JPWO2020141584A1 - Processing system and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve a dilemma in which the guarantee of authenticity of recorded information and the guarantee of the right to delete the information are in conflict with each other. Information is double-encrypted EKA (EKB (information)) and recorded in a blockchain with two half-split common keys KA and KB. The information holder distributes one half-split common key KB to the information requester and decrypts the double-encrypted information with the half-split key KA (EKA (EKB (information))) Information EKB (information). To send. The information requester calculates DKB (EKB (information)) to obtain plaintext information. When the information holder updates the half-split common key KA to R, a DR (EKA (EKB (information))) is transmitted to the information requester, and the information requester decrypts the DKB (DR (EKA (EKB)) with the KB. Information))) Even if it is not possible to obtain plain text. [Selection diagram] Fig. 14

Description

The present invention relates to a processing system and a program for an information recording method that is difficult to falsify or erase, such as a blockchain.
Regarding.

Blockchain has been generally known as an information recording method that is difficult to falsify. For example, Patent Document 1 records various information related to freight transportation using this blockchain.

JP-A-2018-128723

However, recording information using such a blockchain is not only difficult to falsify, but also difficult to erase (hereinafter referred to as "non-erasable"). As a result, once personal information is recorded using the blockchain, even if the personal information owner wants to delete the personal information, it cannot be deleted and the right to erase personal information (so-called right to be forgotten) is impaired. There is a drawback that it is.

In other words, there is a drawback that there is a dilemma in which the guarantee of the authenticity of the recorded information and the guarantee of the right to delete the information are antinomy.

The present invention has been conceived in view of such circumstances, and an object of the present invention is to solve a dilemma in which the guarantee of the authenticity of recorded information and the guarantee of the right to delete the information are antinomy.

The present invention provides an encryption means that performs an encryption process that encrypts information to be recorded.
A recording means for recording information after the encryption process, and
Decoding means that performs decoding processing on the information recorded by the recording means using the first key and the second key to obtain plain text information.
The information recorded by the recording means is provided with an undecryptable means for making the information undecryptable, which cannot be decoded.
The decryption means includes a second key concealment holding means for concealing and holding the second key.
The undecryptable means puts the undecryptable state by updating the second key held by the second key concealment holding means with another key.

Preferably, the decryption means further includes a first key distribution means that distributes the first key to a person who wants to view the information.
Preferably, the search means for searching the information recorded by the recording means in plain text is further provided.

More preferably, the information recorded by the recording means includes personal information.
The undecryptable means puts the personal information of the personal information owner into the undecryptable state at the request of the personal information owner.

Another aspect of the present invention is a step of performing an encryption process for encrypting information to be recorded.
A decryption step of decrypting the information recorded by the recording means for recording the information after the encryption process using the first key and the second key to obtain plain text information.
The step of making the information recorded by the recording means undecipherable, which cannot be decoded,
Let the computer run
The decryption step includes a step of keeping the second key secret.
The step of making the decryption impossible state makes the second key held by the holding step update to another key to make the decryption state impossible.

According to the present invention, the dilemma in which the guarantee of the authenticity of the recorded information and the guarantee of the right to delete the information conflict with each other can be solved as much as possible.

It is a system diagram which shows the whole structure of a processing system. (A) is a diagram for explaining the information stored in the HDD of the user terminal constituting the node of the blockchain, and (B) is a diagram for explaining the information stored in the personal information DB of the authorized business operator. be. (A) is a flowchart showing a main routine program of a user terminal of a public chain, and (B) is a flowchart showing a subroutine program of personal information recording processing and a flowchart of a server of an authorized business operator. It is a flowchart which shows the subroutine program of smart contract processing executed in the user terminal of a public chain. (A) is a flowchart showing the main routine program of the user terminal of the private chain, and (B) is the flowchart showing the subroutine program of the personal information search process. It is a flowchart which shows the subroutine program of smart contract processing executed in the user terminal of a private chain. (A) is a continuation of the flowchart showing the subroutine program of smart contract processing executed on the user terminal of the private chain, and (B) is the flowchart showing the subroutine program of machine learning processing executed on the user terminal of the private chain. be. It is a flowchart which shows the subroutine program of the AI smart contract generation processing executed in the user terminal of a private chain. (A) is a flowchart showing a subroutine program of simulation learning processing executed on a user terminal of a private chain, and (B) is a flowchart showing a subroutine program of AI smart contract group generation processing executed on a user terminal of a private chain. Is. It is a flowchart which shows the subroutine program of smart contract trust contract processing executed by the user terminal of a private chain. (A) is a flowchart showing a subroutine program of reinforcement learning processing of a personalized AI smart contract trained model, and (B) is a flowchart showing a main routine program executed by a user terminal of a consortium chain. (A) is a flowchart showing a subroutine program of IoT sensor data aggregation processing executed by a user terminal of the consortium chain, and (B) is a flowchart showing a subroutine program of simulation processing executed by a user terminal of the consortium chain. .. (A) is a flowchart showing a subroutine program of AI smart contract group generation processing executed by the user terminal of the consortium chain, and (B) is a flowchart showing a subroutine program of smart contract processing executed by the user terminal of the consortium chain. Is. It is an explanatory diagram that makes the information recorded by using the blockchain inaccessible, (A) is a diagram in a normal state that can be viewed, and (B) is a diagram that makes it unreadable so that it cannot be viewed. Is. It is a figure explaining the information stored in the HDD of the user terminal which constitutes the node of a blockchain. It is a flowchart which shows the main routine program of the user terminal of a public chain and the user terminal of a private chain. (A) is a flowchart showing a subroutine program of personal information recording processing on the blockchain, and (B) is a flowchart showing a subroutine program of record unbreakable processing. It is a flowchart which shows the subroutine program of the personal information acquisition processing and the personal information provision processing. (A) is a diagram for explaining the information stored in the HDD of the user terminal constituting the node of the blockchain, and (B) is a diagram for explaining the information stored in the personal information DB of the authorized business operator. be. It is a flowchart which shows the main routine program of the user terminal of a public chain, the server of an authorized business operator, and the user terminal of a private chain. It is a flowchart which shows the subroutine program of the personal information recording processing and the hash value recording processing in a blockchain. It is a flowchart which shows the subroutine program of the record unbreakable processing. It is a flowchart which shows the subroutine program of the personal information provision process, the personal information acquisition process, and the ciphertext transmission process. It is a system diagram which shows the whole structure of a processing system. It is a flowchart which shows the main routine program of a user terminal of a public chain, a server of a key registration center, and a user terminal of a private chain. (A) is a flowchart showing a subroutine program of personal information recording processing and key registration processing on the blockchain, and (B) is a flowchart showing a subroutine program of record decipherability request processing and record decipherability processing. It is a flowchart which shows the subroutine program of one-sided common key provision processing, data acquisition processing, and data decoding processing. It is explanatory drawing of the mirror world as a simulation environment. It is a figure which shows a concrete example of an urban digital twin in a mirror world. It is a flowchart of the main routine of a mirror world server and a user terminal. It is a flowchart which shows the subroutine program of the generation sale process of personal AI. It is a flowchart which shows the subroutine program of the simulation preparation processing and the simulation preparation response processing. It is a flowchart which shows the subroutine program of simulation processing. (A) is a schematic diagram of a multi-service DAO construction system, and (B) is a flowchart of a main routine between a mirror world server and a user terminal. It is a flowchart which shows the subroutine program of the simulation reinforcement learning preparation processing and the simulation reinforcement learning preparation response processing. It is explanatory drawing which registers a multi-service DAO digital twin in a mirror world as a simulation target. It is a schematic system diagram of simulation reinforcement learning of multi-service DAO. It is a flowchart which shows the subroutine program of DAO agent reinforcement learning processing. (A) is a diagram showing a reward table stored as knowledge by the DAO agent, and (B) is a flowchart showing a subroutine program of the persona agent reinforcement learning process. (A) is a flowchart showing a subroutine program of the idea proposal service execution process, and (B) is a flowchart showing a subroutine program of the improvement plan service execution process. (A) is a flowchart showing a subroutine program of commercialization service execution processing, and (B) is a flowchart showing a subroutine program of infringement handling service execution processing. (A) is a flowchart showing a subroutine program of token purchase service execution processing, and (B) is a diagram explaining price fluctuations in a token fluctuation market due to token purchase by a persona agent group. It is a figure which shows the element integrated DAO construction system. (A) is a flowchart of the main routine between the mirror world server and the user terminal, and (B) is a flowchart showing a subroutine program of simulation reinforcement learning preparation processing and simulation reinforcement learning preparation response processing. It is explanatory drawing which registers the element integrated DAO digital twin in a mirror world as a simulation target. The figure shows the schematic system which shows the simulation reinforcement learning of the element integrated DAO digital twin. (A) is a diagram showing a performance and distribution rate calculation algorithm stored as knowledge by the material procurement element agent, and (B) is a diagram showing a performance and distribution rate calculation algorithm stored as knowledge by the assembly element agent. It is a figure which shows, (C) is a figure which shows the calculation algorithm of the performance and distribution rate which the advertising element agent has memorized as knowledge. (A) is a diagram showing a performance and distribution rate calculation algorithm stored as knowledge by the sales element agent, and (B) is a diagram showing a reward table stored as knowledge by the general agent. It is a flowchart which shows the subroutine program of the simulation reinforcement learning process. It is a flowchart which shows the subroutine program of the integrated agent reinforcement learning process. It is a flowchart which shows the subroutine program of the material procurement element agent reinforcement learning process. (A) is a flowchart showing a subroutine program of information collection processing by a crawler, and (B) is a diagram showing various data stored in a material procurement DB. It is a flowchart which shows the subroutine program of the assembly element agent reinforcement learning process. It is a flowchart which shows the subroutine program of the promotion element agent reinforcement learning process. (A) is a flowchart showing a subroutine program of information collection processing by a crawler, and (B) is a diagram showing various data stored in a promotion DB. It is a flowchart which shows the subroutine program of the sales element agent reinforcement learning process. (A) is a flowchart showing a subroutine program of information collection processing, and (B) is a diagram showing various data stored in a sales DB. (A) is a flowchart showing a subroutine program of the personal AI reinforcement learning process in charge of material procurement, and (B) is a flowchart showing a subroutine program of the personal AI reinforcement learning process in charge of assembly. (A) is a flowchart showing a subroutine program of the personal AI reinforcement learning process in charge of advertising, and (B) is a flowchart showing a subroutine program of the personal AI reinforcement learning process in charge of sales. It is explanatory drawing which shows the installation method of a program.

[First Embodiment]
The first embodiment of the present invention will be described with reference to FIGS. 1 to 13. First, referring to the entire system of FIG. 1, three types of blockchain networks, a private chain 2, a consortium chain 3, and a public chain 4, are connected to the centralized oracle 21. Public chain 4 is a completely open mechanism, and any individual or group can trade there. Transactions can be effectively confirmed on the blockchain. Mining (competition of bookkeeping rights) is also free and anyone can participate. The consortium chain 3 is a blockchain that can be used only by partners belonging to associations and unions. The people (each node) among them are designated as bookkeepers. The generation of blocks is also decided in advance, and other people (nodes) can trade but do not have the right to book. Private chain 2 only records with blockchain technology, and the bookkeeping right is not open but monopolized by an individual or a company, and only internal transactions are recorded. Polkadot is used to connect blockchains and exchange tokens and data between blockchains. Polkadot is a blockchain for connecting different blockchains. Blockchains developed using Substrate can be connected to Polkadot, and by connecting to Polkadot, tokens and data can be exchanged with other blockchains connected to Polkadot.

The centralized Oracle 21 is a system that bridges data between the blockchain and the Internet 1, collects various information scattered on the Internet connected to the Internet 1, and provides information to the blockchain smart contract. I do.

Each node 19 of the private chain 2, the consortium chain 3, and the public chain 4 is composed of user terminals such as a personal computer (hereinafter referred to as "PC") 16. This PC (hereinafter, also referred to as "user terminal") 16 is connected to the Internet 1. Further, the server 20 of the SNS (Social Networking Service) 40 and the server 18 of the blockchain authorized business operator 17 are connected to the Internet 1. The server 18 of the certified business operator 17 may participate in the blockchain as a node 19. Further, the server of the certificate authority that issues the digital certificate in PKI (Public Key Infrastructure) may be connected to the Internet 1.

The authorized business operator 17 keeps personal information, issues an electronic ID mainly to the personal quasi-information, and records the hash value of the personal information in the blockchain. The entrusted personal information is stored in the personal information database (hereinafter referred to as "personal information DB") 29. The certified business operator 17 may participate in the blockchain as a node 19.

The PC 16 includes a CPU (Central Processing Unit) 10 as a control center, a RAM (Random Access Memory) 9 that functions as a work area of the CPU 10, a ROM (Read Only Memory) 11 that stores data and programs, and an HDD (hard disk). It is composed of a storage unit such as drive) 12, an input operation unit 7 such as a display and a keyboard, a communication unit 5, a display unit 6, an interface 8, a bus 13, and various other hardware. Various servers such as the server 20 and the server 18 are also configured by the same hardware as the PC 16, and the repetition of illustration and description is omitted here. As the storage unit, an SDD (Solid State Drive) may be used in addition to or instead of the HDD.

An IoT (Internet of Things) device 14 and a wireless sensor network 15 are connected to the node 19 of the consortium chain 3. The sensor signals from the IoT device 14 and the wireless sensor network 15 are input to the node 19, and the drive signal of the IoT device 14 is output from the node 19. The IoT device 14 is various sensors, actuators, and the like for IoT.

The wireless sensor network 15 is a wireless network in which a plurality of wireless terminals with sensors are scattered in a space, and they can cooperate with each other to collect an environment or a physical condition. For example, a sensor device is made of energy harvesting, M2M, or a battery, and for example, deterioration of metal fatigue is constantly monitored by a pressure sensor or gauge sensor, and the change is notified. It is mainly installed in buildings such as bridges and tunnels. Generally, it includes a plurality of sensor nodes and a gateway sensor node. These nodes usually consist of one or more sensors, a wireless chip, a microprocessor, a power source (such as a battery). Wireless sensor networks typically have an ad hoc function and a routing algorithm for sending data from each node to a central node. In other words, there is a function to autonomously reconstruct another communication path when communication between nodes fails. Since the nodes cooperate as a group, there is also an element of distributed processing. In addition, there is also a function that operates for a long period of time without receiving power supply from the outside, and therefore has a power saving function or a self-power generation function.

In the present embodiment, the IoT device 14 and the wireless sensor network 15 are connected to the consortium chain 3 via the node 19, but one or both of the IoT device 14 and the wireless sensor network 15 do not go through the node 19. It may itself be part of node 19 of the consortium chain 3.

Next, the information stored in the HDD 12 of the PC 16 will be described with reference to FIG. 2 (A). The HDD 12 has a user's private key SK, a public key PK, a common key K1, a common key K2 for a trap door, a user's address in a blockchain, a smart contract, a token, and artificial intelligence (also referred to as "AI (Artificial Intelligence)". ) And blockchain data etc. are stored. The user is a broad concept that includes not only natural persons but also corporations.

The private key SK and the public key PK are a key pair used in PKI (Public key Infratructure), and the data encrypted by the public key PK is decrypted by the private key SK. The private key SK is also used for electronic signatures. The common key K1 is a key used for common key encryption such as DES (Data Encryption Standard) and AES (Advanced Encryption Standard). The data encrypted by the common key K1 is decrypted by using the same common key K1. In this embodiment, a different common key is used for each personal information to be encrypted. In the first embodiment, an index (index) for keyword search is provided for the encrypted personal information _{EK1 (personal information) in the ciphertext.} The index is encrypted with the common key K2. In order to perform a keyword search, an encrypted search query (this is called a "trapdoor") in which the keyword (search query) used for the search is encrypted with the common key K2 is used for the search. This common key K2 is stored in the HDD 12 as a trapdoor common key K2.

The user's address on the blockchain is generated through the following process.
1 Generate a public key from the private key with ECDSA.
2 Pass the public key through the hash function SHA-256 to obtain the hash value.
3 The hash value is further passed through the hash function RIPEMD-160 to obtain the hash value.
4 Add 00 as a prefix to the beginning of the hash value.
5 Pass through the hash function SHA-256.
6 Pass it through the hash function SHA-256 again.
Add a 7-byte checksum to the end.
8 Encode in Base58 format.

A smart contract is a computer protocol intended for smooth contract verification, condition confirmation, execution, execution, and negotiation. A token is a unique currency issued on the blockchain by a company or individual.

Next, blockchain data will be described. The data in each block of the blockchain contains the hash value, nonce and data of multiple transactions (also referred to as transactions) of the previous block. Also, although not shown, the time stamp is also embedded in the blockchain. Such a blockchain is generated by each node 19 performing blockchain processing (see S3, S19, S30, S51, S117, S122, S153, etc., which will be described later), and is added as a new blockchain. Blockchain processing mainly consists of three phases: transaction, propagation and recording.

The transaction phase is an act generally called a transaction, and means a legal act such as a buying and selling act, a transfer act, and a lending (lending) act. More specifically, this transaction phase can be divided into three phases: generation → signature → propagation.

The generation phase is to generate a transaction. For example, Mr. A decides to Mr. B that "a dormant PC resource (computational resource) is lent for 39005 seconds to obtain 25.78 tokens." Digitally sign that the transaction was generated. In this electronic prominence, a hash value is generated by passing transaction data through a predetermined hash function, and the hash value is encrypted using the private key SK of the parties (Mr. A and Mr. B) of the transaction (transaction). Is. You may also have a certificate authority issue a digital public key certificate. Figure 2 shows an example of lending PC resources (computational resources), but the lending target is not limited to that, for example, electric power self-generated at home or company, user's specialized knowledge, experience and skills. And the value of personal connections (including human networks on the net) and credit.

The propagation phase is to have other nodes confirm that the transaction has been created and signed correctly. If it is determined that the transaction was not created / signed correctly, the transaction will be discarded.

The recording phase is for the miner to perform mining and record the transaction when it is confirmed that the transaction has been generated and signed correctly. Transactions that have been successfully generated and signed are moved to a location called the mining pool. After that, the miner selects the transaction to be recorded from the mining pool and performs mining.

Mining is the work of calculating nonce. The nonce is a value for adjusting so that a very small hash value such that many 0s are lined up at the beginning is generated when the block data is passed through the hash function. A new block is created if the nonce can be calculated so that the hash value is less than or equal to the target value.

_{The transaction data includes the hash value of EK1} (personal information) in which the user's personal information is encrypted with the key K1, its electronic ID, and the index of personal information, as shown in transaction I on the right side of FIG. and consideration _(provided by the index +2.4 token) E _K2 of the encrypted by key K2 (provided 2.4 token in FIG. 2) also comprise for providing personal information. Specific examples of personal information include vital information such as the user's heart rate, blood pressure, body temperature, and brain waves, behavior history information such as purchase history and website browsing history, user's location information such as GPS, race, creed, and society. Information such as personal status, medical history, electronic medical record data, ID (identification), and information posted on SNS.

The posted information to the SNS or the like is the past posted information already posted on the SNS 25 and stored in the server 20 transferred from the server 20 to the personal information DB 29 and the blockchain. Specifically, the user himself encrypts all the past posted information and stores it in the personal information DB 29 of the authorized business operator, and records the hash value in the blockchain. After that, the user does not post to the SNS 25, but encrypts and stores the posted content in the personal information DB 29 of the authorized business operator, and records the hash value in the blockchain. As a result, the user can retrieve personal information from a business operator such as SNS and put it under self-management.

The consideration for the provision of personal information (provided in 2.4 tokens in FIG. 2) may be recorded in plain text on the blockchain without being encrypted. In that case, even if another user does not obtain the encryption key K2, the blockchain can be searched and the consideration can be known. In addition, smart contracts include transaction conditions such as consideration for providing personal information and consideration for lending PC resources (computational resources) (in Fig. 2, lending PC resources (computational resources) for 39005 seconds to obtain 25.78 tokens). In the form of a code, a smart contract may be used to automate transactions (legal acts).

The encrypted personal information itself, which is the target of the hash value recorded as transaction I, is stored in the personal information DB 29 of the authorized business operator 17. Specifically, as shown in FIG. 2 (B), the encrypted personal information _EK1 (personal information) is associated with the electronic ID issued by the _{authorized business operator 17 for the encrypted personal information EK1} (personal information). Is stored in the personal information DB 29.

The index is an index _{for keyword-searching encrypted personal information EK1} (personal information). In the present embodiment, since the common key encryption method in which the index is encrypted by the common key K2 is adopted, in order to perform a keyword search, the keyword (search query) used for the search is encrypted by the common key K2. Search using a search query (this is called a "trap door"). The common key K2 is a different key for each user, but the same key is used if it is an encrypted index of the same user. Therefore, if, for example, user A performs a transaction to distribute the common key K2 to user B by a smart contract described later, user B uses _EK2 (search query) to perform all the encrypted indexes of user A. It will be searchable on the blockchain.

It should be noted that a searchable cipher such as a homomorphic encryption or a fully homomorphic encryption that can be searched while the ciphertext is encrypted may be used. At that time, personal information may be encrypted by using homomorphic encryption or fully homomorphic encryption, and the encrypted personal information may be recorded directly on the blockchain. Further, the encrypted personal information _EK1 (personal information) may be recorded directly on the blockchain.

Next, a flowchart of the main routine program of the user terminal of the public chain 19 will be described with reference to FIG. 3 (A). The personal information recording process is performed by step S (hereinafter simply referred to as “S”) 1, the smart contract process is performed by S2, and the blockchain process is performed by S3.

The personal information recording process is a process in which the personal information owner encrypts the personal information, registers it with the authorized business operator 17, and records the hash value of the encrypted personal information in the blockchain. The smart contract process is a process that automatically performs legal acts such as contract conclusion and execution in accordance with predetermined rules. The specific contents of the blockchain processing are as described above based on FIG. 2 (A).

The personal information recording process will be described with reference to FIG. 3 (B). According to S5, it is determined whether or not the personal information registration operation has been performed on the user terminals constituting the node 19 of the public chain 19. If not, this personal information recording process returns and shifts to the smart contract process of S2. When it is determined that the personal information has been registered, the personal information stored in the memory (HDD12, etc.) of the user terminal is encrypted with the key K1 by S6, and then digitally signed with the private key SK. At the same time, the index is encrypted with the key K2 and transmitted to the server 18 of the certified business operator 17.

The server 18 of the authorized business operator 17 that received it in S7 issues an electronic ID and generates a hash value of _{EK1 (personal information) that is the received encrypted personal information.} Next, the issued electronic ID is returned to the user terminal (S9). The user terminal that receives it stores the electronic ID in a memory (HDD 12, etc.). In S10, the server 18 of the certified business operator 17 performs a process for recording the electronic ID, the hash value, and the encrypted index on the blockchain.

The flowchart of the subroutine program of the smart contract processing shown in S2 will be described. With reference to FIG. 4, it is determined in S13 whether or not the distribution contract for the common key K2 has been established, and if not, in S15, the loan contract for the PC resources (computational resources) of the user terminal has been established. It is determined whether or not it is established, and if it is not established, it is determined whether or not the contract for providing personal information is established in S16, and if it is not established, an ordering contract for ordering a custom-made product or the like is established in S17. Whether or not it is determined, and if it is not established, it is determined in S22 whether or not the sales contract for the product or the like is established, and if it is not established, a return is made. These decisions are made by smart contracts. For example, when the above-mentioned consideration is coded as a smart contract, the smart contract determines whether or not the conditions such as consideration of both parties are met, and if it is determined that the conditions are met, the contract is concluded and the contract is concluded. Execution is performed automatically.

When it is determined that the distribution contract for the common key K2 has been established, the control proceeds to S14, and after the common key K2 is transmitted to the distribution destination, the control proceeds to S19. In S19, a process for recording the established contract as a transaction in the blockchain is performed. When it is determined that the PC resource (computational resource) loan contract has been established, the control proceeds to S18, and the PC resource (computational resource) loan process is performed. If it is determined that the personal information provision contract has been established, the control proceeds to S20, and the electronic ID of the personal information to be provided and the signature agreeing to provide the personal information are returned to the provider and the personal information to be provided is returned. The common key K1 used for the encryption of the above is encrypted with the public key of the provider and returned to the provider.

If it is determined that the ordering contract has been established, the control proceeds to S21, and after the ordering process is performed, the process proceeds to S19. When it is determined that the sales contract has been established, the control proceeds to S23, the process of obtaining the purchase target is executed, and the process proceeds to S19.

Next, a flowchart of the main routine program of the user terminal constituting the node 19 of the private chain 2 will be described with reference to FIG. 5 (A). Personal information search processing is performed by S28, smart contract processing is performed by S29, blockchain processing is performed by S30, machine learning processing is performed by S31, AI smart contract generation processing is performed by S32, and AI smart contract generation processing is performed by S33. Smart contract trust contract processing is performed. AI smart contract is a concept that includes both "integrated type" and "cooperation type". The "integrated type" is an AI version of the smart contract itself by integrating the AI and the smart contract and performing machine learning based on the data of the contract (legal act). The "cooperation type" is a collaboration between an AI that has performed machine learning based on contract (legal act) data and a smart contract. In the case of the cooperation type, the learned AI (hereinafter referred to as "linkage AI") adds, changes, updates, etc. to the smart contract depending on the situation.

The personal information search process is a process of searching an encrypted index recorded in a blockchain by a trapdoor (encrypted search query). The smart contract process is a process that automatically performs legal acts such as contract conclusion and execution in accordance with predetermined rules. The specific contents of the blockchain processing are as described above based on FIG. 2 (A). The machine learning process is a process of generating a trained model of artificial intelligence by machine learning the personal information of a large number of users as learning data. More specifically, after generating a general learned model of artificial intelligence by machine learning a huge amount of personal information in a form that cannot identify the personal information owner as learning data, data that can identify the personal information owner (for example, block) This is a process of generating a personalized trained model that is personalized for each personal information owner (for each blockchain address) using personal information classified for each (address in the chain).

The AI smart contract generation process is a process of machine learning personal information about a contract (legal act) as learning data to generate a trained model of a smart contract by artificial intelligence. More specifically, after generating a general learned model of smart contract by artificial intelligence by machine learning personal information about a huge contract (legal act) in a form that can not identify the personal information owner as learning data, an individual Personalized AI smart contract personalized for each personal information owner (for example, for each address in the blockchain) using personal information about contracts (legal acts) classified for each data (for example, address in the blockchain) that can identify the information owner. This is a process to generate a trained model.

The smart contract trust contract processing is a processing that provides a service that automatically performs legal acts such as contract conclusion and execution on behalf of the contractor. More specifically, it generates a personalized AI smart contract trained model for the trustee and uses the personalized AI smart contract trained model to perform legal acts on behalf of the trustee. The reward for AI is determined based on the result of the execution, and the personalized AI smart contract trained model is further strengthened and learned by the reward.

Next, a flowchart of the subroutine program for the personal information search process shown in S28 will be described with reference to FIG. 5 (B). It is determined by S37 whether or not the common key K2 is stored, and if not, it returns. When K2 is stored by S45 described later, it is determined by S37 that the common key K2 is stored, and the control proceeds to S38. In S38, a process of searching the encrypted index on the blockchain with a search query (trapdoor) encrypted with K2 is performed. As a result of the search, it is determined by S39 whether or not there is personal information to be obtained. If it is determined that there is no personal information that is desired to be obtained, a return is made, but if it is determined that there is personal information that is desired to be obtained, the electronic ID of the personal information that is desired to be obtained is stored in S40.

Next, a flowchart of the smart contract processing subroutine program shown in S29 will be described with reference to FIGS. 6 and 7 (A). It is determined by S42 whether or not there is personal information desired to be searched. In this determination, for example, when the conditions are met by sequentially negotiating with the smart contract of the personal information owner who has not been searched yet, the personal information of the personal information owner is determined as the personal information desired to be searched. If it is determined by S42 that there is no personal information desired to be searched, it is determined by S46 whether or not the personal information desired to be obtained is stored, and if it is determined that there is no personal information desired to be obtained, control proceeds to S55 in FIG. 7 (A). , It is determined whether or not the PC resource (computational resource) loan contract has been established, and if it is determined that it has not been established, it is determined by S56 whether or not the ordering contract has been established, and if it has not been established, it is determined. If it is determined, it is determined by S57 whether or not the sales contract has been established, and if it is determined that the sales contract has not been established, a return is made.

When it is determined by S42 that there is personal information desired to be searched, the control proceeds to S43 and requests the common key K2 from the personal information main. Specifically, it sends its own address and an Attribute Certificate to the address on the blockchain of the personal information owner of the personal information desired to be searched, and requests the common key K2. It is determined by S44 whether or not there is a reply of K2, and the process waits until there is a reply. The personal information owner or the smart contract of the personal information owner confirms the sent attribute certificate, determines whether or not K2 can be returned, and returns K2 when it is determined that the reply is acceptable. When K2 is returned from the personal information owner, the control proceeds to S45, and after storing the returned K2, the control shifts to S54. In S54, a process of storing the established contract as a transaction in the blockchain is performed. In this case, the contract that the common key K2 used for encrypting the index is distributed from the address of the personal information owner who replied to the address of the user who received the reply is stored in the blockchain.

When it is determined by S46 that the personal information desired to be obtained is stored, the control proceeds to S47, and a process of requesting the personal information desired to be obtained from the personal information is performed. Specifically, the personal information desired to be obtained is requested by sending its own address and attribute certificate to the address on the blockchain of the owner of the personal information desired to be obtained. In the public chain user terminal that receives this, when the smart contract confirms the attribute certificate, determines whether or not the conditions for providing personal information are met, and determines that the personal information may be provided (in S16). YES), the electronic ID of the personal information to be provided and the signature agreeing to provide the personal information are returned, and the common key K1 used for encrypting the personal information to be provided is encrypted with the public key of the provider and returned (YES). See S20).

When the reply is received, it is determined by S48 that there is a reply from the personal information owner, the control proceeds to S49, and the returned encryption common key _EPC (K1) is used as the own private key SK. The operation of decoding with, that is, _DSK (E _PK (K1)) is performed to calculate K1. Next, S50 performs a process of transmitting the electronic ID and signature returned from the personal information owner to the server 18 of the authorized business operator 17. Upon receiving this, the server 18 of the authorized business operator 17 confirms the transmitted signature, searches the personal information DB 29 (see FIG. 2B) based on the received electronic ID, and receives the received electronic ID. The encrypted personal information _EK1 (personal information) stored in association with is read and returned.

If the reply is received, it is determined as YES by S51, the control proceeds to S52, and the returned encrypted personal information _EK1 (personal information) is decrypted by K1 calculated in S49, that is, _{Perform DK1} ( _EK1 (personal information)) to obtain plaintext personal information. The personal information is stored in S53. After that, the process proceeds to S54, and a process for recording the personal information provision contract as a transaction in the blockchain is performed.

Next, if a PC resource (computational resource) loan contract is concluded with the user terminal of the public chain (YES in S15), it is determined as YES by S55 and the control proceeds to S58, and the PC resource (computational resource) Borrowing process is performed. After that, the control shifts to S54, and a process for recording the lease contract as a transaction in the blockchain is performed. If an ordering contract is established with the user terminal of the public chain (YES in S17), it is determined by S56 that the ordering contract has been established, control proceeds to S59, and a process of storing the order for the order is performed. After that, the control shifts to S54, and a process for recording the order contract as a transaction in the blockchain is performed.

If a sales contract is concluded with a user terminal of a public chain (YES in S15), it is determined by S57 that a PC resource (computational resource) loan contract has been concluded, control proceeds to S60, and a sales target is provided. Processing is performed. After that, the control shifts to S54, and a process for recording the sales contract as a transaction in the blockchain is performed. In the smart contract processing described above, actions (for example, S42, S47, S58, S42, S47, S58, S42, S47, S58, S57) are based only on the judgment (for example, S42, S46, S55, S56, S57) of the private chain user terminal 16 based on the smart contract. Instead of executing S56, S60), the act may be executed with the consent of the owner of the smart contract. This owner's consent may also be given when the contract is executed by the smart contract described below.

Next, a flowchart of the machine learning processing subroutine program shown in S31 will be described with reference to FIG. 7B. According to S63, a process of converting a huge amount of personal information stored in a form in which the personal information owner cannot be identified into learning data is performed. Various learning algorithms adopted in this machine learning include, for example, regression and identification as supervised learning, model estimation and data mining as unsupervised learning, reinforcement learning and deep learning as intermediate methods, and the like. Is prepared.

Next, according to S64, machine learning using learning data is performed using the borrowed PC resources (computational resources). For example, in the case of regression as supervised learning, a large amount of data set consisting of input information (vector x) and correct answer information y is used as training data (learning data). In the case of supervised learning, since the function ci (x) (ci: x → y) that maps the input x to the correct answer y is learned, the trained model includes the function ci. The machine learning performed by the machine learning means 34 is not limited to supervised learning, but is an intermediate method between unsupervised learning such as model estimation and pattern mining (data mining), supervised learning and unsupervised learning. It may be any kind of learning with semi-teacher, intensive learning, deep learning, etc.

Machine learning by S64 generates a general trained model and stores it (S65). This general learned model is a model in which a huge amount of personal information, which is personal information of a large number of users and is stored in a form in which the personal information owner cannot be identified, is used as learning data, and can be widely applied to a large number of users. It is an average trained model.

Next, S66 determines whether or not there is an order storage, and if there is no order memory, it returns, but if there is, S67 determines whether or not the order is for artificial intelligence. If it is not an order for artificial intelligence, it will be returned, but if it is an order for artificial intelligence, S68 will perform a process of requesting personal information to the address of the ordering party. If there is a reply of personal information from the orderer, it is determined as YES by S69 and the process proceeds to S70. In S70, a process of personalizing a general trained model based on the returned personal information to generate a personalized trained model is performed. A process of personalizing this general trained model to generate a personalized trained model is described in Japanese Patent No. 6432859. Since personal information required for personalization is collected using the blockchain, there is an advantage that privacy problems can be avoided as much as possible by collecting the personal information while ensuring the anonymity of the blockchain. Next, S71 transmits the personalized trained model to the address of the ordering party.

Next, a flowchart of the subroutine program for the AI smart contract generation process shown in S32 will be described with reference to FIG. According to S80, a process is performed in which personal information related to a contract (legal act) is extracted from a huge amount of personal information stored in a form in which the personal information owner cannot be identified and used as learning data. Next, according to S81, machine learning using learning data is performed using the borrowed PC resources (computational resources). The processing of S80 and S81 is the same as that described in S63 and S64 described above, and the repetition of the description is omitted here.

Next, S82 performs a process of generating and storing a trained model of a general AI smart contract. Next, the simulation learning process is executed by S83. This simulation learning process virtually executes legal acts such as contract verification, condition confirmation, execution, execution, and negotiation, which are originally performed by a large number of human beings, on behalf of a large number of AI smart contract groups in a computer. (Simulation) is performed, and each AI smart contract is rewarded according to the result to perform reinforcement learning. Since reinforcement learning is performed by simulation in a computer instead of reinforcement learning in the real world, there is an advantage that a huge amount of reinforcement learning can be performed in a short time. Reinforcement learning is a mechanism in which an agent placed in a certain environment obtains a measure that maximizes the cumulative reward from the initial state to the goal based on the reward given when an action is selected. That is. In reinforcement learning, learning proceeds by interacting with a software agent (hereinafter referred to as "agent"), which is a type of AI, and the environment. Here, an agent is a kind of AI, and is software that behaves autonomously and acts permanently with a certain degree of judgment ability while communicating with users and software. When the agent performs a certain action a on the environment, the state s of the environment changes, and when a certain target state is reached, the reward r is given to the agent. The agent learns a function that outputs an action a with a state s as an input for the purpose of maximizing the reward r.

Reinforcement learning progresses in time by repeating the following simple steps.
1 The agent receives the observation o (or directly, the state s of the environment) received from the environment, and returns the action a to the environment based on the policy π.
2 The environment changes to the next state s'based on the action a received from the agent and the current state s, and based on the transition, the next observation o'and the good or bad of the immediately preceding action called reward r. Returns the indicated number (scalar amount) to the agent.
3 hours progress: t ← t + 1
Here, ← represents an assignment operation.
As the reinforcement learning, for example, an alpha zero type reinforcement learning algorithm may be used. This alpha zero type reinforcement learning algorithm is different from algorithms such as DQN (Deep Q-Network), and Monte Carlo tree search (MCTS) is used for the search, and all values and policies are used. The configuration is such that a neural network is made to make a prediction, and the prediction is corrected only by the experience obtained by self-play by tree search. Compared to the conventional AlphaGo, the value network that predicts the value and the policy network that predicts the policy are integrated into one neural network, and the prediction accuracy is improved by multi-task learning. In addition, due to the improved performance of the neural network, the processor layout (extending the search tree until a reward is received) processing in the tree search becomes unnecessary, and the search can be performed at a higher speed. In addition, evolutionary computation, genetic algorithms, and Generative adversarial networks may be used.

Next, S84 determines whether or not there is an order storage for the AI smart contract, and if not, returns. When the order is stored in S59 described above, YES is determined in S84, the control proceeds to S85, and it is determined whether or not the stored order is an order for the simulation-learned AI smart contract. If it is not an order for a simulation-learned AI smart contract, it is an order for a personalized AI smart contract-learned model. In that case, control proceeds to S86 and personal information about the contract (legal act) is addressed to the orderer's address. Is performed. When the orderer returns the personal information, S87 determines YES and the control proceeds to S88.

In S88, a process of personalizing a trained model of a general AI smart contract based on personal information regarding the returned contract (legal act) to generate a personalized AI smart contract trained model is performed. A process of personalizing a trained model of this general AI smart contract to generate a personalized AI smart contract trained model is described in Japanese Patent No. 6432859. The personalized trained model is transmitted by S89 to the orderer's address.

On the other hand, in the case of receiving an order for the simulation-learned AI smart contract, S85 determines YES, the control proceeds to S90, and the simulation-learned AI smart contract is transmitted to the orderer's address.

Next, a flowchart of the subroutine program for the simulation learning process shown in S83 will be described with reference to FIG. 9A. According to S334, it is determined whether or not there is a simulation input, and if not, a return is made. This simulation is input by a user terminal of private chain 2, for example, policies and laws that the government is trying to adopt (for example, reduced tax rate due to consumption tax increase, revised immigration law, EU (European Union) in the United Kingdom). ), Partial or full adoption of basic income, amendment of Article 9 of the Japanese Constitution, etc.), transaction simulation in the investment market such as stock trading and futures trading, company management simulation , Or consumption behavior simulation, etc. Furthermore, promotion simulations of new products (including financial products and life insurance) and new services by various media may be used. If it is determined by S334 that the simulation has been input, the control proceeds to S335, and the AI smart contract group generation process is executed.

The flowchart of the subroutine program of the AI smart contract group generation processing will be described with reference to FIG. 9B. According to S344, a process of setting a persona group matching the input simulation is performed by using the borrowed PC resource (computational resource). A persona is generally defined as a virtual person who is a typical target person for a company, product, or service. In the present embodiment, the persona defines a typical person image targeted by the simulation content as a virtual person. For example, in the case of the consumption behavior simulation under the reduced tax rate accompanying the consumption tax increase mentioned above, the persona is equivalent to a general consumer, and each group is grouped by gender, age group, region, annual income, etc. Set as a persona.

The number of personas to be set is proportional to the number of users belonging to the group. For example, the population distribution of general consumers by age group is 5% in their teens, 5% in their 20s, 10% in their 30s, 10% in their 40s, 20% in their 50s, 20% in their 60s, 20% in their 70s, and 5% in their 80s. In the case of 5% in their 90s, the number of personas representing teens is 1, the number of personas representing 20s is 1, the number of personas representing 30s is 2, and the number of personas representing 40s is 2. The number of personas representing the 50s is 4, the number of personas representing the 60s is 4, the number of personas representing the 70s is 4, the number of personas representing the 80s is 1, and the number of personas representing the 90s is 1. , Set.

Next, according to S345, a process of selecting a user group belonging to each persona is performed using the borrowed PC resource (computational resource). Next, according to S346, a process is performed in which a group of users belonging to each persona is grouped and transaction data of the user group is collected from the blockchain for each group. For example, in the case of the consumption behavior simulation under the reduced tax rate accompanying the consumption tax increase described above, the user groups are grouped by gender, age group, region, annual income, etc., and the transaction data of the user group is collected for each group. Collect from the blockchain. For the selection of the user group and the collection of transaction data of the user group in S345 and S346, for example, it is useful to use the database of the questionnaire response monitor members owned by the Internet questionnaire survey company. Internet questionnaire survey companies store the contact information (email address, etc.) of monitor members for questionnaire response in a database in association with attributes such as gender, age group, place of residence, unmarried, married, occupation, and annual household income. Use the monitor member data for each attribute. Similarly, in S145 and S146, S586 and S587, S622 and S623, which will be described later, it is useful to use the database of questionnaire response monitor members owned by the Internet questionnaire survey company. Next, according to S347, a process is performed in which the borrowed PC resource (computational resource) is used to perform machine learning using the transaction data as learning data to generate a trained AI smart contract for each persona. The number of AI smart contracts generated is the same as the number of corresponding persona settings. As a result, the environment for executing the simulation is prepared, and the simulation is performed in that environment.

Returning to FIG. 9A, each AI smart contract generated as described above executes a contract (legal act) in accordance with act a (S336). This "action a" is an action a as a result of reinforcement learning by S338. Next, according to S337, a process of recording the contract concluded between the AI smart contracts on the blockchain is performed.

Next, in S338, using the borrowed PC resources (computational resources), the reward r is calculated based on the contract details that have been concluded, and the process of finding the action a according ^{to the optimal policy π * by TD learning is performed.} Will be. For example, in the case of the consumption behavior simulation under the reduced tax rate accompanying the consumption tax increase described above, the smaller the value (expense amount before tax increase-expenditure amount after tax increase), the higher the reward r is given. Then, it is determined by S339 whether or not the simulation is completed, and if it is not completed yet, the control returns to S336, and the reinforcement learning is advanced by repeatedly patrolling S337 → S338 → S339 → S336. When the simulation is completed, S339 determines YES, the control proceeds to S340, and the AI smart contract with the highest reward r is stored and then returned. It should be noted that the AI smart contract that has received the highest reward r is not limited to that, and for example, the top 5% of AI smart contracts may be stored. Further, it is not always necessary to record to the blockchain by S337. In that case, in the above-mentioned cooperation type, "AI smart contract" in S335, S336, S340 and S347 is changed to "AI for cooperation". do. In other words, in the case of simulation in a computer, it does not involve the execution of contracts (legal acts) in the real world, so if you do not record on the blockchain, you do not have to bother to use smart contracts, and for each cooperation This is because it is sufficient for AIs to perform action a and perform reinforcement learning. At the stage of actual citation after the reinforcement learning is completed, the learned cooperation AI may execute the contract in cooperation with the smart contract and record it on the blockchain.

Next, a flowchart of the subroutine program for the smart contract trust contract processing shown in S33 will be described with reference to FIG. According to S94, it is determined whether or not there is an order storage for the smart contract trust, and if it is determined that there is no order storage, a return is made. The contents of the order storage stored in S59 described above are confirmed, and if the order is for the smart contract trust, it is determined as YES by S94 and the control proceeds to S95. In S95, personal information regarding the contract (legal act) is requested to the address of the ordering party, and when the personal information is returned from the ordering party, S96 determines YES and the control proceeds to S97.

In S97, a process of personalizing a trained model of a general AI smart contract based on personal information regarding the returned contract (legal act) to generate a personalized AI smart contract trained model is performed. A process of personalizing a trained model of this general AI smart contract to generate a personalized AI smart contract trained model is described in Japanese Patent No. 6432859.

Next, according to S98, a process of associating and storing the address of the orderer and the personalized AI smart contract trained model is performed. Using this memorized personalized AI smart contract trained model, the orderer is subjected to trust contract processing (S99). Next, S100 executes reinforcement learning processing of the personalized AI smart contract trained model.

Based on FIG. 11A, a flowchart of a subroutine program for reinforcement learning processing of the personalized AI smart contract trained model shown in S100 will be described. In reinforcement learning, as a method of estimating the Q value when the value when performing the action at in the state st is Q (st, at), the knowledge that models the environment, that is, the state transition probability and the probability distribution of the reward If is given, a model-based method may be used, but if the environmental model is unknown, TD (Temporal Difference) learning is used. First, since it is necessary to search the environment, the ε-greedy method is used. In the early stages of exploration, we try various actions and introduce the concept of temperature so that when we calm down, we choose many of the most suitable actions. Let T be the temperature, and choose the action according to the probability expressed by the following formula.

P (a ｜ s) ＝ {exp (Q (s, a) / T)} / {Σexp (Q (s, b) / T} (Note that b ∈ A is described under Σ, and the above equation The description is omitted)
Here, a is an act, Q (s, a) is the value when performing an act a in the state s,

T is called the temperature in annealing, and if it is high, the action is selected with a probability close to equal probability, and if it is low, it is biased to the optimum one. As the learning progresses, the learning result becomes stable by reducing the value of T. Such a method of estimating the Q value may be applied to all of the reinforcement learning described above and the reinforcement learning described later. Further, as a computer for performing machine learning such as reinforcement learning, a general von Neumann computer is used, but a neural network processor (NNP) may be used. Many "artificial neurons" modeled on real neurons are mounted on the NNP chip, and each neuron cooperates with each other in a network. Further, a quantum computer adopting the "quantum annealing method" may be used. In particular, by using a quantum computer that employs the "quantum annealing method", the time required for optimization calculation in machine learning can be significantly reduced.

In S105, the trustee receives an evaluation of the result of the trust contract processing by the personalized AI smart contract trained model. Next, in S106, a process of calculating the reward r based on the received evaluation is performed. Next, according to S107, a process of obtaining an action a according ^{to the optimum policy π *} is performed by TD learning using the borrowed PC resource (computational resource). Next, according to S108, the contract according to the act a is executed on behalf of the trustee. By receiving the evaluation of the result from the trustee (S105), the processes of S106 to S108 are executed.

Next, a flowchart of the main routine program of the user terminal constituting the nodes of the consortium chain 3 will be described with reference to FIG. 11 (B). IoT sensor data aggregation processing is executed by S114, smart contract processing is executed by S115, simulation processing is executed by S116, and blockchain processing is executed by S117. The specific contents of the blockchain processing are as described above based on FIG. 2 (A).

Next, a flowchart of the subroutine program for the IoT sensor data aggregation process shown in S114 will be described with reference to FIG. 12 (A). According to S120, IoT sensor data is classified and grouped by type, period, region, and the like. This processing includes not only IoT sensor data but also data from the wireless sensor network. Next, S121 performs a process of determining the value of each grouped data. According to this determined value, the consideration (amount of tokens) for providing the data is encoded as a smart contract for each corresponding data. Next, S122 performs a process of recording each grouped data on the blockchain.

Next, a flowchart of the smart contract processing subroutine program shown in S115 will be described with reference to FIG. 13 (B). According to S150, it is determined whether or not a sales contract with a person who wants to obtain the data has been established. This determination is automatically determined to be YES by S150 when both conditions are met with respect to the consideration (amount of tokens) for the provision of data encoded as a smart contract. If NO is determined by S150, the control proceeds to S151, it is determined whether or not the PC resource loan contract has been established, and if not, it returns.

If YES is determined by S150, the control proceeds to S152, data is transmitted to the address of the person who wants to obtain the data, and a process of acquiring a token as a consideration is performed. The established contract is recorded in the blockchain as a transaction by S153. On the other hand, when a PC resource (computational resource) loan contract is concluded, control proceeds to S154, the PC resource (computational resource) is borrowed, and the contract is recorded in the blockchain as a transaction by S153.

Next, a flowchart of the simulation processing subroutine program shown in S116 will be described with reference to FIG. 12 (B). This simulation process virtually executes legal acts such as contract verification, condition confirmation, execution, execution, and negotiation, which are originally performed by a large number of people, on behalf of a large number of AI smart contracts (in a computer). (Simulation) is performed to verify what kind of simulation result will be obtained under certain conditions. Specific examples of the above "conditions" include policies and laws that the government is trying to adopt (for example, reduced tax rates associated with consumption tax increases, revised immigration control laws, withdrawal from the EU (European Union) in the United Kingdom, and basic income. Partial or full adoption of, revision of Article 9 of the Japanese Constitution, etc.), marketing-related conditions (for example, setting prices and consideration for new products (including financial products and life insurance) and new services, promotion by various media (Effects, etc.), investment market-related conditions (for example, weather conditions in futures trading, monetary tightening policies in the stock market, etc.) can be considered.

S134 determines whether or not there is a request for simulation, and if not, returns. If it is determined by S134 that the simulation request has been made, the control proceeds to S135, and the AI smart contract group generation process is executed.

The flowchart of the subroutine program of the AI smart contract group generation processing will be described with reference to FIG. 13 (A). In S144, a process of setting a persona group matching the requested simulation is performed by using the borrowed PC resource (computational resource). A persona is generally defined as a virtual person who is a typical target person for a company, product, or service. In the present embodiment, the persona defines a typical person image targeted by the simulation content as a virtual person. For example, in the case of the simulation of the reduced tax rate associated with the consumption tax increase described above, a persona corresponding to a general consumer is set as a persona for each group grouped by gender, age group, region, annual income, etc. ..

The number of personas to be set is proportional to the number of users belonging to the group. For example, the population distribution of the general public by age group is 5% in their teens, 5% in their 20s, 10% in their 30s, 10% in their 40s, 20% in their 50s, 20% in their 60s, 20% in their 70s, and 5% in their 80s. In the case of 5% in the 90s, the number of personas representing the teens is 1, the number of personas representing the 20s is 1, the number of personas representing the 30s is 2, and the number of personas representing the 40s is 2,50. The number of personas representing the teens is 4, the number of personas representing the 60s is 4, the number of personas representing the 70s is 4, the number of personas representing the 80s is 1, and the number of personas representing the 90s is 1. Set.

Next, according to S145, a process of selecting a user group belonging to each persona is performed using the borrowed PC resource (computational resource). Next, according to S146, a process is performed in which a group of users belonging to each persona is grouped and transaction data of the user group is collected from the blockchain for each group. For example, in the case of the simulation of the reduced tax rate associated with the consumption tax increase described above, the user groups are grouped by gender, age group, region, annual income, etc., and the transaction data of the user group is collected from the blockchain for each group. do. Next, according to S147, using the borrowed PC resource (computational resource), machine learning is performed using the transaction data as learning data to generate a trained AI smart contract for each persona. The number of AI smart contracts generated is the same as the number of corresponding persona settings. As a result, the environment for executing the simulation is prepared, and the simulation is performed in that environment.

Returning to FIG. 12B, each AI smart contract generated as described above executes the contract (legal act) according to the act a (S136). This "act a" is an act a as a result of reinforcement learning by S138. Next, according to S137, a process of recording the contract concluded between the AI smart contracts on the blockchain is performed. Furthermore, the transition status of the situation as the simulation progresses is recorded on the blockchain. For example, in the case of the simulation of the reduced tax rate associated with the consumption tax increase described above, how the domestic demand and the economy have changed as the simulation progresses is recorded on the blockchain.

Next, according to S138, the borrowed PC resource (computational resource) is used to calculate the reward r based on the contract details that have been concluded, and the process of obtaining the action a according ^{to the optimum policy π * by TD learning is performed.} Will be. Then, it is determined by S139 whether or not the simulation is completed, and if it is not completed yet, the control returns to S136 and repeats S137 → S138 → S139 → S136 to proceed with reinforcement learning. When the simulation is completed, S139 determines YES, the control proceeds to S140, and the process returns after the process of deriving the simulation result is performed.

As a specific example of the process of deriving the simulation result, for example, in the case of simulating the economic fluctuation accompanying the adoption of the reduced tax rate due to the consumption tax increase, how each item of the business cycle index fluctuated as a result of the simulation. Derived. In the case of a simulation that accompanies a monetary tightening policy in the stock market, we derive how the stock market fluctuated as a result of the simulation. In addition, the optimum mode of the reduced tax rate while changing the specific mode of the reduced tax rate due to the consumption tax increase (for example, what is the item subject to the reduced tax rate and the reduced tax rate for each item subject to the reduced tax rate) in multiple modes. The simulation optimization method of determining the above may be adopted. In this case, as the optimum mode of the reduced tax rate, (tax revenue increase rate (%) + diffusion index (DI) / 50 as an economic trend index) is set as the expected value E, and the simulation result that maximizes the expected value E is used. Try to get. Let θ be the control parameter (mode of reduced tax rate) in the simulation, let Y (θ) be the result of the simulation, and find θ in maxE [Y (θ)]. Specific methods for optimizing such simulations include, for example, particle swarm optimization (PSO) as a metaheuristic algorithm, information on cases where analytical representation of the objective function is difficult, and information on the differentiation of the objective function. Use DFO (derivative free optimization) etc. to find the optimum solution in the situation where can not be used. It is not always necessary to record to the blockchain by S137. In that case, in the above-mentioned cooperation type, "AI smart contract" in S135, S136 and S147 is changed to "AI for cooperation". In other words, in the case of simulation in a computer, it does not involve the execution of contracts (legal acts) in the real world, so if you do not record on the blockchain, you do not have to bother to use smart contracts, and for each cooperation This is because it is sufficient for AIs to perform action a and perform reinforcement learning.
[Modification example]

(1) The authorized business operator 17 stores the user's encrypted personal information _EK1 (personal information), but instead _{records the encrypted personal information EK1} (personal information) directly on the blockchain. You may try to do so.

(2) Transaction data other than personal information such as transaction C and transaction F in FIG. 2 may be encrypted with a key K1 or the like and recorded on the blockchain in the same manner as personal information.

(3) In the above description, the operation processing of the nodes 19 of the blockchain networks 2, 3 and 4 is shown, but the operation processing shown for the node 19 of the private chain 2 is applied to the operation processing of the nodes 19 of the other blockchain networks 3 and 4. The operation processing shown for the node 19 of the consortium chain 3 may be performed at the node 19 of the other blockchain networks 2 and 4, and the operation processing shown for the node 19 of the public chain 4 may be performed by another block. It may be performed at the nodes 19 of the chain networks 2 and 3. This modification may be similarly applied to the embodiments described later.

(4) The borrowed PC resources (computational resources) may be used for mining (competition of bookkeeping rights) on the blockchain. At that time, PC resources (computational resources) may be lent at an hourly rate as in the first embodiment, but what is the profit (token, etc.) obtained by a miner who succeeds in mining (competition of bookkeeping rights)? The percentage may be distributed (dividend) to the lenders of PC resources (computational resources). The dividend ratio (dividend amount) is controlled so as to be proportional to the loan amount of PC resources (computational resources) (number of PCs lent x loan time, etc.).

(5) Some project may be carried out by borrowing (using) the above-mentioned PC resources (computational resources), privately generated electric power, and the like. Specific examples of projects include research and development (for example, artificial intelligence development, machine learning, human genome analysis, new product development, new drug development, etc.), exploration and excavation of rare metals, oil, natural gas, marine resources, etc., space development, etc. However, it is possible. At that time, as in the first embodiment, the lender (provider) may obtain consideration (tokens, etc.) according to the loan amount (provided amount) to be loaned (provided), but the success of the project. A percentage of the profits obtained by the project executor (individual or corporation or group) may be distributed (dividend) to the resource lender (resource provider). The dividend ratio (dividend amount) is controlled so as to be proportional to the loan amount (provided amount) of resources.

Further, the resource lender (resource provider) may acquire the right to receive the dividend (hereinafter referred to as “dividend enjoyment right”) instead of receiving the dividend itself. This dividend enjoyment right may be controlled to be acquired by the lender (provider) in the form of a token issued by the project executor, for example. Then, the lender (provider) may control the acquired dividend enjoyment right (token) so that it can be transferred to another person at a price (token) according to the market price at that time. With this configuration, the dividend enjoyment right (token) can be managed as if it were a stock transaction in the secondary market in the stock market.

(6) In S71, the generated personalized trained model is sent to the orderer's address for delivery, but in addition to or instead, the generated personalized trained model is used for ordering. A person may be provided with a personalized service.

(7) In the above explanation, the verification, condition confirmation, execution, execution, and negotiation of the contract are automated by the smart contract, but the consent of the user himself / herself is required before concluding and executing the contract (legal act such as transaction). It may be controlled as desired. In addition, instead of seeking the consent of the user for all contracts (legal acts such as transactions), it is determined whether or not they are important contracts (legal acts such as transactions) that have been determined in advance, and important contracts (transactions, etc.) It may be controlled to seek the consent of the user himself / herself when it is determined to be a legal act). Furthermore, it is determined whether or not the contract is urgently concluded and executed (legal act such as transaction), and if it is determined that the contract is urgent (legal act such as transaction), the user himself / herself It may be controlled to conclude and execute a contract (legal act such as a transaction) without obtaining consent, and to report to the user himself / herself later. This modification may be similarly applied to the embodiments described later.

(8) The above-mentioned programs running on the user terminals 16 and the like constituting the node 19 of each blockchain and various servers may be downloaded and installed from a predetermined website or the like, but for example, recording of a CD-ROM 99 or the like. A person who purchases the CD-ROM99 or the like by recording it on a medium (non-transitory recording medium) and distributing it may install the program on the user terminal 16 and various servers (see FIG. 60). ..

(9) Although the centralized oracle 21 is adopted in the above description, the decentralized oracle that is distributed and managed in the entire network may be adopted. Information collected by multiple Oracles distributed throughout the network is collected to extract average information, and the average information is regarded as correct information and incorporated into the blockchain for use in smart contracts. This is what James Surowiecki advocated in the book "Everyone's Opinion Is Unexpectedly Correct", "By gathering information in a group, the conclusions that the group draws lead to better conclusions than anyone in the group thinks. It is based on the theory that "can be done". Then, by giving a reward such as a token to the oracle that collects information close to the average information, an incentive to operate the decentralized oracle is given.

In summary, an extraction means that collects information collected by multiple Oracles distributed on a network and extracts average information, an adoption means that adopts the average information extracted by the extraction means, and an Oracle. The plurality of oracles include a first oracle and a second oracle, and the reward giving means provides information closer to the average information than the first oracle. Give more rewards to the collected second oracle. The reward given to the first oracle may be 0 or may be negative.

(10) Information that identifies the intention of the personal information owner (hereinafter referred to as "intention identification information") with respect to the determination of one or both of the establishment of the K2 distribution contract in S13 and the establishment of the personal information provision contract in S16. ) May be provided. Specifically, when a personal information owner recommends a product or service that matches him / her at a physical store or an electronic shopping mall, he / she obtains specific information of his / her own mobile terminal (smartphone, IC card, etc.). By inputting a password that allows the user to read and conclude a contract, the smart contract of the personal information owner is notified of the intention-specific information consisting of the specific information and the password, and the smart contract is in accordance with the intention-specific information. Judges.

By doing so, the user can utilize the personal information recovered from the business operator such as SNS and placed in self-management for himself / herself by his / her own will.
[Second Embodiment]

Next, the second embodiment will be described. This second embodiment addresses the need for a right to delete personal information (so-called right to be forgotten) to delete one's own personal information by applying encryption technology to personal information recorded using a blockchain. It is a response. The blockchain is characterized in that the information once recorded cannot be tampered with or is extremely difficult to tamper with, and therefore it is impossible or extremely difficult to delete the once recorded information (hereinafter referred to as "undeleteable"). "). On the other hand, the GDPR (General Data Protection Regulation) in Europe requires that the personal information deletion right (so-called right to be forgotten) that the personal information owner can delete the recorded personal information is guaranteed. This request for the right to delete personal information of the GDPR and the impossibility of deletion in the blockchain are in direct conflict and are an antinomy dilemma. That is, this second embodiment solves the dilemma in which the request for guaranteeing the right to delete the information to be deleted and the impossibility of deletion are in conflict with each other. The outline will be described with reference to FIG.

FIG. 14 (A) shows a normal state in which the deletion right is not exercised, and FIG. 14 (B) shows a state in which the information cannot be read by exercising the deletion right. With reference to FIG. 14 (A), the information holder (also referred to as the information owner) 40 double-encrypts the information using the two half-split common keys KA and KB. Expressed as an equation, it is E _KA (E _KB (information)). Next, the encrypted information _EKA ( _EKB (information)) is recorded on the blockchain or the like. The one-sided common key KA is stored in the user terminal or the like of the information holder 40 in a secret state.

In this state, when the information requester 41 requests information from the information master 40, the information master 40 decrypts the recorded encrypted information E _KA (E _KB (information)) with the key KA. Expressed in an equation, D _KA (E _KA (E _KB (information)) = E _KB (information). Then, the _{information owner 40 sends this E KB} (information) and the half-split common key KB to the information requester 41. Send.

They information requester 41 which received the decrypted by halves common key KB that has received the received E _{KB (information).} Expressed in an equation, D _KB (E _KB (information)) = information. As a result, the information requester 41 can obtain the information in plain text.

Next, a state in which the information is made undecipherable by exercising the deletion right will be described with reference to FIG. 14 (B). The information main 40 updates one of the half-split keys KA and KB used for encrypting the information to be made undecipherable to a random number R (≠ KA). Next, when the information requester 41 who has already stored the half-split common key KB requests the information from the information master 40, the information master 40 records the encrypted information E _KA (E _KB (information)). )) Is decrypted with the key R (random number). Expressed by the formula, the _{D R (E KA (E KB} ( information)). Then, the _{D R (E KA (E KB} ( information)) Information main 40 is transmitted to the information requester 41.

Information requester 41 which has received it, if indicated already. Expression of decoding the memory to have counterpart common key KB by _{D R (E KA (E KB} ( _{information)), D KB (D R} (E KA ( E _KB (information)))) ≠ information. In this way, even the information requester 41 who has already stored the half-split common key KB cannot obtain plain text information in the undecipherable state. It is possible to solve the dilemma in which the request to guarantee the right to delete the information to be deleted and the inability to delete it are in conflict with each other. Copying is prohibited so that information such as personal information recorded on the blockchain cannot be copied and pasted. If the processing is performed, the guarantee of the right to delete the information becomes more complete. It is not necessary to limit the information to double encryption using two keys KA and KB, and three or more keys (n). In this case, at least one of the n keys is replaced with a random number R, so that the decryption state is established.

The outline of the second embodiment described above will be described in more detail. Regarding the common points with the first embodiment, the repetition of the description will be omitted, and the differences will be mainly described. FIG. 15 corresponds to FIG. 2 in the first embodiment. With reference to transaction I in the blockchain, in this second embodiment, _EKA ( _EKB (personal information)), which is encrypted personal information, is recorded directly in the block. Therefore, in the second embodiment, the certified business operator 17 is unnecessary. Here, KA and KB are symmetric keys.

Next, with reference to FIG. 16, a flowchart of the main routine of the user terminal constituting the node 19 of the private chain 2 and the user terminal 16 constituting the node of the public chain 4 will be described. This main routine omits the flowchart of the operation process shown in the first embodiment, and shows only the flowchart of the operation process to be added or changed to the operation process shown in the first embodiment. In the user terminal 16 constituting the node of the public chain 4, the personal information recording process to the blockchain is performed by S160, the record unbreakable processing is performed by S161, and the personal information providing processing is performed by S162. At the user terminal constituting the node 19 of the private chain 2, personal information acquisition processing is performed by S170.

The personal information recording process on the blockchain is a process of recording personal information on the blockchain. The record unbreakable process is a process for exercising the deletion right to make the information unreadable. The personal information providing process is a process in which the user terminal 16 of the public chain 4 provides personal information to the user terminal of the private chain 2. The personal information acquisition process is a process in which the user terminal of the private chain 2 acquires personal information from the user terminal 16 of the public chain.

A flowchart of a subroutine program for recording personal information on the blockchain will be described with reference to FIG. 17 (A). In S174, a process of generating two random numbers is performed. For example, in the case of DES, two 56-bit random numbers are generated, and these 56-bit random numbers are used as the half-split common keys KA and KB. In the case of ADS, two 128-bit random numbers are generated, and these 128-bit random numbers are used as the half-split common keys KA and KB.

Next, in S177, _{a process of recording EKA} ( _EKB (personal information)), _EK2 (index + consideration for providing personal information), and a ciphertext identifier is performed on the blockchain. This ciphertext identifier is an identifier _{for identifying EKA} ( _EKB (personal information)), which is encrypted personal information, and corresponds to an electronic ID in the first embodiment.

Next, according to S183, KA and KB are associated with the ciphertext identifier and stored in the HDD 12 of the user terminal 16.

Next, a flowchart of a subroutine program for record unbreakable processing will be described with reference to FIG. 17 (B). S190 determines whether or not there is a ciphertext to be made unbreakable among the ciphertexts recorded on the blockchain (for example, _EKA ( _{EKB (personal information)), etc.).} If there is no ciphertext, it returns, but if there is a ciphertext that you want to make undecipherable, the control proceeds to S191, and the process of searching the HDD 12 for the half-split symmetric key KA that is stored in association with the ciphertext identifier of that ciphertext. Is done.

Next, a random number R is generated by S192. For example, in the case of DES, a 56-bit random number is generated. In the case of ADS, a 128-bit random number is generated. Next, S193 determines whether or not the generated random number R = KA. If the generated random number R is the same as the one-sided symmetric key KA stored in the HDD 12, the control returns to S192 and the random number is generated again. If NO is determined by S193, the control proceeds to S194, and a process of updating the half-split common key KA stored in the HDD 12 to R is performed.

Next, a flowchart of the subroutine program of the personal information providing process and the personal information acquisition process will be described with reference to FIG. In the user terminal of the private chain 2, it is determined by S198 whether or not the half-split common key KB of the personal information desired to be obtained is already stored. If the half-split common key KB of the personal information desired to be obtained has already been distributed from the user terminal 16 of the public chain 4 to the user terminal of the private chain 2, it is determined by S198 that there is a memory, and the control proceeds to S203, but it is still stored. If not, the control proceeds to S199.

In S199, a process is performed in which the ciphertext identifier of the personal information desired to be obtained is transmitted to the user terminal 16 of the public chain 4 to request the half-split common key KB. The user terminal 16 of the public chain 4 that receives it in S200 determines whether or not to perform a transaction that provides personal information specified by the ciphertext identifier by a smart contract (see S16), and a transaction that provides personal information. In this case, S201 returns a signature that agrees to provide personal information and a half-split common key KB corresponding to the ciphertext identifier.

At the user terminal of the private chain 2 that receives it in S202, the signature and the ciphertext identifier of the personal information desired to be obtained are transmitted to the user terminal 16 of the public chain 4 by S203. The user terminal 16 of the public chain 4 that receives it in S206 determines by a smart contract whether or not to carry out a transaction that provides personal information specified by the encryption text identifier (see S16), and a transaction that provides personal information. the S207 when performing the process of replying by calculating D _{KA (encrypted} personal information) or D _R (encrypted personal information) is performed. More specifically, although when the halves common key KA stored in the HDD12 of the user terminal 16 has already been updated to the random number R is returned by calculating the D _R (encrypted personal information), yet random numbers If it has not been updated to R, _{a process of calculating DKA} (encrypted personal information) and returning it is performed.

The user terminal of a private chain 2 received in S208 a response from the user terminal 16 of the public chain 4, the S209, D _KB (D _KA (encrypted personal information)) = plaintext or, D _KB (D _R (Encryption Personal information)) ≠ Processing to calculate plaintext is performed. Specifically, when _DKA (encrypted personal information) is received, _DKB ( _DKA (encrypted personal information)) = _DKB ( _DKA ( _EKA ( _EKB (personal information)))) ) = Compute the plaintext to obtain the plaintext personal information. On the other hand, (D _R (encrypted personal information)) when receiving the, _{D KB} (D _R (encrypted personal _{information)) = D KB (D R} (E KA (E KB ( personal information)))) ≠ Plaintext will be calculated, and personal information in plaintext cannot be obtained. As a result, it is possible to solve the dilemma in which the request for guaranteeing the right to delete the information to be deleted and the impossibility of deletion are in conflict with each other.
[Modification example]

(1) In the above explanation, _EKA ( _EKB (personal information)), which is encrypted personal information, is directly recorded in the blockchain, and this large amount of encrypted personal information is recorded in each node (all in public chain 4). If the information is stored in the node), there is an inconvenience that each node (user terminal) is required to have a huge storage capacity. As a means to solve this, a secret sharing technology for storing the divided data on a plurality of computers. The data is divided and fragmented, and each fragmented data is distributed and stored in a plurality of nodes. The data stored in each node is redundantly (duplicated) stored. Sufficient redundancy is provided. By having it, even if a part of the fragmented data is lost, there is no problem in restoration, and the difficulty of falsification as a blockchain can be guaranteed. Furthermore, the storage amount for storing the fragmented data is stored in each node. It may be controlled so as to be decided by intention, and the consideration may be given to each node in the form of a token or the like according to the amount of storage to be carried.

(2) In the above description, the encrypted personal information E _KA (E _KB (personal information)) was directly recorded on the blockchain, but in the modified examples shown in FIGS. 19 to 23, the encrypted personal information E _KA (E _KB (personal information)) is stored in the personal information DB 29 of the certified business operator 17, and the hash value of the encrypted personal information is recorded in the blockchain. Transaction I shown in FIG. 19 (A). The hash value of E _KA (E _KB (personal information)) + E _K2 (provided by index + 2.4 tokens) + cryptographic identifier and electronic signature are recorded in FIG. 19 (B). personal information DB29 of certified operators 17, encrypted personal information E _KA in association with the ciphertext identifier _{(E KB} (personal information)) is stored.

Based on FIG. 20, the main of the user terminal constituting the node 19 of the private chain 2 in this modification, the user terminal 16 constituting the node 19 of the public chain 4, and the server 18 of the authorized business operator 17. The flow chart of the routine will be described. The server 18 of the certified business operator 17 participates in the blockchain as a node 19. S215 executes personal information recording processing on the blockchain, S216 executes record unbreakable processing, S217 executes personal information providing processing, S220 executes hash value recording processing, and S221 executes ciphertext transmission processing. Is executed, and the personal information acquisition process is executed by S224.

In the personal information recording process on the blockchain, the user terminal 16 constituting the node 19 of the public chain 4 sends the encrypted personal information _EKA ( _EKB (personal information)) to the server 18 of the authorized business operator 17. This is the process of sending. The hash value recording process is a process in which _{the server 18 of the authorized business operator 17 that receives the encrypted personal information EKA} ( _EKB (personal information)) or the like stores it and generates the hash value and records it in the blockchain. Is. The record unbreakable process is a process for making a ciphertext such as encrypted personal information _EKA ( _{EKB (personal information)) that is desired to be unbreakable unbreakable.} The personal information providing process is a process executed by the user terminal 16 constituting the node 19 of the public chain 4 to provide personal information to the user terminal constituting the node 19 of the private chain 2. The personal information acquisition process is a process in which a user terminal constituting the node 19 of the private chain 2 acquires personal information. In the ciphertext transmission process, the server 18 of the authorized business operator 17 transmits a ciphertext such as encrypted personal information _EKA ( _{EKB (personal information)) to the user terminals constituting the node 19 of the private chain 2.} It is a process.

The details of each process will be described below based on the flowchart of each subroutine program, but the differences from the second embodiment will be mainly described.

A flowchart of a subroutine program of personal information recording processing and hash value recording processing on the blockchain will be described with reference to FIG. By S231, _EKA ( _EKB (personal information)) and _EK2 (index + consideration for providing personal information) are transmitted to the server 18 of the certified business operator 17. The server 18 receives it in S240, the _S241, the process of generating the hash value and the ciphertext identifier _{E KA (E KB} (personal information)) is performed. Next, S242 _{performs a process of recording the hash value of EKA} ( _EKB (personal information)), _EK2 (index + consideration for providing personal information), and the ciphertext identifier in the blockchain.

Next, S243 performs a process of transmitting the ciphertext identifier to the user terminal 16 of the public chain 4. In the user terminal 16 of the public chain 4 that receives it in S232, the processing of associating the half-split key KA and KB with the received ciphertext identifier and storing them in the HDD 12 is performed by S233. In the server 18 approved vendors 17, the _S244, the processing of storing the personal information DB29 in association _{E KA (E KB} (personal information)) and the ciphertext identifier is performed.

Since the record unbreakable process shown in FIG. 22 is the same as that already described in FIG. 17 (B) of the second embodiment, the repetition of the description will be omitted.

Next, a flowchart of the subroutine program of the personal information providing process, the personal information acquisition process, and the ciphertext transmission process will be described with reference to FIG. 23. In this modification, when the user terminal of the private chain 2 receives the half-split common key KB corresponding to the signature that agrees to provide the personal information and the ciphertext identifier from the user terminal 16 of the public chain 4 (S264), S265 , The received signature and the ciphertext identifier of the personal information desired to be obtained are transmitted to the server 18 of the authorized business operator 17. The server 18 of the authorized business operator 17 that received it in S266 searches the personal information DB 29 for the ciphertext corresponding to the ciphertext identifier after confirming the signature by S267 and transmits it to the user terminal 16 of the public chain 4. Processing is done.

In the user terminal 16 of the public chain 4 which received it at S268, the S269, and returns by calculating D _{KA (encrypted} personal information) or D _R (encrypted personal information)) to the user terminal of a private chain of processing Is done. More specifically, although when the halves common key KA stored in the HDD12 of the user terminal 16 has already been updated to R are returned by calculating the D _R (encrypted personal information), yet the R If it has not been updated, _DKA (encrypted personal information) is calculated and returned.

The user terminal of a private chain 2 received in S270 a response from the user terminal 16 of the public chain 4, the S271, D _KB (D _KA (encrypted personal information)) = plaintext or, D _KB (D _R (Encryption Personal information)) ≠ Processing to calculate plaintext is performed. Specifically, when _DKA (encrypted personal information) is received, _DKB ( _DKA (encrypted personal information)) = _DKB ( _DKA ( _EKA ( _EKB (personal information)))) ) = Compute the plaintext to obtain the plaintext personal information. On the other hand, (D _R (encrypted personal information)) when receiving the, _{D KB} (D _R (encrypted personal _{information)) = D KB (D R} (E KA (E KB ( personal information)))) ≠ Plaintext will be calculated, and personal information in plaintext cannot be obtained. As a result, it is possible to solve the dilemma in which the request for guaranteeing the right to delete the information to be deleted and the impossibility of deletion are in conflict with each other.

The server 18 of the certified business operator 17 may be connected to the user terminal of the private chain 2 and the user terminal of the pabook chain 4 via the Internet 1 without participating in the blockchain as the node 19. ..

(3) In the above explanation, the personal information owner holds the half-split common key KA (stored in the HDD 12 of the user terminal 16), but instead, the key registration center 30 is an example of a predetermined institution (third-party institution). The one-sided common key KA may be registered in the key DB 32 of the above. The one-sided common key KA is stored in the key DB 32 in a secret state. This modification will be described with reference to FIGS. 24 to 27.

With reference to FIG. 24, the server 31 of the key registration center 30 is connected to the Internet 1. In the key DB 32 connected to the server 31, a ciphertext identifier and a half-split common key KA are stored in association with each user's address, which is each node 19 of the public chain 4. Then, when the user requests to make the record unbreakable, the half-split common key KA stored in association with the ciphertext identifier corresponding to the requested record is updated to the random number R. In FIG. 24, the one-sided common key stored in association with the ciphertext identifier 307cd4 at address 0x6079dd is updated to the random number 1R2, and the one-sided common key stored in association with the ciphertext identifier 4arb56 at address 0x6080dd is The one-sided common key that has been updated to the random number 2Rn and stored in association with the ciphertext identifier e2c87r at the address 0x6978dd has been updated to the random number mR1.

Next, a flowchart of the main routine of the user terminal 16 of the public chain 4, the server 31 of the key registration center 30, and the user terminal of the private chain 2 will be described with reference to FIG. 25. Regarding the common points with the second embodiment, the repetition of the description will be omitted, and the differences will be mainly described.

In the user terminal 16 of the public chain 4, S468 executes the personal information recording process on the blockchain, S469 executes the record decryption request process, and S470 executes the decryption key providing process. In the server 31 of the key registration center 30, the key registration process is executed by S463, the record unbreakable process is executed by S464, and the data decryption process is executed by S465. At the user terminal of the private chain 2, the data acquisition process is executed by S460.

Next, a flowchart of the subroutine program of the personal information recording process and the key registration process on the blockchain will be described with reference to FIG. 26 (A). In the user terminal 16 of the public chain 4, the _{S479, E KA} and _{(E KB} (personal information)) _{E K2} (the consideration for providing index + identity) and the ciphertext identifier is recorded in a block chain, the S480, halves A process of transmitting the common key KA and the ciphertext identifier to the key registration center 30 is performed.

In the server 31 of the key registration center 30 that has received it in S474, a process of associating the received symmetric common key KA with the ciphertext identifier and storing it in the key DB 32 is performed by S475.

Next, a flowchart of the subroutine program of the record unbreakable request process and the record unbreakable process will be described with reference to FIG. 26 (B). In the user terminal 16 of the public chain 4, S494 determines whether or not there is a ciphertext to be unbreakable, and if not, returns, but in some cases, S495 requests a ciphertext identifier that requests unbreakable. Is transmitted to the server 31 of the key registration center 30.

The server 31 of the key registration center 30 that has received it in S485 performs a process of searching the key DB 32 for the half-split common key KA stored in association with the received ciphertext identifier. Next, a random number R is generated by S487, and it is determined by S488 whether or not the random number R = KA. When R = KA, the random number R is regenerated by S487, and when R ≠ KA, S489 performs a process of updating the half-split common key KA to R.

Next, with reference to FIG. 27, a flowchart of a subroutine program of the half-split common key provision process, the data acquisition process, and the data decoding process will be described. In the user terminal of the private chain 2, S503 performs a process of transmitting the ciphertext identifier of the data desired to be obtained to the server 31 of the key registration center 30. The server 31 of the key registration center 30 that has received it in S504 performs a process in S505 to search the blockchain for the ciphertext (encrypted personal information, etc.) corresponding to the ciphertext identifier. Next, S506 performs a process of searching for the half-split common key KA or R corresponding to the ciphertext identifier.

The next S507, processing of transmitting D _{KA (encrypted} personal information) or D _R (encrypted personal information) to the user terminal of a private chain 2 is performed. More specifically, although in the case where counterpart common key KA stored in the key DB32 key registration center 30 has already been updated to R are returned by calculating the D _R (encrypted personal information), yet If it has not been updated to R, _{a process of calculating DKA} (encrypted personal information) and returning it is performed.

The user terminal of a private chain 2 which has received it at S509, the S510, D _KA (D _KB (encrypted personal information)) = plaintext or, D _KA (D _R (encrypted personal information)) is calculated ≠ plaintext Processing is done. Specifically, when _DKA (encrypted personal information) is received, _DKB ( _DKA (encrypted personal information)) = _DKB ( _DKA ( _EKA ( _EKB (personal information)))) ) = Compute the plaintext to obtain the plaintext personal information. On the other hand, (D _R (encrypted personal information)) when receiving the, _{D KB} (D _R (encrypted personal _{information)) = D KB (D R} (E KA (E KB ( personal information)))) ≠ Plaintext will be calculated, and personal information in plaintext cannot be obtained. As a result, it is possible to solve the dilemma in which the request for guaranteeing the right to delete the information to be deleted and the impossibility of deletion are in conflict with each other. Moreover, since the process of updating the half-split common key KA to R is performed at the key registration center 30, it is easy to ensure the reliability that the deletion right is guaranteed by updating KA to R. For example, there is an advantage that it is easy to update KA to R under an audit by a predetermined organization.

(4) As another method for solving the dilemma in which the request for guaranteeing the right to delete the information to be deleted and the impossibility of deletion are in conflict with each other, the encrypted personal information stored in the personal information DB 29 of the certified business operator 17 is used. E _KA (E _KB (personal information)) may be deleted at the request of the personal information owner. In that case, although a hash value of personal information is recorded on the blockchain, a contradictory state occurs in which the corresponding personal information is not stored in the personal information DB 29, but this inconsistency should be tolerated. If possible, deleting personal information is also an effective means.

(5) Information that guarantees the right to delete is not limited to personal information, for example, information posted on SNS and blogs (including data of posted photos and posted videos), notarial acts such as wills and voluntary guardianship contracts, and private documents. Any information may be used, such as the articles of incorporation of the company, the articles of incorporation, and other information that requires a fixed date. Further, in the second embodiment, the information guaranteeing the deletion right is recorded using the blockchain, but the blockchain is only an example, and other information may be used for recording.

(6) The above-mentioned programs running on the user terminals 16 and the like constituting the node 19 of each blockchain and various servers may be downloaded and installed from a predetermined website or the like, but for example, recording of a CD-ROM 99 or the like. A person who purchases the CD-ROM99 or the like by recording it on a medium (non-transitory recording medium) and distributing it may install the program on the user terminal 16 and various servers (see FIG. 60). ..

(7) In the above description, the one encrypted once with the one-sided common key KA is encrypted twice with the one-sided common key KB and then decrypted once with the one-sided common key KB. Is decrypted again with the one-sided common key KA to make it plain text. However, the present invention is not limited to this, and the encryption may be performed with the half-split common key KA or KB a plurality of times, and the decryption may be performed a plurality of times with the one-sided common key KA or KB. Further, the half-split common key KA and KB are not limited to two, and three or more half-split common keys may be used.

Furthermore, the exclusive logical sum (exclusive or) of the one-sided common key KA and KB is calculated to generate one key K (KA (+) KB = K), and the personal information is encrypted with the key K. (E K _(personal information)), and registers the key DB32 halves common key KA key registration center 30 distributes the halves common key KB to the requester of the personal information. Upon receiving the request from the requester of personal information, the personal information owner sends the encrypted personal information to the server 31 of the key registration center 30, and the requester of personal information sends the distributed half-split common key KB to the key registration center 30. Send to server 31. The server 31 of the key registration center 30 calculates an exclusive logical sum (exclusive or) between the received one-sided common key KB and the one-sided common key KA registered in the key DB 32 to generate one key K. (KA (+) KB = K ), received encrypted personal information _{(E K} (personal information)) to decrypt with the key K on which the plaintext _{(D K (E K} (personal information)) = plaintext ), The plaintext personal information may be sent to the requester of the personal information. The above (+) is a symbolic representation of the exclusive OR.

The exclusive OR is only an example, and any algorithm may be used as long as it generates one key K from the half-split common key KA and KB.

Further, in the case of the above method in which the key K is generated by using an additive group such as exclusive OR, there is an advantage that the half-split common keys KA and KB can be periodically updated to maintain security. There is. For example, when one half-split common key KA is updated to KC, the other half-split common key KB = K (+) KC, which can be obtained by calculation. By updating the one-sided common key KA and KB in this way, not only can the leakage of the one-sided common key be countered, but also the personal information requester who once distributed the one-sided common key KB can again be on the blockchain. It is possible to prevent browsing by preventing the encrypted personal information from being decrypted. By executing such an update of the one-sided common key KB for each distribution of the one-sided common key KB, even if a person who receives the distribution of the one-sided common key KB circulates the one-sided common key KB to another person, the one-sided common key KB can be distributed. It is possible to make encrypted personal information on the blockchain unbreakable by a person who has been diverted. That is, the one-sided common key KB to be distributed can be used as a one-time key that can be used only once.

Further, the key is not limited to the common key, and a public key cryptosystem such as RSA or elliptic curve may be used.

Further, in realizing the above-mentioned key update, a cryptographic algorithm satisfying the following conditions may be adopted.
The plaintext is represented by M, the ciphertext is represented by C, and the encryption key is represented by KA, KB, KC and KD.
E _KA (E _KB (M)) = E _KC (E _KD (M)) = C
Algorithm that holds the formula of.

When such an algorithm is a common key cryptographic algorithm, when the plaintext common keys KA and KB are updated to KC and KD, the ciphertext C recorded in the blockchain is decrypted by the plaintext common keys KC and KD. As a result, the plaintext M can be obtained. On the other hand, in the case of the public key cryptographic algorithm, when the private keys KA and KB are updated to KC and KD, the ciphers recorded in the blockchain with the public keys PKC and PKD of the pair corresponding to the private keys KC and KD. The plaintext M can be obtained by decoding the sentence C.
(8) In the above description, the encrypted personal information data holders to information requester (D _{KA (encrypted} personal information) or D _{R (encrypted} personal information)) and transmits a counterpart common key KB (S201 , S207), the information requester himself decrypted the encrypted personal information using the one-sided common key KB to make it plain (S209), but the decryption using the one-sided common key KB was performed by a third-party organization (predetermined). Service organization) may be responsible. In this case, information holder sends a counterpart common key KB and encrypted personal information (D _{KA (encrypted} personal information) or D _{R (encrypted} personal information)) to the third party (predetermined service organization) , Decrypt at a third-party organization (prescribed service organization) and send it to the information requester.
[Characteristics of disclosure content]

Next, the features of the disclosed contents of the embodiments described above are listed below.
(Feature 1)
[Technical field]

Feature 1 relates to a processing system and a program for an information recording method that is difficult to falsify or erase, such as a blockchain.
[Background technology]

Blockchain has been generally known as an information recording method that is difficult to falsify or erase. For example, Japanese Patent Application Laid-Open No. 2018-128723 is used to record various information related to freight transportation using this blockchain.
[Overview of Feature 1]
[Problems that Feature 1 tries to solve]

However, recording information using such a blockchain is not only difficult to falsify, but also difficult to erase (hereinafter referred to as "non-erasable"). As a result, once personal information is recorded using the blockchain, even if the personal information owner wants to delete the personal information, it cannot be deleted and the right to erase personal information (so-called right to be forgotten) is impaired. There is a drawback.

In other words, there is a drawback that there is a dilemma in which the guarantee of the authenticity of the recorded information and the guarantee of the right to delete the information are antinomy.

Feature 1 was devised in view of such circumstances, and its purpose is to solve the dilemma in which the guarantee of the authenticity of the recorded information and the guarantee of the right to delete the information are antinomy.
[Means to solve problems]

The subject of feature 1 is shown, for example, as the following items.
(Item 1)
An encryption means (for example, S174, S177, or S228, S231, or S478, S479) that performs an encryption process for encrypting information to be recorded (for example, personal information).
Recording means (for example, S177 and blockchain, or S231, S240, S242, S244, blockchain and personal information DB29, or S479 and blockchain) for recording the information after the encryption process.
Decoding means (for example, S201, S202, S207 to S209, or S263 to S271) that decodes the information recorded by the recording means using the first key and the second key to obtain plain text information. , Or S500 to S510),
The information recorded by the recording means is provided with decodable means (for example, S191 to S194, or S250 to S254, or S494, S495, S485 to S489) that make the information recorded by the recording means undecryptable.
The decryption means includes a second key concealment holding means (for example, S194, or S233, or S475) that conceals and holds the second key (for example, the half-split common key KA).
The undecryptable means puts the second key held by the second key concealment holding means into another undecryptable state (for example, S190 to S194) by updating the second key to another one (for example, a random number R). , Or S250 to S254, or S494, S495, S485 to S489), processing system.

(Item 2)
The decryption means distributes the first key (for example, a half-split common key KB) to a person who wants to view information (for example, S200, S201, or S2562, S263, or S500, S501). The processing system according to item 1, further comprising.
(Item 3)
The processing system according to item 1 or 2, further comprising search means (for example, S37 to S40, S42 to S45) for searching information recorded by the recording means without making it plain text.

(Item 4)
The information recorded by the recording means includes personal information and includes personal information.
The undecryptable means puts the personal information of the personal information owner into the undecryptable state at the request of the personal information owner (for example, S190 to S194, or S250 to S254, or S494, S495, S485. ~ S489), the processing system according to any one of items 1 to 3.

(Item 5)
A step (for example, S174, S177, or S228, S231, or S478, S479) of performing an encryption process for encrypting information to be recorded (for example, personal information).
It was recorded by a recording means (for example, S177 and blockchain, or S231, S240, S242, S244, blockchain and personal information DB29, or S479 and blockchain) that records the information after the encryption process. A decoding step (for example, S201, S202, S207 to S209, or S263 to S271, or S500 to S510) in which information is decrypted using the first key and the second key to obtain plain text information. When,
The steps (for example, S191 to S194, or S250 to S254, or S494, S495, S485 to S489) that make the information recorded by the recording means undecryptable can be performed.
Let the computer run
The decryption step includes a step (eg, S194, or S233, or S475) of concealing and holding the second key (for example, the half-split common key KA).
The step of making the decryption impossible state makes the second key held by the holding step update to another one (for example, a random number R) to make the decoding impossible state (for example, S190 to S194). , Or S250 to S254, or S494, S495, S485-S489), the program.

(Effect of feature 1)
According to the feature 1, the dilemma in which the guarantee of the authenticity of the recorded information and the guarantee of the right to delete the information conflict with each other can be solved as much as possible.
(Feature 2)
[Technical field]

Feature 2 relates to, for example, a smart contract used in a blockchain or the like.
[Background technology]

A smart contract is a computer protocol intended for smooth contract verification, condition confirmation, execution, execution, and negotiation, and has been conventionally used in blockchains and the like. This smart contract has been conventionally known as a device for automating contracts, transactions, etc. (for example, Japanese Patent No. 6403177).
[Overview of Feature 2]
[Problems that Feature 2 tries to solve]

In the field of such smart contracts, advanced smart contracts that can execute legal acts such as various contracts represented by sales contracts and lease contracts or various transactions on behalf of the user are desired.

The purpose of Feature 2, which was conceived in view of such circumstances, is to provide an advanced smart contract that can execute legal acts on behalf of the user.
[Means to solve problems]
The subject of feature 2 is shown, for example, as the following items.
(Item 1)
Machine learning means (for example, S80 to S82) that input information on legal acts performed by a plurality of natural persons or corporations as data for machine learning to generate a general model, and
A means for personalizing the general model into a model suitable for the user, which is a personalization means (for example, S86 to S88, or S94 to) for personalizing based on information about a legal act performed by the user. S98) and
A computer comprising a smart contract generating means (eg, S86-S88, or S94-S98) that generates a smart contract for performing a legal act on behalf of the user using the personalized model. system.

(Item 2)
It is a means for personalizing a general model generated by inputting information on legal acts performed by multiple natural persons or corporations as data for machine learning into a model suitable for the user, and is performed by the user. Personalization means (eg, S80-S82, S86-S88, or S94-S98) that personalize based on information about legal acts.
A computer comprising a smart contract generating means (eg, S86-S88, or S94-S98) that generates a smart contract for performing a legal act on behalf of the user using the personalized model. system.

(Item 3)
It is a means for personalizing a general model generated by inputting information on legal acts performed by multiple natural persons or corporations as data for machine learning into a model suitable for the user, and is performed by the user. Personalization means (eg, S80-S82, S86-S88, or S94-S98) that personalize based on information about legal acts.
A computer system comprising a service providing means (for example, S99) that provides a service of using the personalized model as a smart contract to execute a legal act on behalf of the user.

(Item 4)
Reinforcement learning means (for example, S99) for learning a measure for maximizing the accumulation of the reward by giving a reward for the legal act executed by the service providing means (for example, S99) to the executed model. For example, the computer system according to item 3, further comprising S105 to S108).

(Item 5)
Predetermined themes (eg, policies and laws that the government is trying to adopt (eg, reduced tax rates associated with consumption tax increases, revised immigration laws, withdrawal from the UK's European Union), partial or full basic income Simulation of trading in the investment market such as stock trading and futures trading, company management simulation, consumption behavior simulation, etc.) on the assumption that the target adoption, revision of Article 9 of the Japanese Constitution, etc.) has been adopted in the computer. It is a computer system that progresses reinforcement learning by going to
Selection means (for example, S344, S345) for selecting a group of users belonging to a plurality of personas matching the theme of the simulation, and
A collection means (for example, S346) for grouping a group of users selected by the selection means for each of the plurality of personas and collecting information on legal acts performed by the group of users for each group.
A generation means (for example, S347) that performs machine learning using the collected information on legal acts as learning data to generate a trained smart contract model group for each persona.
It is provided with a simulation means (for example, S336 to S339) for executing a simulation of performing a legal act between the generated trained smart contract models in a computer.
The simulation means is an enhanced learning means (eg, for example,) in which the learned smart contract model learns how to maximize the accumulation of the reward by giving a reward for the executed legal act to the executed learned smart contract model. A computer system including S336, S338).

(Item 6)
A computer system that performs simulations in a computer to advance reinforcement learning.
The trained smart is created by executing a simulation of performing legal actions between trained smart contract models generated by machine learning in a computer and giving a reward for the executed legal actions to the executed learned smart contract model. A computer system comprising a strengthening learning means (eg, S336, S338) in which a contract model performs a simulation strengthening learning process that learns measures to maximize the accumulation of rewards.

(Item 7)
Item 6 further includes a selection means (for example, S340) for selecting a learned smart contract model to be actually used based on the results of the reinforcement learning result by the reinforcement learning means from the trained smart contract model group. Described computer system.

(note)
The above-mentioned "data for machine learning" for model generation and "data for machine learning" used for personalization need only include "information on legal acts", other than "information on legal acts". Information (for example, access history to a website, GPS location information, etc.) may also be included. The above-mentioned "smart contract generating means" includes, for example, a case where artificial intelligence such as a personal assistant who has been machine-learned based on information on legal acts plays a role as a smart contract.

(Effect of feature 2)
According to Feature 2, it is possible to provide an advanced smart contract that can execute various legal acts on behalf of the user himself / herself.

(Feature 3)
[Technical field]
Feature 3 is, for example, the policies and laws that the government is trying to adopt (for example, the reduced tax rate due to the consumption tax increase, the revised immigration control law, the withdrawal from the EU (European Union) in the United Kingdom, and the partial or full adoption of basic income. , Amendment of Article 9 of the Japanese Constitution, etc.), Marketing-related conditions (for example, setting prices and consideration for new products (including financial products and life insurance), new service prices, promotion effects by various media, etc.), Investment market-related Conditions (for example, weather conditions in futures trading, monetary tightening policy in the stock market, etc.) are set, and simulations are performed in the computer under those conditions to predict in advance what kind of simulation results will be obtained. Regarding computer systems.
[Background technology]

As a computer system of this kind, a new statement of changes in net assets has been set up to help policy-level decision-making by clarifying future liabilities and resources available to the public in the future, and for the year. An accounting method has been proposed that can clarify the asset fluctuations due to the policy decision-making and simulate the future burden on the people (for example, Japanese Patent Application Laid-Open No. 2006-155233).
[Overview of Feature 3]
[Problems that Feature 3 tries to solve]

In the field of such simulations, a computer system that can faithfully execute simulations that faithfully imitate the activities of natural persons and corporations in the real world and derive simulation results with minimal deviation from the real world is desired.

The purpose of Feature 3, which was conceived in view of such circumstances, is to enable simulations that faithfully imitate the activities of natural persons and corporations in the real world.
[Means to solve problems]

The subject of feature 3 is shown, for example, as the following items.
(Item 1)
Predetermined conditions (eg, policies and laws that the government is trying to adopt (eg, reduced tax rates associated with consumption tax increases, revised immigration laws, withdrawal from the UK's European Union), partial or full basic income Recruitment, revision of Article 9 of the Japanese Constitution, etc.), marketing-related conditions (for example, setting of prices and consideration for new products (including financial products and life insurance), promotion effects by various media, etc.), investment A computer system that performs simulations under market-related conditions (for example, weather conditions in futures trading, monetary tightening policies in the stock market, etc.) in a computer.
Selection means (for example, S144, S145) for selecting a group of users belonging to a plurality of personas that match the conditions of the simulation, and
A collection means (for example, S146) for grouping a group of users selected by the selection means for each of the plurality of personas and collecting information on legal acts performed by the group of users for each group.
A generation means (for example, S147) that performs machine learning using the collected information on legal acts as learning data to generate a trained smart contract model group for each persona.
A simulation means (for example, S136 to S139) that executes a simulation of performing a legal act between the generated trained smart contract models in a computer, and
A derivation means (for example, S140) for deriving the result of simulation by the simulation means is provided.
The simulation means is an enhanced learning means (eg, for example,) in which the learned smart contract model learns how to maximize the accumulation of the reward by giving a reward for the executed legal act to the executed learned smart contract model. A computer system including S136, S138).

(Item 2)
A computer system that performs simulations in a computer
A simulation of performing legal actions between trained smart contract models generated by machine learning (for example, S144 to S146) is executed in a computer, and the reward for the executed legal actions is given to the executed learned smart contract model. By giving, the trained smart contract model learns a strategy for maximizing the accumulation of the rewards.
A computer system including a derivation means (for example, S140) for deriving the result of a simulation of performing a legal act between trained smart contract models in which reinforcement learning by the reinforcement learning means has progressed.
(Effect of feature 3)

According to Feature 3, it is possible to simulate the activities of natural persons and corporations in the real world as faithfully as possible.
[Third Embodiment]

Next, the third embodiment will be described. In this third embodiment, for example, an optimum solution that predicts the future can be derived by performing a simulation in a mirror world (cyber space) composed of digital twins in the real world as a simulation environment. It relates to a system for deriving the optimum solution of incentive design in DAO (Decentralized Autonomous Organization) and performing AI machine learning (for example, reinforcement learning).

A digital twin is a digital representation of a real-world entity or system. The mirror world is a mirror world composed of digital twins in which all information in the physical world (real world) such as real nations, cities, societies, local governments, organizations such as companies, and people is digitized. .. Specifically, machine learning (for example, an agent) for learning a person's digital twin as knowledge of the life log of the person's behavior (both real and virtual behavior) and assisting the optimum behavior for the person. It is composed of assistant AI (hereinafter referred to as "personal AI") who has performed reinforcement learning by. This reinforcement learning is a multi-agent reinforcement learning in which a plurality of personal AIs cooperate to perform reinforcement learning. The personal AIs of the people who make up the organization of the company in the real world make up the digital twin of the organization, and the personal AIs of the people who make up the local government in the real world make up the digital twin of the local government. , The personal AIs of the people who make up the city in the real world make up the digital twins of the city, and the personal AIs of the people who make up the nations of the real world make up the digital twins of the country.

In this third embodiment, organizations and people such as real nations, cities, societies, local governments, companies, etc. in the real world participate in the construction of a mirror world as a simulation environment and cooperate in the construction. Prepare a mechanism. Specifically, by performing various simulations in the mirror world, the personal AI of the digital twins participating in the simulation is machine-learned (for example, reinforcement learning), and the learned personal AI that has been learned more highly is realized. Give back (feedback) to the world. With the enjoyment of this merit as an incentive, we encourage the organizations and people of real-world nations, cities, societies, local governments, companies, etc. to take the initiative in participating and cooperating in the construction of the mirror world.

With reference to FIG. 28, the mirror world data is stored in the data center 45 in which a plurality of mirror world servers (including storage servers) 46 are installed. Since the hardware configuration of the mirror world server 46 is the same as the hardware configuration of the user terminal 16 shown in FIG. 1, the illustration and description thereof will be omitted here. Digital twins in which all the information of the real nation (for example, Japan 49), cities 50, society, local governments, companies and other organizations, people and the earth 48 in the real world 47 is digitized (real nation digital twin (for example, Japan) The entire mirror world 51 composed of the national digital twin 53), the urban digital twin 54, organizations such as society, local governments, companies, people, and the earth digital twin 52) is stored in the data center 45 as digital data.

In the data center 45, simulation is performed using this mirror world 51 as a simulation environment, and for example, the optimum solution foreseeing the future is derived by simulation optimization. Simulations include, for example, the policies and laws that the government intends to adopt (eg, reduced tax rates associated with consumption tax increases, revised immigration laws, withdrawal from the EU (European Union) in the United Kingdom, partial basic income, or Assuming that full adoption, revision of Article 9 of the Japanese Constitution, etc.) has been adopted, transaction simulations in the investment market such as stock trading and futures trading, company management simulations, consumption behavior simulations, etc. can be considered. Furthermore, promotion simulations of new products (including financial products and life insurance) and new services by various media may be used. The optimum solution derived by simulation optimization is fed back (reduced) to the real world, and the benefits of the optimum solution are provided to the real world. In addition, the learned personal AI that has been machine-learned (for example, reinforcement learning) by simulation is returned to the real world so that tasks by more advanced learned personal AI can be performed.

By linking this personal AI and a smart contract, the above-mentioned linkage type AI smart contract is configured. The data center 45 is connected to the Internet 1 shown in FIGS. 1 and 24. In FIG. 28, various blockchains 2, 3, 4, SNS 19, key registration center 30, and the like are not shown.

FIG. 29 shows a specific example of the urban digital twin 54 in the mirror world 51. Within the city 50 of Real World 47, there are ABC56 Co., Ltd., Taro 55 who is a person, and Taro's family 56. Corresponding urban digital twins 54 include ABC digital twins 59, Taro digital twins (Taro's personal AI) 57, and Taro's family digital twins 58. Urban digital twin data consisting of these data is stored in the mirror world server 46. If there are changes to various objects such as ABC56 Co., Ltd. in Real World 49, Taro 55 who is a person, and Taro's family 56 (for example, personnel transfer at a company, employment or retirement, marriage or childbirth about a person, etc.) Various corresponding digital twins will be updated with the changed contents. Such urban digital twin data is stored in the data center 45 for each city and becomes the data of the digital twin 53 of the Japanese state 49, and the urban digital twin data in each country is stored in the data center 45 for each city. It becomes the digital twin data of each country, and the data of all the digital twin data becomes the data of the digital twin 52 of the earth 48.

As a specific example, the mirror world server 46 has a Taro digital twin (Taro's personal AI) 57 as a name: Taro, AI identification number: 82km9, personal AI data, and Taro's personal data (for example, life log, profile, preference). Data, electronic medical record data, vital data, etc.) are stored. As Taro's family digital twin 58, names: Taro, Sakura, Shiro, family composition: husband, wife, eldest son, AI identification number: 82km9, 11zk9, gf43y are memorized. As ABC Digital Twin 59 Co., Ltd., names: Taro, Hanako ... Saburo, titles: CEO, managing director, general manager ... regular employees, AI identification numbers: 82km9, ba935, 2es14, ... 9w1c2 are memorized. There is.

The flowchart of the main routine program of the user terminal 16 and the mirror world server 46 will be described with reference to FIGS. 30 to 33. With reference to FIG. 30A, the CPU 10 of the user terminal 16 executes the member registration request process S555, the simulation preparation response process S556, and the simulation response process S557 for requesting participation registration in the mirror world 51 as a simulation environment. The CPU 10 of the mirror world server 46 executes the member registration process 550, the simulation preparation process S551, and the simulation process S552.

The member registration process and the member registration request process will be described with reference to FIG. 30B. Both of these processes are for registering members who want to participate as digital twins in a simulation using the mirror world 51 as a simulation environment. In the member registration request process, the CPU 10 of the user terminal 16 determines whether or not to apply for registration in S560, and if it determines that the application for registration is not performed, the member registration process ends and returns. If it is determined that a registration application will be made, in S561, the predetermined items required for the registration application will be transmitted to the mirror world server 46, and if there is a person who does not have a personal AI, that fact and the blockchain address of that person will be sent. Is also transmitted to the mirror world server 46. Specifically, the prescribed items required for registration application are the AI identification number and personal AI data of the person's personal AI in the case of the person's digital twin, and the family name and family composition in the case of the family's digital twin. The AI identification number of the company, the name and position of the employee in the digital twin of the company, the AI identification number of each, and the like.

The CPU 10 of the mirror world server 46, which receives it in S565, determines in S566 whether or not it already owns the personal AI. If the information transmitted by S561 includes the information "not having a personal AI", the control proceeds to S567 to generate and sell the personal AI, but "has a personal AI". If the information "not done" is not included, the control proceeds to S568, and the predetermined items including the AI identification number sent by S562 are registered in the mirror world 51.

The personal AI generation / sales process shown in S567 will be described with reference to FIG. The CPU 10 of the mirror world server 46 transmits the transaction data recorded in the blockchain address (the blockchain address of the user who does not have the personal AI) received by S565 in S573 and the posted data such as SNS from the blockchain. collect. Next, in S574, machine learning is performed using transaction data and posted data such as SNS as learning data to generate a learned personal AI. Next, in S575, the learned personal AI is sold to the corresponding user.

The simulation preparation process shown in S551 and the simulation preparation response process shown in S555 will be described with reference to FIG. The CPU 10 of the mirror world server 46 determines in S577 whether or not the simulation request has been received. If it is determined that the simulation has not been received, this simulation preparation process ends and returns. When it is determined that the simulation request has been received, the control proceeds to S578, and the process of determining the personal AI group and the digital twin that match the requested simulation is performed. For example, in the case of the consumption behavior simulation under the reduced tax rate accompanying the consumption tax increase mentioned above, it is a personal AI group corresponding to general consumers, and according to population statistics such as gender, age group, region, and annual income. The personal AI group and the consumer goods manufacturer digital twins and retailer digital twins that are subject to the reduced tax rate are identified at the same rate. Next, a process of requesting the consent of the simulation is performed on the indexed personal AI group and the digital twin. Specifically, the content of the simulation is transmitted to the user terminals 16 of each of the indexed personal AI group and the user group corresponding to the digital twin, and it is asked whether or not to agree.

The CPU 10 of the user terminal 16 of each of the indexed personal AI group and the user group corresponding to the digital twin receives the transmitted contents of the simulation in S580, and agrees to participate in the execution member of the simulation in S581. Determine whether or not to do so. This determination may be determined by the personal AI, or may be determined by the user himself. If it is determined that they do not agree, the simulation preparation response process ends and returns. However, if it is determined that they agree, S582 returns a message to that effect to the mirror world server 46.

The CPU 10 of the mirror world server 46 that has received it in S583 determines whether or not the amount of consent required to execute the requested simulation has been obtained. If it is determined that the result has been obtained, in S584, the AI group and the digital twin for which consent has been obtained are copied and registered in the mirror world 51 as a simulation target. The registered state is shown in FIG. 29 above.

On the other hand, if it is determined that consent has not been obtained from the required personal AI group and digital twin, control proceeds to S585, and the persona group (manufacturer or retailer's) that matches the lacking personal AI group and digital twin. The process of setting the persona corresponding to the digital twin) is performed, the user group belonging to each persona (including the user group engaged in the manufacturer or the dealer) is selected in S586, and the user belonging to each persona in S587. The group is grouped and the transaction data of the user group (including the transaction data as a manufacturer or a dealer) is collected from the blockchain for each group, and in S588, machine learning is performed using the transaction data as learning data for each persona. After generating and replenishing the learned personal AI group and digital twin, the process proceeds to S584. S585 to S588 are the same processes as S344 to S347 of FIG. 9B, and the repetition of detailed description is omitted here.

Next, specific control of the simulation process shown in S552 and the simulation response process shown in S557 will be described with reference to FIG. 33. S593 to S595 are the same processes as S336, S338, S339 in FIG. 9 and S136, S138, S139 in FIG. 12, and detailed description thereof will be omitted here. In S596, the simulation result is notified to the simulation requester. Specifically, the simulation result is transmitted to the user terminal 16 of the simulation requester. Next, in S597, each personal AI used in the simulation (including a personal AI engaged in a digital twin of a company organization or the like) is transmitted to the user terminal 16 of each owner.

The CPU 10 of the user terminal 16 that has received it in S598 determines in S599 whether or not to erase the received personal AI. The received personal AI is a learned AI that has been reinforcement-learned (machine-learned) by participating in a simulation, and its performance is improved by that amount, and it is possible to execute advanced task processing. However, depending on the content of the simulation, the user may have received undesired reinforcement learning (machine learning). In such a case, S599 determines YES, and S601 erases the received personal AI. do. On the other hand, if the simulation is the content desired by the user and it is determined that the received personal AI has received the desired reinforcement learning (machine learning), the control proceeds to S600 and the received learned personal AI is overwritten and saved. .. As a result, the user has an advantage that he / she can obtain a personal AI with improved performance by receiving desirable reinforcement learning (machine learning). With the enjoyment of this advantage as an incentive, organizations and people of real-world nations, cities, societies, local governments, companies, etc. can be encouraged to take the initiative in participating and cooperating in the construction of the mirror world. By overwriting and saving the personal AI by this S600, the data is updated in the mirror world server 46 to the digital twin of the new personal AI after the overwriting and the digital twin of the organization consisting of the new personal AI (FIG. 29). In addition, instead of overwriting and saving, both the existing personal AI and the learned personal AI may be stored together and used properly as needed.

Next, a system for deriving an optimum solution for incentive design in DAO by performing a simulation in a mirror world will be described with reference to FIGS. 34 to 59. FIG. 34 (A) is a schematic diagram of a multi-service DAO construction system. For example, Bitcoin is a type of DAO, but a node (minor) has only one type of service, mining (competition of bookkeeping rights), and gives an incentive to give Bitcoin to those who succeed in mining. This will add blocks and the Bitcoin system will continue autonomously. On the other hand, a DAO having multiple types of services is called a multi-service DAO. For example, in the case of DAO, which is a company organization, there are multiple services such as material procurement, assembly, promotion, and sales, and it is optimal to distribute what proportion and how much reward to the nodes that performed these multiple services. It becomes a difficult problem whether it will be an incentive design. A system for deriving an optimum solution for incentive design in such a multi-service DAO will be described.

With reference to FIG. 34 (A), the mirror world server 46 that stores the multi-service DAO data is a DAO agent 61, a persona agent group 62 that performs service 1, a persona agent group 63 that performs service 2, ... The data of the persona agent group 64 that performs n is stored. Further, the mirror world server 46 also stores the types of rewards r1, r2, ... rn given to each persona agent group in accordance with reinforcement learning.

The terminal 16 of the multi-service DAO builder downloads and installs the DAO agent 61, the necessary persona agent group, and the types of rewards r1, r2, ... rn from the mirror world server 46. The terminal 16 is each node 19 constituting the public chain 4. The digital twin 66 of the multi-service DAO 65 operated in the public chain 4 composed of each node 19 is subjected to simulation reinforcement learning in the mirror world 51, and the optimum solution of the incentive design in the multi-service DAO is derived. The optimum solution of the incentive design is applied to the actual multi-service DAO65 in the real world 47 to create the multi-service DAO65 of the optimum incentive design. The simulation reinforcement learning in the mirror world 51 is executed in the mirror world server 46. The control will be described below.

With reference to FIG. 34 (B), the CPU 10 of the terminal 16 performs the simulation reinforcement learning preparation response processing in S606 and the simulation reinforcement learning response processing in S607.

The CPU 10 of the mirror world server 46 performs simulation reinforcement learning preparation processing in S611, simulation reinforcement learning processing in S612, and DAO agent reinforcement learning processing in S613.

Specific control of the simulation reinforcement learning preparation process shown in S611 and the simulation reinforcement learning preparation response process shown in S606 will be described with reference to FIG. 35. In the simulation reinforcement learning preparation response processing, the CPU 10 of the terminal 16 determines in S615 whether or not to request the simulation reinforcement learning. If not requested, this simulation learning preparation and strengthening answer processing is completed and returned. When requesting, the control proceeds to S616, and the multi-service DAO data is transmitted to the mirror world server 46 to request. This multi-service DAO data includes service types. For example, in the case of the innovation-induced DAO described later in FIGS. 36 to 42, there are five types of services: idea idea, improvement idea, commercialization, infringement detection, and token purchase, and these services are transmitted.

The CPU 10 of the mirror world server 46 that receives it in S620 sets a persona group that matches each of the multi-services by S621. For example, in the case of the above innovation-induced DAO, people who often think of inventions as a group of idea and improvement idea, people who are interested in commercialization as a group of commercialization personas, and patent law and copyright as a group of infringement discovery personas. People who are familiar with the law, people who are interested in investing as a group of personas for purchasing tokens, etc. can be considered.

Next, in S622, a user group belonging to each persona is selected. For example, in the case of the above innovation-induced DAO, the user group listed as the inventor of the patent application as the user group belonging to the persona group of the idea idea and the improvement idea, and the company manager as the user group belonging to the commercialization persona group. The user group of patent attorneys and lawyers as the user group belonging to the persona group of infringement detection, the user group of users who have purchased virtual currency such as bit coin as the user group belonging to the persona group of token purchase, etc. can be considered. ..

Next, in S623, the user groups belonging to each persona are grouped and the transaction data of the user group is collected from the blockchain for each group, and in S624, machine learning is performed using the transaction data as learning data to learn for each persona. Generate a group of completed persona agents. Both of these controls are the same processes as those of S346 and S347 of FIG. 9B described above, and the repetition of detailed description thereof will be omitted here. Next, in S625, the persona agent group is deployed in the multi-service DAO 65 to generate the multi-service DAO digital twin 66, and the persona agent group is registered in the mirror world 51 as a simulation target. The state is shown in FIG.

With reference to FIG. 36, a digital twin 66 of a multi-service DAO 65 composed of a public chain is constructed in a mirror world 51. In the multi-service DAO digital twin 66, a persona agent group is deployed in S625, and one persona agent is deployed for each node. The identification numbers of those persona agents are classified for each service (idea idea, improvement idea, commercialization, infringement detection, token purchase) and stored in the mirror world server 46. For example, kc29m, 1w13a, ... 9nad8 are stored in the mirror world server 46 as the identification number of the persona agent for the idea idea service. This multi-service DAO 65 is the above-mentioned innovation-induced DAO, and the multi-service DAO will be described below by taking the innovation-induced DAO as an example.

Further, the mirror world server 46 also stores a DAO agent that gives a reward (incentive) for the actions of each persona agent. This DAO agent derives the optimum solution for incentive design by performing reinforcement learning (machine learning) on the distribution ratio of the reward to be given and the reward amount. A schematic system of the reinforcement learning (machine learning) is shown in FIG. 37.

With reference to FIG. 37, if the persona agent group posts the origin idea as actions a11, a12, ... a1n, the environment state S1 is input to the DAO agent 61, and the persona agent group performing the idea idea service. Rewards r11, r12, ... r1n are given to 67. This idea-creating service is a broad concept that includes the ideas of dreams, ideas, business plans, technical ideas, copyrighted works, and the like. The environmental state S1 is also given to the persona agent group 67 who has performed the idea-creating service. The state S1 of this environment is, for example, the content of the origin idea posting, the floating exchange rate price of the token A1 given as a reward to each of the persona agents 68 who performed the idea idea service, and the like.

If the persona agent group 68 performs actions a21, a22, ... a2 such as posting an improvement plan to the above origin idea, the environmental state S2 is input to the DAO agent 61, and the persona who performed the improvement plan service. Rewards r21, r22, ... r2n are given to the agent group 68. The environmental state S1 is also given to the persona agent group 68 who performed the improvement proposal service. The state S2 of this environment is, for example, the content of the improvement proposal submission, the number of “likes” given to the improvement proposal submission, and the like. The entity that gives this "like" is limited to, for example, only those who have purchased the token A1 given as a reward for posting the origin idea (persona agent group 71). In this way, the reason for limiting (restricting) the granting party of "Like!" To interested parties (stake folders) is to prevent fraudulent activities. If the number of "Like" grantees is expanded without limit, for example, the person who posted the improvement plan (Persona agent group 68) colludes with a large number of people (Persona agents) and a large number of "Like!" This is to prevent fraudulent acts such as having people attach. The person who gave "Like" to the persona agent group 69 who performed the commercialization service and the persona agent group 70 who performed the infringement countermeasure service also purchased the token A1 for the same reason (persona agent group 71). Limited to only.

If the persona agent group 69 who performed the commercialization service for the above origin idea performs actions a31, a32, ... a3n, the environmental state S3 is input to the DAO agent 61, and the persona who performed those commercialization services. Rewards r31, r32, ... r3n are given to the agent group 69. Actions a31, a32, ... a3n of the persona agent group 69 include, for example, posting a business plan, posting the progress of commercialization, posting the actual status of commercialization, and profit from the commercialized business. Posting of the amount can be considered. The environmental state S3 is also given to the persona agent group 69 who performed the commercialization service. This environmental state S3 is, for example, the number of “likes” given to the posting of the business plan, the posting of the progress of commercialization, the posting of the profit amount from the commercialized business, and the like.

If the persona agent group 70 who performed the infringement countermeasure service for the above origin idea performs actions a41, a42, ... a4n, the environmental state S4 is input to the DAO agent 61, and the persona who performed the infringement countermeasure service. Rewards r41, r42, ... r4n are given to the agent group 70. As the actions a31, a32, ... a3n of the persona agent group 70, for example, an infringement discovery report posting, an infringement handling report posting, a license negotiation report posting, and the like can be considered. Furthermore, the service of submitting a report of a patent application and the service of submitting a report of the acquisition of rights, which are the premise of these services, may be included. The environmental state S4 is also given to the persona agent group 70 who has performed the infringement countermeasure service. The state S4 of this environment is, for example, the number of “likes” given to the infringement discovery report posting, the infringement countermeasure report posting, the license negotiation report posting, and the like.

If the persona agent group purchases the token A1 granted to the origin idea idea service as actions a51, a52, ... a5n, the environment state S5 is input to the DAO agent 61 and those tokens are purchased. Rewards r51, r52, ... r5n are given to the persona agent group 71 who performed the service. The environmental state S5 is also given to the persona agent group 71 who has performed the token purchasing service. The state S5 of this environment is, for example, the number of tokens purchased (or the purchase price). Persona agent group 71 consumes virtual currency (for example, ETH of Ethereum) to purchase token A1. The purchased token can be converted (cashed) into virtual currency according to the price in the floating market, and the virtual currency can be converted (cashed) into legal currency such as yen or dollar according to the price in the floating market.

The rewards r1 to r5 given to each persona agent group are determined by the DAO agent 61 based on the reward table (see FIG. 39 (A)). For the persona agent who performed the idea idea service, r1 = A1 + B1 ・ b + G1 ・ g, for the persona agent who performed the improvement service, r2 = A2 ・ e ＋ B2 ・ b + G2 ・ g, business R3 = A3 ・ e ＋ B3 ・ b for the persona agent who performed the conversion service, r4 = A4 ・ e ＋ B4 ・ b + G4 ・ g for the persona agent who performed the infringement countermeasure service, and token purchase service. For the persona agent, r5 = B5 ・ b + G5 ・ g is determined.

Here, A2 to A4, B1 to B5, G1, G2, G4, and G5 are coefficients, and the DAO agent 61 converges to the optimum one by reinforcement learning. A1 is token, g is license revenue, e is the number of likes, and b is commercialization revenue.

Only the license revenue g or the commercialization revenue b generated after each persona agent group 68 to 71 performs the service is considered as the reward r2 to r5. This is to prevent fraudulent acts such as performing improvement services or token purchase services later on an origin idea for which license revenue g or commercialization revenue b has already been generated.

In addition, a group of persona agents who have performed improvement services, commercialization services, or infringement countermeasure services may also perform token purchase services. Further, the persona agent group 67 who has performed the idea idea service may also perform the improvement service, the commercialization service, or the infringement countermeasure service.

The details of the DAO agent reinforcement learning process shown in S613 will be described with reference to FIG. 38. In this process, the DAO agent 61 performs reinforcement learning by itself to optimize the rewards r1 to r5. The DAO agent 61 determines in S630 whether or not each act a of the persona agent group has been received. If it is not received, it proceeds to S632, but if it is determined that it is received, the control proceeds to S631 and each received action a is stored.

In S632, it is determined whether or not "Like!" Is given, and if it is not given, the process proceeds to S634, but if it is determined that it is given, in S633, each persona agent likes it! Remember e. In S634, it is determined whether or not there is commercialization profit, and if it is not, the process proceeds to S636, but if it is determined that there is, in S635, the commercialization profit b is stored. In S636, it is determined whether or not there is a license profit g, and if it is not present, the process proceeds to S638, but if it is determined that there is, the license profit g is stored in S637.

In S638, it is determined whether or not it is time to calculate the reward, and if it is not determined, it is determined that the process proceeds to S640. Calculate r1 to r5 and assign it to the corresponding persona agent. In S640, it is determined whether or not the learning update time has come, and if not, the DAO agent reinforcement learning process ends and returns. When it is determined that the learning update time has come, in S641, the total grant price TT of the token A1 granted as a reward and the current total price TB in the floating exchange rate of the token granted are calculated, and in S642, the TB / TT The reward R of the DAO agent is calculated from the value. For example, the value of TB / TT at the time of the previous learning update is compared with the value of TB / TT at the time of the current learning update, and when the value of TB / TT at the time of the current learning update is larger, a large reward R And if it is small, it will be a small reward R. As a result, the reward R that the DAO agent 61 can obtain increases when the total price TB in the floating exchange rate of the token rises, and decreases when the total price TB in the floating exchange rate of the token falls.

Next, in S643, based on the reward R, the processing of obtaining the actions A1 to A4, B1 to B5, G1, G2, G4, G5 according ^{to the optimum policy π * by TD learning is performed, and in S644, the reward table is performed.} A1 to A4, B1 to B5, G1, G2, G4, G5 are updated to the requested actions A1 to A4, B1 to B5, G1, G2, G4, G5. As a result, the DAO agent 61 learns the optimum actions A1 to A4, B1 to B5, G1, G2, G4, and G5 for raising the total price TB in the floating exchange rate of tokens. This learning goal is just an example, and other learning goals include increasing the number of posts of origin ideas, increasing the total number of posts of origin ideas and improvement plans, and increasing the number of commercializations. That may be to increase the total commercialization revenue.

Next, in S645, it is determined whether or not the reinforcement learning is completed, and if it is not completed yet, it returns. When it is determined that the completion is completed, the learned multi-service DAO is transmitted to the requester of the simulation reinforcement learning in S646.

The client of simulation reinforcement learning can operate a trained multi-service DAO (innovation-induced DAO) 65 with an optimized incentive design in the real world 47. As a result, in this multi-service DAO (innovation-induced DAO) 65, "each persona agent group 67 to 71" shown in FIG. 37 becomes an actual user group, and the trained DAO agent is used for the user group performing each service. The optimally designed reward (incentive) is distributed by 61. At the stage of actual operation in this real world 47, services such as the contents of each post and the contents of token sales transactions are recorded on the blockchain with a time stamp. As a result, the blockchain acts as a notary for the content of origin idea submissions and the content of improvement proposals, and the exception application of loss of novelty (Patent Law Article 30) and measures against denial application (Patent Law Article 49, Paragraph 1, Line 7). , Article 74, Article 123, Paragraph 1, Item 2) will also be easier to perform.

In addition, even at the stage of operating the multi-service DAO (innovation-induced DAO) 65 in the real world 47, the DAO agent 61 will continue machine learning (reinforcement learning), and an even more optimal incentive that matches the operation situation of one's own. It may be designed. In addition, each trained persona agent group 67 to 71 (learned persona agent group according to FIGS. 40 (A) (B) and 41 (A) (B)) is also included in the multi-service DAO (innovation-induced DAO) 65. It may be transmitted to the requester of the simulation reinforcement learning, and each persona agent group 67 to 71 may function as a consultant for the user group performing each service. Furthermore, at the stage of operating the multi-service DAO (innovation-induced DAO) 65 in the real world 47, even if each service is a mixed type multi-service DAO 65 executed by both the user group and each persona agent group 67 to 71. In addition, it may be a persona agent operation type multi-service DAO (innovation-induced DAO) 65 in which each service is executed only by each persona agent group 67 to 71. It should be noted that this innovation-induced DAO is not limited to the one generated through the simulation reinforcement learning in the mirror world described above, but is artificially generated based on another method, for example, an artificial design. Further, it is not limited to DAO, and may be an organization having a specific manager or entity (for example, an ordinary corporation).

Next, the main routine of the process in which the persona agent performs reinforcement learning will be described based on FIG. 39 (B). In S648, the idea idea service execution process is performed, in S649, the improved service execution process is performed, in S650, the commercialization service execution process is performed, in S651, the infringement countermeasure service execution process is performed, and in S652, the infringement countermeasure service execution process is performed. Token purchase service execution process is performed.

The details of the idea-initiating service execution process shown in S648 will be described with reference to FIG. 40 (A). In S655, it is determined whether or not to propose an idea, and if not, a return is made. If it is determined to be performed, the process of creating an idea is performed in S656. The creation of this idea uses, for example, an AI called DABUS. For example, Persona Agent 67 and DABUS collaborate to create ideas. In S657, the posted content of the idea idea is generated, and in S658, the idea idea posting act a1i is executed.

In S659, it is determined whether or not the reward r1i has been received from the DAO agent 61, and if not, the reward r1i is returned. If it is determined that the signal has been received, in S660, the action a according ^{to the optimum policy π * is obtained by TD learning based on the reward r1i.} In this act a, if the received reward r1i is satisfactory, the act of inventing an idea is repeatedly continued, but if the reward r1i is not satisfactory, another act (for example, improvement service, commercialization) is continued. You will have to select services, infringement handling services, token purchasing services, or no services).

The details of the improved service execution process shown in S649 will be described with reference to FIG. 40 (B). In S664, it is determined whether or not to post the improvement plan, and if it is not posted, a return is made. If it is determined to post, the process of creating an improvement plan is performed in S665. For the creation of improvement plans, for example, AI called DABUS is used. For example, Persona Agent 68 and DABUS collaborate to create an improvement plan. In S666, the content of the improvement proposal posting is generated, and in S667, the improvement proposal posting act a2i is executed.

In S668, it is determined whether or not the reward r2i has been received from the DAO agent 61, and if not, the reward r2i is returned. If it is determined that the message has been received, in S669, the action a according ^{to the optimum policy π * is obtained by TD learning based on the reward r2i.} If the reward r2i received is satisfactory, this act a will continue the act of posting the improvement proposal repeatedly, but if the reward r2i is not satisfactory, another act (for example, an idea idea service, etc.) You will have to select commercialization services, infringement handling services, token purchasing services, or no services).

The details of the commercialization service execution process shown in S650 will be described with reference to FIG. 41 (A). In S674, it is determined whether or not to commercialize, and if it is not commercialized, a return is made. If it is determined to be commercialized, a business plan is generated in S675, the business plan posting act a3i is executed in S676, the commercialization service is executed in S677, and the execution status posting act a3i is executed in S678. To execute. This performance status posting act a3i also includes posting of profits obtained from the above-mentioned commercialization.

In S679, it is determined whether or not the reward r3i has been received from the DAO agent 61, and if not, the reward r3i is returned. If it is determined that the message has been received, in S680, the action a according ^{to the optimum policy π * is obtained by TD learning based on the received reward r3i.} If the reward r3i received is satisfactory, this act a will continue the act of the commercialization service repeatedly, but if the reward r3i is not satisfactory, another act (for example, the idea idea service, etc.) You will have to select improved services, infringement handling services, token purchasing services, or no services).

The details of the infringement countermeasure service execution process shown in S651 will be described with reference to FIG. 41 (B). In S684, it is determined whether or not to execute the infringement countermeasure service. If it is not executed, it returns, but if it is determined to be executed, the infringing act is investigated in S685, and it is determined in S686 whether or not the infringing act is found. Before investigating the infringing act, as described above, the patent application and the act of acquiring the right may be filed. If the infringing act is not found, it returns, but if it is determined that the infringing act is found, a warning letter to the suspected infringer is generated in S687, and the warning letter posting act a4i is executed in S688. Then, in S689, the infringement coping act a4i such as negotiation with the suspected infringer is executed, and in S690, the execution status posting act a4i is executed.

Next, in S691, it is determined whether or not the reward r4i has been received from the DAO agent 61, and if not, the reward r4i is returned. If it is determined that the message has been received, in S692, the action a according ^{to the optimum policy π * is obtained by TD learning based on the received reward r4i.} If the reward r4i received is satisfactory, this act a will continue the act of the commercialization service repeatedly, but if the reward r4i is not satisfactory, another act (for example, the idea idea service, etc.) You will have to select improvement services, commercialization services, token purchase services, or no services).

Next, the details of the token purchase service execution process shown in S652 will be described with reference to FIG. 42 (A). In S969, it is determined whether or not to purchase the token, and if it is not purchased, a return is made. If it is determined to purchase, the token purchase act a5i is executed in S697. Next, in S698, it is determined whether or not the reward r5i has been received from the DAO agent 61, and if not, the reward r5i is returned. If it is determined that the message has been received, in S699, the action a according ^{to the optimum policy π * is obtained by TD learning based on the received reward r5i.} If the reward r5i received is satisfactory, this act a will continue the act of the commercialization service repeatedly, but if the reward r5i is not satisfactory, another act (for example, the idea idea service, etc.) Improvement services, commercialization services, infringement handling services, or no services) will be selected.

Based on FIG. 42 (B), the price fluctuation of the token 72 in the floating exchange rate due to the token purchase by the persona agent group 71 will be described. 50 tokens (market capitalization 50,000 yen) 72 are given to the persona agent 67 who performed the idea idea service as a reward A1, and a part of the token 72 (10 tokens) is paid a virtual currency equivalent to 10,000 yen. Purchased by Persona Agent 71a. Next, the persona agent 71b purchased the 10 tokens by paying a virtual currency equivalent to 15,000 yen. As a result, the value of 10 tokens will rise to 15,000 yen. Persona Agent 71c purchased it by paying a virtual currency equivalent to 20,000 yen. As a result, the value of 10 tokens will rise to 20,000 yen. Persona Agent 71d purchased it by paying a virtual currency equivalent to 25,000 yen. As a result, the value of 10 tokens will rise to 25,000 yen. Persona Agent 71e purchased it by paying a virtual currency equivalent to 30,000 yen. As a result, the value of 10 tokens will rise to 30,000 yen.

As a result, 40 tokens (market capitalization of 40,000 yen) on hand of Persona Agent 67 will soar to a market capitalization of 120,000 yen. This token soars in direct proportion to expectations, the more popular origin ideas are, the more popular (likes) improvements are posted, and the more popular (likes!) Are posted. The more commercializations are posted, the higher the price, and the more popular (many likes) infringement countermeasures are posted, the higher the price.

At the stage of actually operating the multi-service DAO65 in the real world 47, as described above, “each persona agent group 67 to 71” in FIG. 37 becomes a user group in the real world. In that case, not only the token 72 given to the group of users who have performed the idea-creating service, but also the tokens of the users themselves who perform various services (hereinafter referred to as "my token") may be bought and sold. When another user who browses the posted service content purchases the poster's own My Token in anticipation of the poster, the price of the My Token in the floating exchange rate rises. In this case, a part of the poster's income may be distributed to the My Token purchaser at a rate according to the purchase amount. This My Token may be issued within the Multi-Service DAO65, but the My Token issued by the specialist who issues and distributes the My Token is linked to the My Token issued to the user, and the My Token issued by the Specialist is linked to the Multi-Service DAO65. The user may be able to buy and sell. Currently, VALU Inc. is a specialist company that issues and distributes My Tokens.

Next, a system in which a plurality of functional elements work together to construct one well-controlled DAO will be described with reference to FIGS. 43 to 59. Such DAO is hereinafter referred to as "element integrated DAO".

With reference to FIG. 43, this element integrated DAO makes it possible to easily construct a company organization or the like that already exists in the real world 47 with DAO, and an element DAO is generated and prepared in advance for each functional element. There is. That is, a modularized element DAO is prepared for each functional element, and it is configured so that a desired element integrated DAO can be easily constructed by selecting and combining necessary element DAOs. The element DAO provider 73 is provided with a server 74 and an element DAO protocol DB 75. The element DAO protocol DB75 includes, for example, an element DAO prepared for each functional element required for a company relationship, an element DAO prepared for each functional element required for an NPO (Nonprofit Organization) relationship, and a functional element required for a local government. The element DAO and the like prepared for each are stored.

The element integrated DAO builder receives an order for element integrated DAO construction from the client, and installs the element DAO corresponding to the required functional element on the PC terminal 76 via the server 74. In the example of FIG. 43, A1 element DAO (including A1 element agent), A2 element DAO (including A2 element agent), A5 element DAO (including A5 element agent), and A9 element DAO (including A9 element agent) are installed. .. Each element agent of A1 to A9 is an AI for reinforcement learning (machine learning) so that each corresponding element DAO can exhibit the best performance.

In addition, a general agent is also installed in the PC terminal 76. This supervising agent supervises the element agents of each element DAO and controls so that the entire element integrated DAO is optimized, and the supervising agent itself also performs reinforcement learning (machine learning) to achieve overall optimization. do. Since each element agent is for maximizing the performance of the element DAO in charge, there is a possibility that the element agent alone may fall into partial optimization and the overall optimization of the element integrated DAO cannot be achieved. Therefore, a central agent that controls the entire element integrated DAO to be optimized is required. This can be said to be the same as searching for a Pareto optimal solution in a perfect information game, for example.

The element integrated DAO installed on the PC 67 is installed on a plurality of terminals 16 in order to perform simulation reinforcement learning using the mirror world 51 as a simulation environment, and the block chain digital consisting of a private chain 2 having the terminals 16 as nodes 19 is digital. Twin 2T is generated in the mirror world 51.

The main routine of simulation reinforcement learning of this element integrated DAO will be described with reference to FIG. 44 (A). The CPU 10 of the client user terminal 16 requesting the simulation reinforcement learning of the element integrated DAO performs the simulation reinforcement learning preparation response processing in S674 and the simulation reinforcement learning response processing in S675. The CPU 10 of the mirror world server 46 performs a simulation reinforcement learning preparation process in S679 and a simulation reinforcement learning process in S680.

The details of the simulation reinforcement learning preparation response processing shown in S674 and the simulation reinforcement learning preparation processing shown in S679 will be described with reference to FIG. 44 (B). In the simulation reinforcement learning preparation response processing, the CPU 10 of the user terminal 16 determines in S679 whether or not to request the simulation reinforcement learning, and returns if not requested. If it is determined that the request is made, the DAO data and the personal AI group are transmitted in S680 to make the request. DAO data is a functional element of an organization that wants simulation reinforcement learning. For example, in the case of a furniture assembly and sales company, it is a material procurement element, an assembly element, an advertising element, and a sales element. The personal AI group is a personal AI of people who are actually engaged in element integrated DAO in the real world 47. If there is a worker who does not have a personal AI, or if the worker has not been decided yet, simulation reinforcement learning will be performed as described based on S561, S565-S568, S562, and S573 to S575 described above. Generate and prepare a personal AI that matches the target element integration DAO.

In the simulation reinforcement learning preparation process, the CPU 10 of the mirror world server 46 determines in S683 whether or not there is a request for simulation reinforcement learning, and returns if there is no request. When it is determined that there is a request for simulation reinforcement learning, in S684, the personal AI group is copied and deployed in the element integrated DAO to generate the element integrated DAO digital twin, and it is registered in the mirror world 51 as a simulation target. ..

The state is shown in FIG. The digital twin 78 of the element integrated DAO77 of the real world 47 is registered in the mirror world 51. The element-integrated DAO digital twin 78 shown in FIG. 45 is, for example, an element-integrated DAO digital twin 78 of a furniture assembly and sales company, has each functional element of material procurement, assembly, promotion, and sales, and engages in each functional element. The identification numbers of the personal AI groups of those who do so are stored in the mirror world server 46.

For this element-integrated DAO digital twin 78, the optimum incentive design is derived by simulation reinforcement learning. A schematic system of the reinforcement learning (machine learning) is shown in FIG.

With reference to FIG. 46, in the element integrated DAO digital twin 78, the material procurement element agent 80, the personal AI group 84 in charge of material procurement, and the assembly element agent correspond to each functional element of material procurement, assembly, advertisement, and sales. 81 and personal AI group 85 in charge of assembly, advertising element agent 82 and personal AI group 86 in charge of advertising, material procurement element agent 80 and personal AI group 84 in charge of material procurement, sales element agent 83 and personal AI group 87 in charge of sales Is formed. The controlling agent 79 controls each of these element agents 80 to 83.

The personal AI group 84 in charge of material procurement performs actions a11, a12, ... a1n such as proposals at an internal meeting, and finally executes the set actions a1 to the digital twin group 88 of the material provider. The state S1 of the digital twin group 88 of the material provider for the act a1 is input to the material procurement element agent 80 and the personal AI group 84 in charge of material procurement. This state S1 is, for example, the number of material requests, the number of response materials for the act of price negotiation a1, the response price, and the like. In addition, each act a11, a12, ... a1n of the personal AI group 84 in charge of material procurement is also input to the material procurement element agent 80, and the collective action a1 is also input to the general agent 79 and the material procurement element agent 80. Will be done.

The material procurement element agent 80 calculates the performance p1 by the personal AI group 84 in charge of material procurement based on the act a1 and the state S1, and transmits the performance p1 to the controlling agent 79. The general agent 79 determines the reward r1 based on the performance p1, and transmits the reward r1 to the material procurement element agent 80. The material procurement element agent 80 determines the reward distribution rate based on each action a11, a12, ... a1n of the personal AI group 84 in charge of material procurement, and sets the reward r1 to the personal in charge of each material procurement according to the reward distribution rate. Distribute to AI84.

The personal AI group 85 in charge of assembly performs actions a21, a22, ... a2i such as proposals at an internal meeting, and finally executes the set actions a2 on the digital twin group 89 of the assembly facility. The state S2 of the digital twin group 89 of the assembly equipment with respect to the act a2 is input to the assembly element agent 81 and the personal AI group 85 in charge of assembly. This state S1 is, for example, the power consumption of the assembly equipment digital twin group 89 and the total working hours of the personal AI group 85 in charge of assembly engaged in the assembly equipment digital twin group 89. The actions a21, a22, ... a2i of the personal AI group 84 in charge of assembly are also input to the assembly element agent 81, and the combined actions a2 are also input to the control agent 79 and the assembly element agent 81.

The assembly element agent 81 calculates the performance p2 by the personal AI group 85 in charge of assembly based on the act a2 and the state S2, and transmits the performance p2 to the control agent 79. The controlling agent 79 determines the reward r2 based on the performance p2, and transmits the reward r2 to the assembly element agent 81. The assembly element agent 81 determines the reward distribution rate based on each action a21, a22, ... a2i of the personal AI group 85 in charge of assembly, and distributes the reward r2 to the personal AI 85 in charge of assembly according to the reward distribution rate. ..

The personal AI group 86 in charge of advertising performs actions such as proposals a51, a52, ... a5j at an internal meeting, and finally executes a set action a5 on the consumer's personal AI group 90. The state S5 of the consumer's personal AI group 90 with respect to the act a5 is input to the promotion element agent 82 and the personal AI group 86 in charge of promotion. This state S5 is, for example, whether or not the consumer has purchased the product, the purchase price, and the like for the product recommendation act a5 to the consumer. The actions a51, a52, ... a5j of the personal AI group 86 in charge of promotion are also input to the promotion element agent 82, and the combined actions a5 are also input to the general agent 79 and the promotion element agent 82.

The promotion element agent 82 calculates the performance p5 by the personal AI group 86 in charge of promotion based on the act a5 and the state S5, and transmits the performance p5 to the controlling agent 79. The general agent 79 determines the reward r5 based on the performance p5, and transmits the reward r5 to the promotion element agent 82. The advertising element agent 82 determines the reward distribution rate based on each act a51, a52, ... a5j of the personal AI group 86 in charge of advertising, and distributes the reward r5 to the personal AI 86 in charge of advertising according to the reward distribution rate. ..

The personal AI group 87 in charge of sales performs actions such as proposals a91, a92, ... a9m at an internal meeting, and finally executes a set action a9 for the store and the consumer's digital twin group 91. The state S9 of the store and consumer digital twin group 91 with respect to the act a9 is input to the sales element agent 83 and the personal AI group 87 in charge of sales. This state S9 is, for example, the total sales amount at the store. The state S5 of the digital twin 88 of the material provider for the above-mentioned act a5 is also input to the sales element agent 83. The actions a91, a92, ... a9m of the personal AI group 87 in charge of sales are also input to the sales element agent 83, and the collective actions a9 are also input to the general agent 79 and the sales element agent 83.

The sales element agent 83 calculates the performance p9 by the personal AI group 87 in charge of sales based on the act a9 and the states S5 and S9, and transmits the performance p9 to the controlling agent 79. The general agent 79 determines the reward r9 based on the performance p9, and transmits the reward r9 to the sales element agent 83. The sales element agent 83 determines the reward distribution rate based on each action a91, a92, ... a9m of the personal AI group 87 in charge of sales, and distributes the reward r9 to the personal AI 87 in charge of sales according to the reward distribution rate. ..

The calculation method of each of the performances p1 to p9 and each reward distribution rate will be described with reference to FIGS. 47 (A) (B) (C) and 48 (A). The material procurement element agent 80 stores the calculation algorithm of the performance p1 and the distribution rate as knowledge. The calculation algorithm of the performance p1 and the distribution rate will be described with reference to FIG. 47 (A). The material procurement element agent 80 sets the material purchase amount u and the current inventory quantity z of this time (between the previous YES time point and the present YES time point by S689) as the state S1, and the performance p1 = {2 (average purchase amount). Calculate performance p1 using the formula / u) + (z / average stock quantity)} / 3. The average purchase price is the average purchase price of materials from the start of simulation reinforcement learning to the present. The average inventory quantity is the average inventory quantity of materials from the start of simulation reinforcement learning to the present. As a result of this calculation formula, the performance p1 increases as the material purchase price u this time decreases, and the performance p1 decreases as the current inventory quantity z increases.

In addition, when p1 ≧ 1, the reward distribution rate is calculated in proportion to the degree of approval for the collective act a1, and conversely, when p1 <1, it is inversely proportional to the degree of approval for the collective act a1. calculate. This includes not only those proportional to or inversely proportional to the first power of the "degree of approval" but also those proportional or inversely proportional to the nth power of the "degree of approval", and the optimum proportional function or inverse proportional function is used in the material procurement element agent 80. Is obtained by performing reinforcement learning (machine learning). In addition, as for the "degree of approval", the personal AI that proposed the collective action a1 itself has the highest degree of approval, and the material procurement element agent 80 is each personal based on each action a11, a12, ... a1n of the personal AI. Judge (calculate) the degree of approval of AI.

The calculation algorithm of the performance p2 and the distribution rate stored as knowledge by the assembly element agent 81 will be described with reference to FIG. 47 (B). The assembly element agent 81 sets the power consumption e of the assembly facility this time (between the previous YES time point and the YES time point of this time by S710) and the total working time t of the current assembly worker as the state S2, and the performance p2 = { Calculate performance p2 using the formula (average power consumption / e) + (average total working hours / t)} / 2. The total working hours t of the assembly workers this time is the total working hours of the personal AI group 85 in charge of assembly engaged in the assembly equipment digital twin group 89 this time. The average power consumption is the average power consumption of the assembly equipment from the start of simulation reinforcement learning to the present. The average total working hours is the average total working hours of assembly workers from the start of simulation reinforcement learning to the present. As a result of this calculation formula, the performance p2 increases as the current power consumption e of the assembly equipment decreases, and the performance p2 decreases as the total labor time t of the assembly workers increases.

In addition, when p2 ≧ 1, the reward distribution rate is calculated in proportion to the degree of approval for the collective act a2, and conversely, when p2 <1, it is inversely proportional to the degree of approval for the collective act a2. calculate. This includes not only those proportional to or inversely proportional to the first power of the "degree of approval" but also those proportional or inversely proportional to the nth power of the "degree of approval", and the assembly element agent 81 assembles the optimum proportional function or inverse proportional function. Obtained by performing reinforcement learning (machine learning). As for the "degree of approval", the personal AI that proposed the collective action a2 itself has the highest degree of approval, and the assembly element agent 81 has each personal AI based on each action a21, a22, ... a2i of the personal AI. Judge (calculate) the degree of approval of.

The calculation algorithm of the performance p5 and the distribution rate stored as knowledge by the promotion element agent 82 will be described with reference to FIG. 47 (C). The advertising element agent 82 sets the total purchase amount k of the recommended consumer personal AI this time (between the previous YES time and the YES time of this time by S728) as the state S5, and the performance p5 = k / recommended consumer personal. Calculate performance p5 using the formula for the average total purchase price K of AI. The average total purchase amount K is the average of the total purchase amount of the recommended consumer personal AI group 90 from the start of simulation reinforcement learning to the present. As a result of this formula, the performance p5 will increase if the total purchase amount k of the recommended consumer personal AI this time is higher.

In addition, when p5 ≧ 1, the reward distribution rate is calculated in proportion to the degree of approval for the collective act a5, and conversely, when p5 <1, it is inversely proportional to the degree of approval for the collective act a5. calculate. This includes not only those proportional to or inversely proportional to the first power of the "degree of approval" but also those proportional or inversely proportional to the nth power of the "degree of approval", and the advertising element agent 82 advertises the optimum proportional function or inverse proportional function. Obtained by performing reinforcement learning (machine learning). As for the "degree of approval", the personal AI that proposed the collective act a5 itself has the highest degree of approval, and the advertising element agent 82 has each personal AI based on each act a51, a52, ... a5j of the personal AI. Judge (calculate) the degree of approval of.

The calculation algorithm of the performance p5 and the distribution rate stored as knowledge by the sales element agent 83 will be described with reference to FIG. 48 (A). The sales element agent 83 sets the total sales amount h and the average total sales amount H of this time (between the previous YES point and the YES point of this time by S749) at the store as the state S9, and sets the state S9 and the above state S5. Therefore, the performance p9 is calculated by the formula of performance p9 = (h−k) / (HK). The average total sales amount H is the average of the total sales amount at the store from the start of the simulation reinforcement learning to the present. Further, k is the total purchase amount of the recommended consumer personal AI this time, and K is the average of the total purchase amount of the recommended consumer personal AI group 90 from the start of simulation reinforcement learning to the present (Fig.). See 47 (C) and its description). As a result of this calculation formula, the performance p9 becomes larger as the value obtained by subtracting the current total purchase amount k of the recommended consumer personal AI from the current total sales amount h at the store becomes larger. The total purchase amount k of the recommended consumer personal AI this time is the credit of the personal AI group 86 in charge of advertising, and the credit of only the personal AI group 87 in charge of sales is recommended from the total sales amount h of this time at the store. This is because the value is obtained by subtracting the total purchase amount k of the consumer personal AI this time.

In addition, when p9 ≧ 1, the reward distribution rate is calculated in proportion to the degree of approval for the collective act a9, and conversely, when p9 <1, it is inversely proportional to the degree of approval for the collective act a9. calculate. This includes not only those proportional to or inversely proportional to the first power of the "degree of approval" but also those proportional or inversely proportional to the nth power of the "degree of approval", and the advertising element agent 83 advertises the optimum proportional function or inverse proportional function. Obtained by performing reinforcement learning (machine learning). As for the "degree of approval", the personal AI that proposed the collective action a9 itself has the highest degree of approval, and the sales element agent 83 has each personal AI based on each action a91, a92, ... a9m of the personal AI. Judge (calculate) the degree of approval of.

Next, the reward table 92 stored as knowledge by the controlling agent 79 will be described with reference to FIG. 48 (B). In the reward table 92, the calculation formula of the reward distributed by the controlling agent 79 to the element agents 80 to 83 is stored. The reward to be distributed is calculated by multiplying the coefficient x (profit for the current term) x (performance sent from the target element agent) / (total performance sent from all element agents). Here, the "current term" is the period from the previous YES time point to the current YES time point according to S675.

Specifically, the reward r1 to be distributed to the material procurement element agent 80 is r1 = A1, Lt, p1 / (p1 + p2 + p5 + p9). The reward r2 distributed to the assembly element agent 81 is r2 = A2 ・ Lt ・ p2 / (p1 + p2 + p5 + p9). The reward r5 to be distributed to the advertising element agent 82 is r5 = A5 ・ Lt ・ p5 / (p1 + p2 + p5 + p9). The reward r9 distributed to the sales element agent 83 is r9 = A9 ・ Lt ・ p9 / (p1 + p2 + p5 + p9). Here, Lt is the profit for this term, and A1, A2, A5, A9 are the coefficients as the actions decided by the controlling agent 79.

Next, the specific contents of the simulation reinforcement learning process shown in S680 will be described with reference to FIG. 49. S687 executes the integrated agent reinforcement learning process, S688 executes the material procurement element agent reinforcement learning process, S689 executes the assembly element agent reinforcement learning process, S690 executes the advertising element agent reinforcement learning process, and S691 executes the advertising element agent reinforcement learning process. Sales agent reinforcement learning process is executed, material procurement person personal AI reinforcement learning process is executed by S692, assembly person personal AI reinforcement learning process is executed by S693, advertising person personal AI reinforcement learning process is executed by S694, and S695. Executes the personal AI reinforcement learning process in charge of sales.

The details of the integrated agent reinforcement learning process shown in S687 will be described with reference to FIG. In S699, the general agent 79 determines whether or not each performance p sent from each element agent 80 to 83 has been received. If it is not received, the control proceeds to S671, but if it is determined that it is being received, each received performance p is stored in S670.

Next, in S671, it is determined whether or not each of the actions a1 to a9 has been received, and if not, the control proceeds to S673. If it is determined that the signal has been received, the received actions a1 to a9 are stored in S672. Next, in S673, it is determined whether or not there is an input of the state S9 sent from the personal AI group 91 of the store and the consumer, and if not, the control proceeds to S675. If it is determined that there is an input, sales = ΣS9 is calculated in S674.

Next, in S675, it is determined whether or not it is time to calculate the reward, and if it is not, the control proceeds to S677, but if it is determined, in S676, the reward table 92 is referred to. The rewards r1, r2, r3, r5, and r9 are calculated and transmitted to the corresponding element agents 80 to 83.

Next, in S677, it is determined whether or not it is time to update the reinforcement learning (machine learning), and if not, a return is made. If it is determined that the renewal time has come, the profit Lt = sales-expense for the current period is calculated in S687. Next, in S679, each reward r1, r2, r5, r9 is calculated and distributed to the corresponding element agents 80 to 83.

Next, in S680, the reward R of the controlling agent 79 is calculated from the profit Lt. This reward R is proportional to the profit Lt. Next, in S681, based on the above reward R, the actions (coefficients) A1, A2, A5, A9 according ^{to the optimum policy π * are obtained by TD learning.} Next, in S682, A1, A2, A5, A9 of the reward table 92 are updated to the actions (coefficients) A1, A2, A5, A9 obtained in S681. As a result, the supervising agent 79 learns the actions (coefficients) A1, A2, A5, and A9 that maximize the profit Lt.

The details of the material procurement element agent reinforcement learning process shown in S688 will be described with reference to FIG. 51. The material procurement element agent 80 performs information collection processing by the crawler in S684. A crawler is a program that periodically acquires documents and images on the Web and automatically creates a database. Also called "bot", "spider", "robot", etc.

The details of the information collection process by the crawler will be described with reference to FIG. 52 (A). The material procurement element agent 80 receives the information collected by the crawler patrolling the net in S702. Next, in S703, the received information is stored in the material procurement DB 93.

The information stored in the material procurement DB 93 is shown in FIG. 52 (B). As shown in the figure, the material procurement DB 93 stores various information necessary for material procurement operations, such as economic information, social information, weather information, inventory information, market information, and so on. ..

Returning to FIG. 51, the material procurement element agent 80 determines whether or not the actions a11, a12, ... a1n from the personal AI group 84 have been received in S685. If it is not received, the control proceeds to S687, but if it is determined that it has been received, the received actions a11, a12, ... a1n are stored in S686. In S687, it is determined whether or not the state S1 from the digital twin group 88 of the material supplier has been received, and if not, the control proceeds to S689. If it is determined that the signal has been received, the received S1 is stored in S688.

In S689, it is determined whether or not it is the calculation time of the performance p1, and if it is not the calculation time, the control proceeds to S692. If it is determined that it is the calculation time, the performance p1 = {2 (average purchase amount / u) + (z / average inventory quantity)} / 3 is calculated in S690. The performance p1 is transmitted to the controlling agent 79 (S691).

In S692, it is determined whether or not the reward r1 transmitted from the control agent 79 has been received, and if not, the reward r1 is returned. If it is determined that the reward has been received, the reward distribution rate is calculated based on the reward distribution rate algorithm shown in FIG. 47 (A) (S693). In S694, each reward r11, r12 ... r1n is calculated by multiplying the reward by each distribution rate, and in S695, each reward r11, r12 ... r1n is given to each material procurement personal AI group 84. In S696, based on the received reward r1, the action (proportional function or inverse proportional function) according ^{to the optimum policy π * is obtained by TD learning.} In S697, the proportional function or the inverse proportional function is updated to the one obtained in S696. As a result, the material procurement element agent 80 learns the proportional function or the inverse proportional function that maximizes the performance p1.

Next, the details of the assembly element agent reinforcement learning process shown in S689 will be described with reference to FIG. 53. The assembly element agent 81 determines in S706 whether or not the actions a21, a22, ... a2n from the personal AI group 85 have been received. If it is not received, the control proceeds to S708, but if it is determined that it has been received, each of the received actions a21, a22, ... a2n is stored in S707. In S708, it is determined whether or not the state S2 from the digital twin group 89 of the assembly equipment has been received, and if not, the control proceeds to S710. When it is determined that the signal has been received, the received state S2 is stored in S709.

In S710, it is determined whether or not it is the calculation time of the performance p2, and if it is not the calculation time, the control proceeds to S713. If it is determined that it is the calculation time, the performance p2 = {(average power consumption / e) + (average total working hours / t)} / 2 is calculated in S711. The performance p2 is transmitted to the controlling agent 79 (S712).

In S713, it is determined whether or not the reward r2 sent from the controlling agent 79 has been received, and if not, the reward r2 is returned. If it is determined that the reward has been received, the reward distribution rate is calculated based on the reward distribution rate algorithm shown in FIG. 47 (B) (S714). In S715, each reward r11, r12 ... r1n is calculated by multiplying the reward by each distribution rate, and in S695, each reward r21, r22 ... r2i is assembled and given to the personal AI group 86 in charge. In S717, based on the received reward r2, the action (proportional function or inverse proportional function) according ^{to the optimum policy π * is obtained by TD learning.} In S718, the proportional function or the inverse proportional function is updated to the one obtained in S717. As a result, the assembly element agent 81 learns a proportional function or an inverse proportional function that maximizes the performance p2.

The details of the promotion element agent reinforcement learning process shown in S690 will be described with reference to FIG. 54. The promotion element agent 83 performs information collection processing by the crawler in S723. The details of this process will be described with reference to FIG. 55 (A). The promotion element agent 83 receives the information collected by the crawler patrolling the net in S740, and stores the received information in the promotion DB 94 in S741.

The collected data stored in the promotion DB 94 is shown in FIG. 55 (B). Various behavior data of consumers such as Taro, Jiro, ... Hanako, etc. are stored in the promotion DB 94. For example, in the case of Taro, it is judged that there is a high possibility of purchasing furniture from the information that "I ordered a detached house", and the furniture is advertised to Taro. In the case of Jiro, it is judged that there is a high possibility that he will buy furniture in his new house to get married soon, based on the information that he will buy a couple, and he will advertise the furniture to Jiro.

Returning to FIG. 54, the promotion element agent 83 determines whether or not the action from each personal AI group 86 has been received in S742. If it is not received, the control proceeds to S726, but if it is determined that it is being received, each act received is stored in S725. In S726, it is determined whether or not the state S5 sent from the consumer's personal AI group 90 has been received, and if not yet received, the control proceeds to S728. If it is determined that the signal has been received, the received state S5 is stored in S727.

In S728, it is determined whether or not it is the time to calculate the performance p5, and if it is not the time to calculate the performance p5, the control proceeds to S731. If it is determined that it is time to calculate the performance p5, the average total purchase amount K of the performance p5 = k / recommended consumer personal AI is calculated in S729. Next, in S730, the performance P5 is transmitted to the controlling agent 79.

In S731, it is determined whether or not the reward r5 sent from the controlling agent 79 has been received, and if it has not been received yet, the reward r5 is returned. When it is determined that the reward has been received, the reward distribution rate is calculated in S732 based on the reward distribution rate algorithm shown in FIG. 47 (C) (S732). In S733, each reward r51, r52 ... r5j is calculated by multiplying the reward by each distribution rate, and in S734, each reward r51, r52 ... r5j is given to the personal AI group 87 in charge of promotion. In S735, based on the received reward r5, the action (proportional function or inverse proportional function) according ^{to the optimum policy π * is obtained by TD learning.} In S736, the proportional function or the inverse proportional function is updated to the one obtained in S735. As a result, the advertising element agent 82 learns a proportional function or an inverse proportional function that maximizes the performance p5.

Next, the details of the sales element agent reinforcement learning process shown in S691 will be described with reference to FIG. 56. The sales element agent 84 performs information collection processing by the crawler in S744. The details of this process will be described with reference to FIG. 57 (A). The sales element agent 84 receives the information collected by the crawler patrolling the net in S760, and stores the received information in the sales DB 95 in S761. Further, in S762, the POS data in the store is stored in the sales DB 95.

The collected data stored in the sales DB 95 is shown in FIG. 57 (B). Various data such as weather data and POS data are stored in the sales DB 95. "By date" in meteorological information is a concept that includes each day of the week. Based on the weather information (hourly weather temperature data by date) and POS data (hourly sales product data by date), for example, it is possible to rearrange the displayed products in consideration of the day, time, and weather conditions. can.

Returning to FIG. 56, the sales element agent 84 determines in S745 whether or not the action from each personal AI group 87 has been received. If it is not received, the control proceeds to S747, but if it is determined that it is being received, each act received is stored in S746. In S747, it is determined whether or not the state S9 sent from the personal AI group 91 of the store and the consumer has been received, and if not yet received, the control proceeds to S747. If it is determined that the signal has been received, the received state S9 is stored in S748.

In S749, it is determined whether or not it is the calculation time of the performance p9, and if it is not the calculation time of the performance p9, the control proceeds to S752. If it is determined that it is time to calculate the performance p9, the performance p9 = (h-k) / (HK) is calculated in S750. Next, in S751, the performance p9 is transmitted to the controlling agent 79.

In S752, it is determined whether or not the reward r9 sent from the controlling agent 79 has been received, and if it has not been received yet, the reward r9 is returned. If it is determined that the reward has been received, the reward distribution rate is calculated in S753 based on the reward distribution rate algorithm shown in FIG. 48 (A) (S753). In S754, each reward r91, r92 ... r9m is calculated by multiplying the reward by each distribution rate, and in S755, each reward r91, r92 ... r9m is given to the personal AI group 88 in charge of sales. In S756, based on the received reward r9, the action (proportional function or inverse proportional function) according ^{to the optimum policy π * is obtained by TD learning.} In S757, the proportional function or the inverse proportional function is updated to the one obtained in S756. As a result, the sales element agent 83 learns the proportional function or the inverse proportional function that maximizes the performance p9.

Next, the details of the material procurement personal AI reinforcement learning process shown in S692 will be described with reference to FIG. 58 (A). In S765, the personal AI group 84 in charge of material procurement determines whether or not to negotiate with the digital twin group 88 of the material supplier, and if not, the control proceeds to S770. If it is determined to negotiate, in S766, the stored data of the fund procurement DB 93 is browsed, the action a1 is determined while conducting an internal meeting with reference to the stored data (S767), and negotiations are made with the digital twin group 88 of the material supplier. (S768). In S769, it is determined whether or not the negotiation has been completed, and if it has not been completed, the process returns to S766 and the loop of S767 → S768 → S769 → S766 is patrolled. When it is determined in S769 that the negotiation is completed, the control proceeds to S770.

In S770, it is determined whether or not the rewards r11, r12 ... r1n have been received from the material procurement element agent 80, and if not, the reward is returned. When it is determined that the reward has been received, in S771, the action (a11, a12 ... a1n) according ^{to the optimum policy π * is obtained by TD learning based on the received reward.} This act a1i also includes a move (change of job) to another company DAO Digital Twin (for example, DAO Digital Twin 59 of ABC Co., Ltd. in FIG. 45) if the received reward r1i is not satisfactory. As a result of this reinforcement learning, each of the personal AIs in charge of material procurement will learn the above-mentioned act of increasing the performance p1.

Next, the details of the assembly-in-charge personal AI reinforcement learning process shown in S693 will be described with reference to FIG. 58 (B). In S775, the personal AI group 85 in charge of assembly determines whether or not to have an internal meeting, and if not, the control proceeds to S779. If it is determined that a meeting is to be held, in S776, each personal AI in charge of assembly decides the action a2 while conducting an internal meeting. Next, in S777, the assembly equipment digital twin group 89 is commissioned according to the act a2, and the validity of the act a2 is verified. In S778, it is determined whether or not the meeting has been completed, and if it has not been completed, the process returns to S777 and the loop of S777 → S778 → S776 is patrolled. If the action a2 is appropriate as a result of the trial run of S777, it is determined by S778 that the meeting is completed, and the control proceeds to S779.

In S779, it is determined whether or not the rewards r21, r22 ... r2n have been received from the assembly element agent 81, and if not, the reward r21, r22 ... r2n is returned. When it is determined that the reward has been received, in S780, the action (a21, a22 ... a2i) according ^{to the optimum policy π * is obtained by TD learning based on the received reward.} This act a2i also includes a move (change of job) to another company DAO Digital Twin (for example, DAO Digital Twin 59 of ABC Co., Ltd. in FIG. 45) if the received reward r2i is not satisfactory. As a result of this reinforcement learning, each of the personal AIs in charge of assembly learns the above-mentioned act of increasing the performance p2.

Next, the details of the promotion personal AI reinforcement learning process shown in S694 will be described with reference to FIG. 59 (A). In S784, the personal AI group 86 in charge of promotion determines whether or not to have an internal meeting, and if not, the control proceeds to S789. If it is determined that a meeting is to be held, in S785, each personal AI in charge of advertising decides the act a5 while conducting an internal meeting. Next, in S787, the act a2 for the consumer is executed. In S788, it is determined whether or not the meeting has been completed, and if it has not been completed, the process returns to S785 and the loop of S786 → S787 → S788 is patrolled. Control proceeds to S789 when it is determined by S788 that the meeting is completed.

In S789, it is determined whether or not the rewards r51, r52 ... r5j have been received from the advertising element agent 82, and if not, the reward r51, r52 ... r5j is returned. When it is determined that the reward has been received, in S790, the action (a51, a52 ... a5j) according ^{to the optimum policy π * is obtained by TD learning based on the received reward.} This act a5i also includes a move (change of job) to another company DAO Digital Twin (for example, DAO Digital Twin 59 of ABC Co., Ltd. in FIG. 45) if the received reward r5i is not satisfactory. As a result of this reinforcement learning, each of the personal AIs in charge of advertising will learn the above-mentioned act of increasing the performance p5.

Next, the details of the sales person in charge personal AI reinforcement learning process shown in S695 will be described with reference to FIG. 59 (B). In S791, the sales person in charge of personal AI group 87 determines whether or not to have an internal meeting, and if not, the control proceeds to S795. If it is determined that a meeting is to be made, in S792, each sales person in charge of personal AI decides the act a9 while conducting an internal meeting. Next, it is determined in S793 whether or not the meeting has been completed, and if it has not been completed, the process returns to S792 and the loop of S792 → S793 → S792 is patrolled. Control proceeds to S794 when it is determined by S793 that the meeting is completed. In S794, the act determined in the above meeting is executed for the consumer and the store.

Next, in S795, it is determined whether or not the rewards r91, r92 ... r9m have been received from the sales element agent 83, and if not, the reward is returned. When it is determined that the reward has been received, in S796, the action (a91, a92 ... a9m) according to the optimum policy π * is obtained by TD learning based on the received reward. This act a9i also includes a move (change of job) to another company DAO Digital Twin (for example, DAO Digital Twin 59 of ABC Co., Ltd. in FIG. 45) if the received reward r9i is not satisfactory. As a result of this reinforcement learning, each of the sales person personal AIs learns the above-mentioned act of increasing the performance p9.

The element integrated DAO for which simulation reinforcement learning has been completed is operated as an actual organization in the real world 47. At that stage, an actual human (user) in the real world will be in charge of "each personal AI group 84 to 87" in FIG. 46. At that time, each personal AI group 84 to 87 that has undergone simulation reinforcement learning acts as a consultant for an actual human (user), and the knowledge, experience, and know-how gained through simulation reinforcement learning can be provided to the actual human (user). can.

The construction of the element integrated DAO described above shows that the entire organization such as a company, NPO, or local government is created by combining element DAOs by function, but it is not the entire organization but a part of the organization (for example, material procurement). ) May be constructed with the element DAO.

The above-mentioned programs running on the user terminal 16 and the like and various servers may be downloaded and installed from a predetermined website or the like, but for example, a recording medium such as a CD-ROM 99 (non-transitory recording). The program may be installed on the user terminal 16 and various servers by a person who has purchased the CD-ROM99 or the like by recording it on a medium) and distributing it (see FIG. 60).
[Modification example]

(1) For example, the names of Taro, Jiro, Sakura, Saburo, etc. in the digital twin data shown in FIG. 29 are identified by using pseudonyms (anonymous) from the viewpoint of personal information protection, although they can be identified as the same person. It may not be possible to identify the individual. In that case, the AI identification number or blockchain address may be used as a pseudonym (anonymous). Similarly, digital twins such as ABC Co., Ltd. also use a pseudonym (anonymous) for the company name (organization name) so that they can be identified as the same company (same organization) but not a specific company (organization). It may be. Further, as the human digital twin, a plurality of digital twins by a plurality of personal AIs may be prepared for one person. Further, one digital twin in one person may be composed of a set of a plurality of personal AIs (for example, a set of specialized personal AIs in various fields).

(2) In FIGS. 34 to 59, a system for deriving the optimum solution of incentive design in DAO by performing a simulation using a multi-service DAO having a plurality of types of services as an example has been described, but the system is not limited to the multi-service DAO. , The system may be a system for deriving the optimum solution of incentive design by simulation for DAO in which only one type of service exists.

(3) In FIG. 35, a trained persona agent group is generated for each persona, but each personal AI of the user group belonging to each persona is set to the existing personal AI group registered in the mirror world 51. You may choose from them and use it as a group of persona agents. In this case, it is necessary to inquire the user group belonging to each persona whether or not the personal AI may be used for the simulation, and obtain the consent that the personal AI may be used. Each personal AI of the user group with the consent is copied and used for the simulation, and the learned personal AI group after the simulation is completed is transmitted to each corresponding user. When each user who receives it determines that the learned personal AI is useful (necessary), it overwrites the learned personal AI with the existing personal AI and saves it. In addition, both the existing personal AI and the learned personal AI may be stored together and used properly as needed.

(4) As multi-agent reinforcement learning, the general agent (master agent) responsible for overall optimization distributes rewards to each agent, and the general agent itself also performs reinforcement learning to converge the reward distribution action to the optimum one. The master agent method is shown. However, multi-agent reinforcement learning is not limited to this, for example, D-learning that can converge to the optimum solution under the Markov decision process, or Bucket Brigade as a reinforcement learning algorithm in the Classifier System. Or Profit Sharing may be used.

(5) The simulation using the digital twin is not limited to the digital twin of a person or an organization composed of people (for example, a corporation, an NPO, etc.). For example, in an object such as an AI-equipped machine or an electric product (for example, an AI-equipped vacuum cleaner), a digital twin in an environment in which the object operates (for example, a user's house in which an AI-equipped vacuum cleaner that moves independently operates). Is generated in the cyber space, the AI mounted on the object is simulated in advance in the environment digital twin, reinforcement learning (machine learning) is performed, and the customized (personalized) trained AI-equipped object is applicable. It may be provided to the user who does.

Since the dilemma in which the guarantee of the authenticity of the recorded information and the guarantee of the right to delete the information conflict with each other can be solved as much as possible, it can be used for an information recording method having an indelible property such as a blockchain.

1 Internet 2 Private chain 3 Consortium chain 4 Public chain 12 HDD
16 User terminal 19 Node 30 Key registration center 32 Key DB
46 Mirror World Server 51 Mirror World 52 Earth Digital Twin 53 Japan Digital Twin 54 Town Digital Twin 57 Taro Digital Twin 58 Taro Family Digital Twin 59 ABC Digital Twin Co., Ltd. 61 DAO Agent 72 Token 78 DAO Digital Twin 79 General Agent.

Claims (5)

An encryption means that performs encryption processing to encrypt the information to be recorded,
A recording means for recording information after the encryption process, and
Decoding means that performs decoding processing on the information recorded by the recording means using the first key and the second key to obtain plain text information.
The information recorded by the recording means is provided with an undecryptable means for making the information undecryptable, which cannot be decoded.
The decryption means includes a second key concealment holding means for concealing and holding the second key.
The decryption-disabled means is a processing system that puts the second key held by the second key concealment-holding means into a decryption-disabled state by updating it with another key. The processing system according to claim 1, wherein the decryption means further includes a first key distribution means for distributing the first key to a person who wants to view information. The processing system according to claim 1 or 2, further comprising a search means for retrieving information recorded by the recording means without making it plain text. The information recorded by the recording means includes personal information and includes personal information.
The processing system according to any one of claims 1 to 3, wherein the undecryptable means puts the personal information of the personal information owner into the undecryptable state at the request of the personal information owner. Steps to perform encryption processing to encrypt the information to be recorded,
A decryption step of decrypting the information recorded by the recording means for recording the information after the encryption process using the first key and the second key to obtain plain text information.
The step of making the information recorded by the recording means undecipherable, which cannot be decoded,
Let the computer run
The decryption step includes a step of keeping the second key secret.
The step of making the decryption impossible state is a program that makes the decryption state by updating the second key held by the holding step to another one.

Images