallet. It may be configured to provide real-time access to at least one of the real name account and wallet.

Figure 46 is a flowchart showing an example of a method of selling virtual assets using a cash deposit/withdrawal account according to an embodiment of the present invention. As shown, the method may be executed by at least one of an enterprise server 7010, a custody system 7020, and a virtual asset exchange server 7040. Referring to FIG. 46, the method of selling virtual assets may begin with a step (S7046) in which the corporate server 7010 requests cash withdrawal from the custody system 7020.

In response to the cash withdrawal request, the custody system 7020 may transmit the virtual asset purchase request and the virtual asset to the virtual asset exchange server 7040 (S7052). Accordingly, the virtual asset exchange server can allocate the delivered virtual asset to the sub-wallet (S7052). That is, at least some of the virtual assets stored in the wallet of the custody system 700 may be transferred to the sub-wallet of the virtual asset exchange server 7040.

The virtual asset exchange server 7040 can sell virtual assets allocated to the sub-wallet (S7054). The virtual asset exchange server 7040 may deposit cash obtained by selling virtual assets into the second real-name account of the custody system 7020. Accordingly, the custody system 7020 can store the deposited cash in the second real name account (S7058). Then, the custody system 7020 can return the cash minus the fee to the corporate server 7010 by approving the withdrawal (S7062) (S70).

Figure 47 is a configuration diagram of a custody-based digital asset (token security) issuance management system that constitutes a digital asset custody system according to an embodiment of the present invention. Referring to Figure 47, the custody-based digital asset (token security) issuance management system 8000 according to an embodiment of the present invention includes at least one consignor terminal (8000: 8100(1) to 8100(n)) and a trustee. It may include a terminal (8200) and a trustee DB (8300). However, in order to operate the custody-based digital asset (token security) issuance management system 8000, it may include fewer or more components than the components shown in FIG. 47.

In this specification, token securities are included in digital assets issued based on blockchain technology. They are issued in digital form and can be implemented as digital tokens that represent ownership of real assets. Hereinafter, digital assets will be described using the same meaning as token securities.

At least one consignor terminal (8100(1) to 8100(n)) may be a terminal related to a consignor that entrusts the issuance, storage, and management of digital assets to the trustee terminal (8200). For example, the consignor terminal (8100(1) to 8100(n)) may be a terminal belonging to an organization (consignor) such as a company, institution, or organization that entrusts the issuance, storage, and management of digital assets.

At least one consignor terminal (8100(1) to 8100(n)) may generate an issuance and distribution plan for digital assets and provide it to the trustee terminal (8200). At this time, the digital asset issuance and distribution plan includes a management plan that describes the consignor's issuance and distribution authority for the issuance of digital assets, the type and quantity of digital assets, and distribution conditions, including the amount and distribution period of digital assets. can do.

At least one consignor terminal (8100(1) to 8100(n)) may be provided as a terminal such as, for example, a desktop PC, laptop, notebook, smartphone, or smart pad, but is not limited thereto.

In an embodiment of the present invention, the digital asset that is entrusted by the consignor to the trustee may include one or more virtual currencies selected from virtual currencies including Bitcoin, Ethereum, and Klaytn.

The trustee terminal 8200 can receive an issuance and distribution plan for digital assets from at least one trustee terminal 8100(1) to 8100(n) and verify the trustor's authority to issue and distribute digital assets.

The trustee terminal (8200) is entrusted with the issuance, storage, and management of digital assets from at least one trustee terminal (8100(1) to 8100(n)) and manages the digital asset from at least one trustee terminal (8100(1) to 8100(n)). ) can provide custody services that issue, store, and manage digital assets requested from.

When an event including the issuance and distribution of digital assets occurs, the trustee terminal 8200 may provide event information related to the digital asset to one of at least one recipient terminal 8400 or 8500.

Alternatively, when an event including the issuance and distribution of digital assets occurs, the trustee terminal 8200 may provide event information related to the digital asset to at least one of the recipient terminals 8400 and 8500.

The trustee database (DB, 8300) stores information on the issuance and distribution plan of digital assets entrusted from at least one trustee terminal (8100(1) to 8100(n)), and digital asset wallet information that stores digital assets. You can.

At least one recipient terminal 8400 may be a terminal owned by a securities company or exchange for trading digital assets. At least one recipient terminal 8400 may be a terminal owned by a financial authority or public disclosure company for monitoring digital asset transactions.

Figure 48 is a flowchart of a control method of a custody-based digital asset (token security) issuance management system that constitutes a digital asset custody system according to an embodiment of the present invention. FIG. 49 is a conceptual diagram for explaining steps (S8100, S8200) of FIG. 48. Hereinafter, the description will be made with reference to FIGS. 48 and 49 together. The consignor terminal (8100, A) can create an issuance and distribution plan for digital assets and provide it to the trustee terminal (8200) corresponding to the custody system. The trustee terminal (8100, A) can entrust the issuance, storage, and management of digital assets (or token securities) (a) to the trustee terminal (8200) corresponding to the custody system.

Specifically, the consignor terminal (8100, A) can access the custody system of the consignor terminal (8200) through the network and enter into an affiliate contract. The consignor terminal (8100, A) can register as a member and create a wallet address for the digital asset (or token security) (a).

The consignor terminal (8100, A) is the consignor (A)'s issuance and distribution authority for the issuance of digital assets (or, token securities) (a), and digital assets (or token securities) (a) to the trustee terminal (8200). An issuance and distribution plan can be provided that includes a management plan that describes the type and quantity of digital assets (or token securities) (a) and distribution conditions, including the distribution volume and distribution period (S8100).

The consignor terminal (8100, A) can entrust the trustee terminal (8200) with the issuance, storage, and management of the digital asset (or token security) (a) included in the issuance and distribution plan. The trustee terminal (8200) corresponding to the custody system uses the pre-stored verification logic (8201) to determine the issuance and distribution authority for the digital asset (or token security) (a) of the consignor (A) from the issuance and distribution plan. Verification can be performed (S8200).

Figure 50 is a conceptual diagram for explaining steps (S8200, S8300) of Figure 48. Hereinafter, the description will be made with reference to FIGS. 48 and 50 together. If the trustor's (A) issuance and distribution authority for the digital asset (or token security) (a) is verified by the pre-stored verification logic (8201), the trustee terminal (8200) corresponding to the custody system is in accordance with the management plan. Digital assets (or token securities) (a) can be issued corresponding to the type and quantity of the included digital asset (or token securities) (a).

In addition, the trustee terminal (8200) corresponding to the custody system monitors the distribution conditions of the digital asset (or token security) (a) according to the distribution amount and distribution time included in the management plan and at least one recipient terminal (8400, 8500) ), a target recipient terminal (e.g., 8400) to receive event information can be set and stored in the smart contract 8210 (S8300). If the trustor's (A) issuance and distribution authority for the digital asset (or token security) (a) is not verified by the pre-stored verification logic (8201), the trustee terminal (8200) corresponding to the custody system (or, token securities) (a) may not be issued.

FIG. 51 is a conceptual diagram for explaining step S8400 of FIG. 48. Hereinafter, the description will be made with reference to FIGS. 48 and 51 together. The trustee terminal 8200 issues digital assets (or token securities) (a) to any one of the at least one recipient terminal (8400, 8500) (hereinafter referred to as the target recipient terminal). Event information related to distribution can be provided (S8400).

Specifically, the consignee terminal (8200) provides information about the type and quantity of digital assets, distribution conditions according to distribution volume and distribution time, and event information from the consignor terminal (8100, A) to the recipient terminal (8400 or 8500). Information can be provided in advance.

When an event occurs, the consignee terminal (8200) provides information on the type, quantity, distribution volume and distribution time of the digital asset provided in advance by the consignor terminal (8100, A), and the recipient terminal to receive event information. It can be determined whether the information on the type, quantity, distribution volume and distribution period of digital assets (or token securities) (a) included in the information and management plan, and information on the target recipient terminal are consistent.

Information on the recipient terminal to receive distribution conditions and event information according to the type, quantity, distribution volume, and distribution time of the digital asset provided in advance from the consignor terminal (8100, A), and the digital assets included in the management plan (or, If the information on the type, quantity, distribution volume and distribution period of the token security (a), and the information on the target recipient terminal are consistent with each other, the digital asset (or token security) stored in the smart contract (8210) (a) ) can be provided to the recipient terminal in real time (S8400).

Information on the recipient terminal to receive distribution conditions and event information according to the type, quantity, distribution volume, and distribution time of the digital asset provided in advance from the consignor terminal (8100, A), and the digital assets included in the management plan (or, If the information on the type, quantity, distribution volume and distribution period of the token security (a), and the information on the target recipient terminal do not match, the digital asset (or token security) stored in the smart contract (8210) ( Event information for a) may not be provided to the recipient terminal in real time.

As described above, the custody-based digital asset (token security) issuance management system 8100, which is an embodiment of the present invention, allows a company (or consignor) and a securities company or exchange (or recipient) to issue digital assets including coins, etc. , breaking away from the existing method of directly performing storage and management, the issuance, storage, and management of digital assets can be relayed between a company (or consignor) and a securities company or exchange (or beneficiary).

The custody-based digital asset (token security) issuance management system 8100, which is an embodiment of the present invention, issues and distributes digital assets of the company (or consignor) based on the issuance and distribution plan provided by the company (or consignor). Verification of authority can be performed.

In addition, the custody-based digital asset (token security) issuance management system 8100, which is an embodiment of the present invention, can achieve transparency in transaction and management of digital assets by performing issuance, storage, and management of digital assets.

Figure 52 is a configuration diagram of a custody-based digital asset (token security) distribution and trading management system that constitutes a digital asset custody system according to an embodiment of the present invention. Referring to Figure 52, the custody-based digital asset (token securities) distribution and trading management system 9000 includes at least one consignor terminal (9100: 9100(1) ~ 9100(n)), a trustee terminal (9200), It may include a trustee DB (9300), at least one recipient terminal (9400), and at least one user terminal (9500).

However, in order to operate the custody-based digital asset (token securities) distribution and trading management system 9000, it may include fewer or more components than the components shown in FIG. 52. In this specification, token securities are included in digital assets issued based on blockchain technology. They are issued in digital form and can be implemented as digital tokens that represent ownership of real assets. Hereinafter, digital assets will be described using the same meaning as token securities.

At least one consignor terminal (9100: 9100(1) to 9100(n)) may be a terminal related to the consignor that entrusts the purchase and sale of digital assets to the trustee terminal (9200). At least one consignor terminal (9100: 9100(1) ~ 9100(n)) may be, for example, a terminal belonging to a group (consignor) such as a company, institution, or organization that entrusts the purchase and sale of digital assets. there is.

At least one consignor terminal (9100: 9100(1) to 9100(n)) may generate an issuance and distribution plan for digital assets and provide it to the trustee terminal (9200). At this time, the digital asset issuance and distribution plan contains a management plan that describes the trustor's issuance and distribution authority for the issuance and distribution of digital assets, the type and quantity of digital assets, and distribution conditions, including the distribution volume and distribution period of digital assets. It can be included.

At least one consignor terminal (9100: 9100(1) to 9100(n)) may be provided as a terminal such as, for example, a desktop PC, laptop, notebook, smartphone, or smart pad, but is not limited thereto. In an embodiment of the present invention, the digital asset that is entrusted by the consignor to the trustee may include one or more virtual currencies selected from virtual currencies including Bitcoin, Ethereum, and Klaytn.

At least one consignor terminal (9100: 9100(1) to 9100(n)) is responsible for the purchase or sale of digital assets provided from at least one user terminal 9500 and at least one recipient terminal 9400, which will be described later. Approval may be instructed to the trustee terminal 9200. The trustee terminal 9400 may receive in advance an issuance and distribution plan for digital assets from at least one trustee terminal 9100 (9100(1) to 9100(n)).

The trustee terminal 9400 can issue digital assets according to the digital asset issuance and distribution plan provided in advance. The trustee terminal 9400 may verify the approval instruction provided from at least one trustee terminal 9100 based on the digital asset issuance and distribution plan provided in advance and sell or purchase the digital asset.

When an event such as sale or purchase of a digital asset occurs, the trustee terminal 9400 may provide event information related to the digital asset to at least one recipient terminal 9400 in real time. The trustee database (DB, 9300) contains information on the issuance and distribution plan of digital assets provided by at least one trustee terminal (9100: 9100(1) ~ 9100(n)), and information on changes due to the sale or purchase of digital assets. , digital asset wallet information that stores digital assets, etc. can be stored.

At least one recipient terminal 9400 may be a terminal owned by a securities company or exchange for trading digital assets. Alternatively, it may be a terminal owned by financial authorities or public disclosure companies to monitor digital asset transactions. At least one recipient terminal 9400 can receive event information related to digital assets from the trustee terminal 9200 in real time and announce it.

For example, at least one recipient terminal 9400 may determine the type and quantity of digital assets stored by the trustee terminal 9200, the consignor's issuance authority and distribution authority, the amount and distribution period of digital assets, and the purchase of digital assets. Alternatively, event information including changes due to sales can be disclosed.

At least one user terminal 9500 may request the purchase or sale of a digital asset from at least one recipient terminal 9400 based on event information announced on the at least one recipient terminal 9400. At least one user terminal 9500 may purchase digital assets from the trustee terminal 9200 or sell digital assets to the trustee terminal 9200 according to the results of verification of instructions performed by the trustee terminal 9200.

Figure 53 is a flowchart of a control method of a custody-based digital asset (token security) distribution and trading management system that constitutes a digital asset custody system according to an embodiment of the present invention. FIG. 54 is a conceptual diagram for explaining steps (S9100 to S9400) of FIG. 53. Hereinafter, the description will be made with reference to FIGS. 53 and 54 together.

At least one recipient terminal (9400, C) receives from the trustee terminal (9200) the type and quantity of digital assets (or token securities) (a) stored by the trustee terminal (9200), the issuance authority of the consignor (B), and Receive event information in real time, including distribution rights, distribution volume and distribution period of digital assets (or token securities) (a), and changes due to purchase or sale of digital assets (or token securities) (a) Can be disclosed (S9100).

At least one user terminal (9400, A) requests the at least one recipient terminal (9400, C) to purchase or sell a digital asset based on event information announced in the at least one recipient terminal (9400, C). (S9200). At least one recipient terminal (9400, C) sends at least one consignor terminal (9100, You can request the purchase or sale of digital assets to B) (S9200).

At least one trustee terminal (9100, A) may instruct the trustee terminal (9200) to approve the purchase or sale of digital assets in response to the request (S9300). The trustee terminal 9200 verifies the instruction using the pre-stored verification logic 9201 (see FIG. 55), and according to the verification result, digital assets (or token securities) (a) are received from at least one user terminal 9400 (A). You can purchase or sell digital assets (or token securities) (a) to at least one user terminal (9400, A) (S9400).

Figure 55 is a conceptual diagram for verifying instructions and explaining the purchase or sale of digital assets according to an embodiment of the present invention. The trustee terminal 9200 can use the pre-stored verification logic 9201 to verify the issuance authority and distribution authority for the trustee's digital assets from the issuance and distribution plan. The trustee terminal (9200) does not purchase digital assets from the user terminal (9400, A) or provides digital assets to the user terminal (9400, A) if verification of the issuance authority and distribution authority for the digital asset of the consignor has not been completed. You may not sell.

The trustee terminal 9200 may receive announced event information in real time from at least one recipient terminal 9500 (C). The trustee terminal (9200) has completed the verification of the consignor's issuance authority and distribution authority, and has event information and issuance and distribution plan in which the type and quantity of the digital asset included in the request of the user (A) and the purchase or sale period are disclosed. If it all matches the information about the digital asset included in the user (A), the digital asset corresponding to the type and quantity can be purchased or sold from at least one user terminal (A) at the time of purchase or sale included in the request of the user (A). You can.

The trustee terminal (9200) has completed the verification of the consignor's issuance authority and distribution authority, and has announced the type and quantity of digital assets included in the user's (A) request, as well as the purchase or sale time, and among the events and issuance and distribution plan. If it does not match the information about the digital asset included in at least one of the digital assets, the digital asset may not be purchased from at least one user terminal (A) or the digital asset may not be sold to at least one user terminal (A).

The trustee terminal 9200 may provide event information that occurs according to the purchase or sale of digital assets to at least one recipient terminal 9500 (C) in real time. At least one recipient terminal (9500, C) can receive and announce event information in real time, which is shared and confirmed with at least one consignor terminal (9100, B) and at least one user terminal (9400, A). It can be.

As described above, the custody-based digital asset (token securities) distribution and sales management system 9100, which is an embodiment of the present invention, purchases or sells digital assets between a customer (or user) and a company (or consignor), or allows customers to purchase or sell digital assets. Rather than the existing method of purchasing or selling digital assets between a customer (or user) and a securities company or exchange (or recipient), the purchase or sale of digital assets is conducted between a customer (or user) and a securities company or exchange (or recipient). It can be relayed.

The custody-based digital asset (token securities) distribution and sale management system 9100, which is an embodiment of the present invention, receives approval from the company (or consignor) in response to the purchase or sale of digital assets requested by the customer (or user). The instructions can be verified. Specifically, the authority to approve the purchase or sale of a company's (or consignor's) digital assets can be verified.

In addition, the custody-based digital asset (token securities) distribution and trading management system 9100, which is an embodiment of the present invention, receives event information announced by a securities company or exchange (or recipient) in real time and provides information to customers (or users). Transparency in transactions and management can be achieved by purchasing or selling digital assets only if the request matches both the event information and the issuance and distribution plan for digital assets provided in advance.

In the above, the digital asset custody system and the digital asset management method performed by the same according to various embodiments of the present invention have been described with reference to the drawings. Among the various embodiments described, two or more embodiments or two or more functions/services may be integrated and implemented within a digital asset custody system.

The embodiments described above may be implemented with hardware components, software components, and/or a combination of hardware components and software components. For example, the devices, methods, and components described in the embodiments may include, for example, a processor, a controller, an Arithmetic Logic Unit (ALU), a Digital Signal Processor, a microcomputer, and a Field Programmable Gate (FPGA). It may be implemented using one or more general-purpose computers or special-purpose computers, such as an array, PLU (Programmable Logic Unit), microprocessor, or any other device that can execute and respond to instructions.

The processing device may execute an operating system and one or more software applications that run on the operating system. Additionally, a processing device may access, store, manipulate, process, and generate data in response to the execution of software. For ease of understanding, a single processing device may be described as being used; however, those skilled in the art will understand that a processing device may include multiple processing elements and/or multiple types of processing elements. You will understand that it can be included.

For example, a processing device may include multiple processors or one processor and one controller. Additionally, other processing configurations, such as parallel processors, are also possible. Software may include a computer program, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired, or to process independently or collectively. You can command the device.

Software and/or data may be used on any type of machine, component, physical device, virtual equipment, computer storage medium or device to be interpreted by or to provide instructions or data to a processing device. It can be embodied in . Software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. Computer-readable media may include program instructions, data files, data structures, etc., singly or in combination. Program instructions recorded on the medium may be specially designed and configured for the embodiment or may be known and available to those skilled in the art of computer software.

Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CDROMs and DVDs, and ROM, RAM, and flash memory. Includes hardware devices specifically configured to store and execute program instructions, such as: Examples of program instructions include machine language code, such as that produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

As described above, although the embodiments have been described with limited examples and drawings, various modifications and variations can be made by those skilled in the art from the above description. For example, the described techniques are performed in a different order than the described method, and/or components of the described system, structure, device, circuit, etc. are combined or combined in a different form than the described method, or other components are used. Alternatively, appropriate results may be achieved even if substituted or substituted by an equivalent. Therefore, other implementations, other embodiments and equivalents of the claims also fall within the scope of the following claims.

Claims (10)

In the digital asset management method of the digital asset custody system performed by a computing device,
Receiving a digital asset management request from the customer company server;
In response to the digital asset management request, creating a digital asset wallet for digital asset management; and
Including providing a digital asset custody service using the digital asset wallet,
The step of providing the digital asset custody service is,
Steps in which digital assets are entrusted to customers;
A step in which digital asset transaction data for each exchange is collected;
Based on the collected digital asset transaction data, predicting the price increase rate and accuracy of digital assets by period based on machine learning; and
A digital asset management method of a digital asset custody system, including the step of operating digital assets entrusted from the customer based on predicted results based on machine learning. According to claim 1,
The step of operating the entrusted digital assets is,
Analyzing customer investment tendencies through customer surveys; and
A digital asset management method of a digital asset custody system, comprising the step of configuring the customer's digital asset portfolio based on the customer's investment propensity and the results predicted based on the machine learning. According to clause 2,
The step of predicting the price increase rate and accuracy of the digital asset by period is,
Preprocessing the digital asset transaction data collected for each of a plurality of exchanges;
Extracting investment factors from the digital asset transaction data;
Performing machine learning-based time series data learning based on the extracted investment factors;
Extracting a prediction model based on the results of the time series data learning; and
A digital asset management method of a digital asset custody system, comprising predicting investment factors for digital assets using the prediction model. According to clause 3,
The investment factors include forecast accuracy, price increase rate, price stability, and appropriate investment amount. A digital asset management method of a digital asset custody system. According to clause 4,
The step of predicting the price increase rate and accuracy of the digital asset by period is,
Using training transaction data collected for each exchange, generating a plurality of different artificial intelligence models based on different machine learning algorithms;
Inputting transaction data collected for each of the plurality of exchanges into the plurality of different artificial intelligence models, and obtaining the price increase rate and accuracy of digital assets by period from the output of the plurality of artificial intelligence models according to the transaction data. step; and
Combining the period-specific price increase rate and accuracy of digital assets output from each of the plurality of artificial intelligence models based on the weight set for each artificial intelligence model, and finally predicting the period-specific price increase rate and accuracy of the digital asset for each exchange. A digital asset management method of a digital asset custody system, including. According to clause 5,
The step of predicting the price increase rate and accuracy of the digital asset by period is,
A digital asset management method of a digital asset custody system, further comprising determining the weight corresponding to each artificial intelligence model based on accuracy based on training transaction data of each of the artificial intelligence models. According to clause 6,
The plurality of different artificial intelligence models are,
LGBM (Light Gradient Boosting Machine) artificial intelligence algorithm, RNN (recurrent neural network), a recursive recurrent artificial neural network based on Long Short Term Memory (LSTM), fully connected layer-based ANN (artificial neural network), K- KNN (K-Nearest Neighbors) algorithm, which is a nearest neighbors artificial neural network, K-Means Clustering algorithm, PCA (Principal Component Analysis) algorithm, SVM (Support Vector Machine) artificial neural network based on the support vector machine algorithm, and a digital asset management method of a digital asset custody system, including both an Auto Encoder artificial neural network. According to clause 7,
The step of predicting the price increase rate and accuracy of the digital asset by period is,
The LGBM (Light Gradient Boosting Machine) artificial intelligence algorithm, the RNN (recurrent neural network), which is a recursive recurrent artificial neural network based on the Long Short Term Memory (LSTM), and the fully connected layer-based ANN (artificial neural network) , the KNN (K-Nearest Neighbors) algorithm, which is the K-nearest neighbors artificial neural network, the K-Means Clustering algorithm, the PCA (Principal Component Analysis) algorithm, and the SVM (based on the support vector machine algorithm) For training each of the artificial intelligence models on the period price increase rate and accuracy of digital assets output from a plurality of different artificial intelligence models, including both the Suport Vector Machine (Support Vector Machine) artificial neural network and the Auto Encoder artificial neural network. A digital asset management method in a digital asset custody system, comprising the step of reflecting and combining the weights determined based on accuracy based on transaction data to finally predict the price increase rate and accuracy by period of digital assets for each exchange. A computer-readable recording medium on which a program for executing the digital asset management method of the digital asset custody system according to any one of claims 1 to 8 is recorded. Receive a digital asset management request from the customer server;
In response to the digital asset management request, create a digital asset wallet for digital asset management;
We provide digital asset custody services using the digital asset wallet,
We collect digital asset transaction data by exchange, and based on the collected digital asset transaction data, predict the price increase rate and accuracy of digital assets by period based on machine learning.
A digital asset custody system that operates digital assets entrusted to customers based on predicted results based on machine learning.