Rights management system for learning data

Abstract

To provide a rights management system for learning data used for machine learning models such as generative AI.SOLUTION: A rights management system allocates a learning data ID to learning data to be used for generating a generative AI model for generating a new content or new data, and receives registration of rights information of learning data to which the learning data ID is allocated. Registration of the generative AI model that is generated using licensed learning data is received from a model developer, and model registration information associated with the learning data ID of the used learning data, is generated. License applications for the registered generative AI model are received from business operators, usage fees corresponding to the rights information of the learning data associated with the licensed generative AI model, are calculated for the business operators who obtain business revenues through a content or data generated by the registered generative AI model, and the usage fees deposited from the business operators are aggregated and a dividend is calculated for each learning data provider.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] Embodiments of the present invention relate to a right management technology for learning data used in machine learning models such as generative AI models.

Background Art

[0002] The generation of machine learning models such as AI models and the business using machine learning models are booming. In particular, in recent years, technologies for generating new content and new data using artificial intelligence called generative AI have attracted attention.

[0003] Generative AI can learn patterns and the like based on learning data and generate new content and new data. This generative AI is a machine learning model constructed using machine learning such as deep learning.

Prior Art Documents

Non-Patent Documents

[0004]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] An object is to provide a right management system for learning data.

Means for Solving the Problems

[0006] The rights management system of the embodiment manages the rights of learning data used in a generative AI model that generates new content or new data. The rights management system assigns a learning data ID to the learning data of the learning data provider, accepts the registration of the rights information of the learning data to which the learning data ID is assigned, and accepts a license application for the learning data from the model developer. A learning data management unit, accepts the registration of a generative AI model generated using the licensed learning data, generates model registration information associated with the learning data ID of the used learning data, and accepts a license application for the registered generative AI model from the business operator. A model management unit, and calculates a usage fee according to the rights information of the learning data associated with the licensed generative AI model for a business operator who obtains business revenue through the content or data generated by the generative AI model, and aggregates the usage fees received from the business operator. A usage fee management unit that calculates a distribution amount for each learning data provider.

Brief Description of the Drawings

[0007]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Best Mode for Carrying Out the Invention

[0008] Hereinafter, embodiments will be described with reference to the drawings.

[0009] A machine learning model is generated through a learning process using learning data. The machine learning model can be broadly divided into two categories.

[0010] The first machine learning model is an AI model that organizes and classifies information to perform predictions, inspections, etc., and examples include image recognition models and voice recognition models. The feature of this first machine learning model is to automatically perform determined actions and output appropriate answers such as prediction results and inspection results for input data based on the learned data.

[0011] On the other hand, the second machine learning model is a generative AI model that generates (creates) content and data, and there are various types such as text generation AI models, voice generation AI models, image generation AI models, and video generation AI models. The generative AI model is characterized by newly generating content such as text, voice, music, images, and videos.

[0012] And in recent years, with the emergence of generative AI, various opinions and issues have emerged regarding how to handle learning data. For example, a voice generation AI that has undergone a learning process using the voice data of voice actors (actors, narrators) who only appear by voice in voice works and video works can newly generate voice content that depends on the voice data of the voice actors. Regardless of the discussion of whether the content is created by humans or by generative AI, those who have rights to the learning data are not unrelated to the benefits obtained by the voice generation AI that uses their voice data as learning data.

[0013] Since the new content and new data generated by generative AI depend on the training data used to build the generative AI, the unregulated use of training data is not an appropriate situation from the perspective of protecting those who have rights to the training data. Therefore, there is an urgent need to create a platform that protects those who have rights to the training data, in other words, a platform that can effectively connect relevant parties when generating new content and businesses using generative AI.

[0014] The rights management system of the embodiment provides a rights management platform that centrally manages training data providers, model developers, and business operators who conduct business using generative AI models, and realizes a mechanism that can distribute profits to the training data contributed to the generative AI models that bring business benefits.

[0015] From another perspective, when a business operator uses a generative AI model with unknown origin training data, there may be risks. For example, when copyrighted works are used in the training data, permission may be required. Therefore, if it is clearly understood what training data was used to develop the generative AI model, risks can be avoided and the quality based on the training data and the value of the generative AI model can be promoted. Also, the model developer can clarify the origin of the training data to the business operator and promote it to the business operator who is the trading partner. The value of the development of the generative AI model and the business using the generative AI model can be improved through the rights management platform.

[0016] (First Embodiment) Figures 1 to 6 are diagrams for explaining the first embodiment.

[0017] FIG. 1 is a schematic explanatory diagram of a rights management system for learning data used in a machine learning model including a generative AI model. The rights management system includes a management device 100. The management device 100 provides a rights management platform used by learning data providers, model developers, and businesses.

[0018] In the following description, as an example of a machine learning model, a generative AI model will be described. The learning data is data used in a generative AI model, such as text data (articles), voice data, music data, image data, video data, etc. The rights management platform of the present embodiment performs rights management of the learning data used in the learning process (model generation process) of the generative AI model regardless of the type and content of the learning data.

[0019] A person who has rights to learning data is called a learning data provider. For example, when the learning data is voice data, the learning data provider is an individual who provides the voice or an organization such as a company to which the individual belongs. For example, individuals such as voice actors and actors, and companies to which voice actors and actors belong correspond to this.

[0020] A model developer is a developer of a machine learning model that performs a learning process using the learning data provided by the learning data provider and constructs and generates a generative AI model. The types and uses of the generative AI models registered in the rights management platform are not particularly limited, and any generative AI model using the learning data registered in the rights management platform may be used. Note that the model developer can develop not only a generative AI model but also a computer system including the generative AI model and provide it to a business. A model developer is an organization such as an individual or a company.

[0021] An operator is a person who conducts business by receiving the provision of a generative AI model from a model developer. For example, an operator purchases a generative AI model from a model developer, or purchases the right to use a computer system including the generative AI model provided by the model developer, provides the content and data generated by the generative AI model, or provides services that utilize the content and data generated by the generative AI model. In addition, an operator can enter into a transaction contract with a model developer and commission the development of a generative AI model. An operator is an organization such as an individual or a company that conducts business for profit by deploying a generative AI model or a computer system including the generative AI model to obtain profits.

[0022] As a preliminary preparation for using the rights management system (management device 100), the learning data provider, the model developer, and the operator perform user registration.

[0023] The learning data provider registers the learning data in the rights management system. The rights management system identifies the registered learning data. Specifically, a learning data ID is assigned to the learning data. Then, it accepts the registration of the rights information of the learning data to which the learning data ID has been assigned. By identifying the learning data and registering the rights information, the preparations for granting permission to use the learning data are completed.

[0024] Note that the learning data can be registered as a plurality of data groups, that is, a dataset. When a dataset composed of a plurality of learning data groups is registered, the learning data ID is assigned so that each of the dataset and the plurality of learning data included in the dataset can be identified. Note that the registration of the learning data does not have to be a group of datasets, and for example, it can be controlled to register one learning data.

[0025] Model developers can use functions such as searching for training data registered in the rights management system, viewing the rights information associated with the training data, and applying for a license to use. The rights management system accepts an application for a license to use the training data from the model developer and conducts license processing. The model developer can obtain the licensed training data from the rights management system.

[0026] The model developer uses the obtained training data to generate a generative AI model. The model developer registers the generative AI model generated using the training data in the rights management system. The rights management system accepts the registration of the generative AI model and generates model registration information associated with the training data ID of the used training data. That is, the generated generative AI model is associated with the used training data to identify the generative AI model. This prepares for the license to use the generative AI model.

[0027] Note that there are two cases for how the model developer obtains the training data: the first case is to download and obtain the training data from the management device 100 through the rights management system, and the second case is to obtain the training data without going through the rights management system. In the second case, for example, the model developer can directly obtain the training data from the training data provider.

[0028] In the case of the first scenario, since the training data is downloaded from the rights management system, the generative AI model is generated using licensed training data. However, in the case of the second scenario, as the training data is not provided via the rights management system, the generative AI model may be generated using training data that is not licensed on the rights management system. In this case, the model developer can later apply for a license for the training data used to generate the generative AI model. For example, when registering the generative AI model, the used training data can be searched and configured to be associated with the training data registered in the rights management system and the generative AI model. Note that even in the second scenario, before generating the generative AI model using the training data, the rights management system can be accessed to apply for a license for the corresponding training data. In this case, the generative AI model will be generated using the licensed training data.

[0029] The business operator conducts transactions individually with the model developer. The business operator obtains the desired generative AI model from the model developer or obtains the right to use the computer system including the generative AI model provided by the model developer. The business operator applies for a license to the rights management system for the generative AI model used in the business. The rights management system accepts the license application for the registered generative AI model from the business operator and associates the business operator with the generative AI model. As a result, the business using the generative AI model will be associated with the training data.

[0030] Here, the license for each of the training data and the generative AI model will be described. The license for using the training data is a process that associates the training data with the model developer. That is, the license for using the training data records the history of which training data the model developer has used. At this time, the form of the license is arbitrary. For example, it may be a mechanism that applies a known license agreement, or it may be a mechanism that authenticates (permits) the use of the training data registered by the model developer in the rights management system. Therefore, the application for the license to use the training data is an act in which the model developer requests to use (has used) the training data registered in the rights management system, and the license process is a process that associates the model developer who has requested to use the training data with the corresponding training data (training data provider).

[0031] Similarly, the license for the generative AI model is a process that associates the model developer with the business operator, and is a process for recording the history of which generative AI model the business operator uses to conduct business. The form of the license for the generative AI model is also arbitrary. For example, it may be a mechanism that applies a known license agreement. Therefore, the application for the license to use the generative AI model is an act in which the business operator requests to use (has used) the generative AI model registered in the rights management system, and the license process is a process that associates the business operator who has requested to use the generative AI model with the corresponding generative AI model.

[0032] Next, the rights management system calculates the usage fee that the business operator pays according to the rights information of the training data associated with the licensed generative AI model, and conducts a billing process for the business operator. At this time, the business operator registers (reports) business information (revenue-related information) including the business revenue related to the licensed generative AI model in the rights management system. The rights management system can calculate the usage fee to be billed to the business operator based on the declared business revenue. The business operator pays the billed usage fee. The rights management system aggregates the usage fees received from the business operator and calculates the distribution amount for each training data provider. The calculated distribution amount is deposited into the bank account of the training data provider.

[0033] In this way, the rights management system can provide a rights management platform that centrally manages learning data providers, model developers, and businesses, and can implement a mechanism for distributing profits for the learning data contributed to the generative AI model that brings business benefits.

[0034] Figure 2 is a diagram showing the system configuration and functional blocks of the rights management system of this embodiment. The rights management system includes a management device 100. The management device 100 includes a communication device 110, a control device 120, and a storage unit 130. The learning data provider, model developer, and business connect to the management device 100 from their respective user terminals through a network such as an IP network and use the rights management platform.

[0035] The control device 120 includes a user interface control unit 121, a user management unit 122, a learning data management unit 123, a model management unit 124, and a usage fee management unit 125.

[0036] The user interface control unit 121 is a processing unit that provides various screens for using the rights management system to each user terminal of the learning data provider, model developer, and business, and performs screen control according to the operations of the user terminals.

[0037] The user management unit 122 provides a user registration function through a user registration screen. The user management unit 122 includes a registration control unit 1221. The registration control unit 1221 stores the user information input through the learning data provider registration screen for the learning data provider in the storage device 130 separately for each learning data provider. The user information of the learning data provider includes a learning data provider ID, a login password, registrant information (name, address, contact information, email address, etc.), organization information (organization name, address, contact information, email address, etc.), a registration type (a type indicating a learning data provider), transfer destination information (usage fee transfer destination information, etc.).

[0038] Furthermore, the registration control unit 1221 stores the user information input through the model developer registration screen for each model developer in the storage device 130. The user information of the model developer includes a model developer ID, a login password, registrant information (name, address, contact information, email address, etc.), organization information (organization name, address, contact information, email address, etc.), a registration type (a type indicating a model developer), and the like.

[0039] In addition, the registration control unit 1221 stores the user information input through the business operator registration screen for each business operator in the storage device 130. The user information of the business operator includes a business operator ID, a login password, registrant information (name, address, contact information, email address, etc.), organization information (organization name, address, contact information, email address, etc.), a registration type (a type indicating a business operator), usage fee settlement information (credit card information, account transfer information, etc.), and the like.

[0040] Next, the learning data management unit 123 includes a learning data registration control unit 1231, a learning data rights information management unit 1232, a learning data identity management unit 1233, and a learning data use permission management unit 1234.

[0041] The learning data registration control unit 1231 assigns a learning data ID to the learning data uploaded to the management device 100 through the learning data registration screen and stores it in the storage device 130.

[0042] The learning data right information management unit 1232 controls the input of right information through the user terminal of the learning data provider, and associates the input right information with the learning data ID and stores it in the storage device 130. The right information can include, for example, the permission to use, the scope of the permission to use, the information on non-usable content, the rate setting information, etc. The scope of the permission to use is information that defines the scope of the purpose and application of the generated AI model generated using the learning data. For example, when the generated AI model is used for non-profit purposes, it is permitted, but when the generated AI model is used for profit purposes, it is not permitted. The purpose of using the learning data can be set, such as when the generated AI model is used for in-house education, the application can be permitted. The information on non-usable content is information that defines the scope of the content and data generated by the generated AI model, or the scope of the service provided using the generated AI model, age restrictions, etc. For example, restrictions can be set such as not permitting the application to the generated AI model used for content, data, and services provided to children under 12 years old.

[0043] The rate setting information is information that defines the usage fee for the learning data. The definition of the usage fee is optional. For example, the usage fee can be set according to the number of learning data used in the learning process of the generated AI model, the usage fee according to the selling price of the generated AI model, the usage fee for the revenue obtained by the business operator in the business using the generated AI model, etc.

[0044] The learning data identity management unit 1233 performs the generation process and the identity check process of the identity check information of the learning data. For example, using a hash function based on a predetermined algorithm, the hash value of the learning data is generated. This hash value is managed in association with the learning data ID and is used to confirm (prove) that the learning data is genuine and has not been tampered with or edited. The learning data identity management unit 1233 can, for example, when there is an identity check request from the model developer, generate the hash value of the learning data received from the user terminal of the model developer, compare it with the registered hash value, and provide the identity check result to the model developer.

[0045] In addition, the learning data identity management unit 1233 can provide a model developer's user terminal with an identity check information generation program including a hash function based on a predetermined algorithm. The model developer executes the provided identity check information generation program on the learning data to generate a hash value. The model developer transmits the generated hash value from the user terminal to the management device 100. The learning data identity management unit 1233 compares the received hash value with the registered hash value, performs an identity check process, and provides the result to the user terminal.

[0046] This learning data identity check function is a function for determining whether the learning data is the learning data registered in the rights management system, and has the meaning for the model developer to confirm the authenticity of the learning data and the meaning for the model developer to use it to confirm whether the learning data obtained from a place other than the rights management system is the learning data registered in the rights management system. The model developer can confirm in advance whether the learning data is registered as a preparation for generating a generation AI model. If the learning data obtained from outside the rights management system is registered learning data, the model developer can submit an application for permission to use it to the rights management system.

[0047] The learning data use permission management unit 1234 performs a use permission process for the learning data registered in the rights management system. The learning data use permission management unit 1234 accepts an application for permission to use the registered learning data from the model developer's user terminal and performs a use permission process for each learning data ID.

[0048] The learning data management unit 123 centrally manages learning data attribute information including a learning data ID, identity check information, a learning data provider ID, a rights information ID, an explanation of the data set, etc. for each learning data. Then, the learning data management unit 123 can control the downloaded learning data to be downloadable to the model developer's user terminal.

[0049] Note that, as described above, there are cases where the training data includes a data set composed of a plurality of data. However, since the data set has a large data volume, it may be difficult to upload it to the management device 100. Therefore, it may be configured to input only the registration information of the training data into the management device 100 without uploading the training data itself to the management device 100.

[0050] Specifically, the training data management unit 123 controls so that the above-described identity check information generation program managed by the training data identity management unit 1233 can be provided to the user terminal of the training data provider. Then, the training data provider executes the provided identity check information generation program on the training data to generate a hash value. The training data provider transmits the generated hash value to the management device 100. The training data registration control unit 1231 can assign a training data ID to the received hash value for registration and generate training data attribute information.

[0051] In this case, the model developer can obtain the training data by means other than the management device 100. For example, the training data can be obtained directly from the training data provider or through the download site used by the training data provider. When the model developer obtains the training data through a route other than the management device 100, the model developer accesses the management device 100 and performs an identity check of the obtained training data through the identity check function provided by the training data identity management unit 1233, and can apply for permission to use the obtained training data through the permission function provided by the training data use permission management unit 1234.

[0052] The model management unit 124 includes a model registration control unit 1241, a model identity management unit 1242, and a model use permission management unit 1243.

[0053] The model registration control unit 1241 assigns a model ID for each generation AI model generated using the learning data registered in the management device 100, and generates model registration information associating the learning data ID of the used learning data with the model ID. Specifically, the model developer registers the generation AI model generated using the learning data registered in the rights management system in the rights management system. The model registration control unit 1241 receives model property information through the generation AI model registration screen and assigns a model ID. Then, along with the registration of the generation AI model, a identity check information generation program including a hash function for generating identity check information is provided to the user terminal. The model developer executes the identity check information generation program on the source data of the generation AI model including the model property information to generate a hash value. The generated hash value is input through the generation AI model registration screen together with the model property information.

[0054] The model registration control unit 1241 generates model registration information including the model ID, model developer ID, model property information, and identity check information (hash value), and stores it in the storage device 130.

[0055] Here, the model property information will be described. The model property information includes the learning data ID and the number of data used in the learning process of the generation AI model, and can be configured to include any property information of the generation AI model in addition to the learning data ID and the number of data. The model registration control unit 1241 can be configured to generate model registration information including identity check information generated for the source data of the generation AI model including the model property information. By including the learning data ID and the number of data in the hash value for checking the identity of the generation AI model in this way, even if the generation AI models use the same learning data ID but have different numbers of data, different hash values (individual model registration information) will be generated. The rights management system registers and manages the generation AI model, the learning data ID, and the number of data in an inseparable state.

[0056] The model identity management unit 1242 provides the same functions as the learning data identity management unit 1233 of the learning data management unit 123. That is, it provides a function to determine whether the generated AI model is the one registered in the rights management system, and is used by the business operator to confirm whether the generated AI model used is the one registered in the rights management system. The business operator can check in advance whether the generated AI model is registered as a preparation for conducting business using the generated AI model. If the generated AI model to be used by the business operator is a registered generated AI model, the business operator shall submit an application for license to use to the rights management system.

[0057] The model license management unit 1243 conducts license processing for the use of the generated AI model for business operators. The model license management unit 1243 receives an application for license to use the registered generated AI model from the user terminal of the business operator, and conducts license processing to associate the model ID with the business operator.

[0058] The usage fee management unit 125 is composed of a business information reception unit 1251, a usage fee calculation unit 1252, a usage fee payment management unit 1253, and a usage fee distribution management unit 1254.

[0059] The business information reception unit 1251 provides a function to receive revenue-related information of the business using the generated AI model licensed for use from the user terminal of the business operator. The usage fee calculation unit 1252 calculates the usage fee according to the rights information associated with the learning data based on the revenue-related information. Specifically, the usage fee to be paid by the business operator is calculated based on the rate setting information included in the rights information and the business revenue obtained by using the generated AI model included in the revenue-related information. The usage fee payment management unit 1253 conducts the billing process for the usage fee to be paid by the business operator. The billing process is a process to demand the business operator to pay the usage fee. Note that the billing process may be configured to include settlement processes such as credit settlement and account transfer associated with the billing. The usage fee distribution management unit 1254 totals the usage fees received from the business operator and calculates the distribution amount for each learning data provider. The usage fee management unit 125 conducts a process to pay the calculated distribution amount to the transfer destination of the learning data provider.

[0060] The storage device 130 stores user information 1301, learning data 1302, right information 1303, learning data attribute information 1304, learning data usage permission information 1305, model registration information 1306, model usage permission information 1307, business information 1308, usage fee charging information 1309, and usage fee distribution information 1310.

[0061] Figure 3 is a diagram showing a processing flow centered on a learning data provider in the rights management system.

[0062] As shown in Figure 3, the learning data provider accesses the management device 100 to perform user registration (S301). The management device 100 assigns a learning data provider ID and performs registration processing of the input user information (S101). The learning data provider logs in based on the learning data provider ID through the user terminal (S302), and the management device 100 performs login authentication processing (S102).

[0063] After login authentication, the learning data provider performs registration processing of learning data (S303). The learning data provider uploads learning data to the management device 100 through the user terminal. The management device 100 performs generation processing of identity check information for the uploaded learning data (S103). The management device 100 assigns a learning data ID to the uploaded learning data, associates the learning data ID with the identity check information, and stores the learning data (S104). Along with the upload of the learning data, the learning data provider inputs the right information of the learning data to be registered (S304). The management device 100 performs registration processing of the input right information (S105), assigns a right information ID to the input right information, and associates it with the learning data ID. The right information associated with the learning data ID is stored (S106). The management device 100 generates learning data attribute information through the learning data registration process and the right information registration process, and stores it in the storage device 130.

[0064] Figure 4 is a diagram showing a processing flow centered on a model developer in the rights management system.

[0065] As shown in FIG. 4, the model developer accesses the management device 100 to perform user registration (S401). The management device 100 assigns a model developer ID and performs registration processing of the input user information (S111). The model developer logs in based on the model developer ID through the user terminal (S402), and the management device 100 performs login authentication processing (S112).

[0066] After login authentication, the model developer can search for learning data (S403). The management device 100 performs a search process based on the search conditions specified by the model developer and provides the search results to the user terminal of the model developer (S113).

[0067] The management device 100 controls the right information of the searched learning data to be viewable (S114), and the model developer checks the right information through the user terminal (S404).

[0068] The model developer applies for a license to use the desired learning data (S405). The management device 100 accepts the license application and performs license processing, associating the applied learning data with the model developer (S115). The management device 100 controls the licensed learning data to be downloadable, and based on the acquisition request of the learning data from the model developer (S406), provides the learning data to the user terminal of the model developer (S116).

[0069] The model developer develops and generates a generative AI model using the learning data registered in the management device 100 (S407). The model developer registers the generated generative AI model in the management device 100. At this time, the model developer records the learning data ID and the number of data of the used learning data in the model property information of the generated generative AI model.

[0070] The model developer acquires the identity check information generation program from the management device 100 (S409, S117), and executes the identity check information generation program on the user terminal for the source data including the model property information in which the learning data ID and the number of data are recorded, to generate the identity check information (S410).

[0071] The model developer inputs the identity check information and the model property information through the model registration screen (S411). The management device 100 assigns a model ID, generates (S118) and saves (S119) the model registration information including the identity check information and the model property information.

[0072] FIG. 5 is a diagram showing a processing flow centered on an operator in the rights management system.

[0073] As shown in FIG. 5, the operator accesses the management device 100 and performs user registration (S501). The management device 100 assigns an operator ID and performs registration processing of the input user information (S131). The operator logs in based on the operator ID through the user terminal (S502), and the management device 100 performs login authentication processing (S132).

[0074] After login authentication, the operator can search for the generated AI model (S503). The management device 100 performs a search process based on the search conditions specified by the operator, and provides the search result to the operator's user terminal (S133).

[0075] The operator applies for a license to use the desired generated AI model (S404). The management device 100 accepts the license application and generates usage fee billing information based on the rights information of the learning data associated with the generated AI model (S134). The management device 100 provides the generated usage fee billing information to the operator's user terminal, and the operator makes a declaration of consent to the license to use through the user terminal (S505). The management device 100 performs license processing based on the operator's consent to the license, and associates the applied generated AI model with the operator (S135).

[0076] The usage fee billing information is, for example, the billing information of usage fees by business operator ID and model ID, which is generated based on the right information of the training data associated with the generative AI model. The usage fee billing information can be configured to include revenue-related information registered by the business operator, the billing history of usage fees, and the payment history (deposit history) of usage fees.

[0077] The business operator conducts a business using the registered machine learning model and creates revenue-related information regarding the usage fee billing information (S506). The business operator registers (reports) the created revenue-related information to the management device 100 (S507).

[0078] The management device 100 calculates the usage fee corresponding to the revenue-related information based on the usage fee billing information (S136). The management device 100 performs the billing process for the calculated usage fee (S137), and the business operator confirms the billing content (S508).

[0079] The business operator conducts the deposit process for the billed usage fee, with the operator operating the management device 100 (right management platform) as the payee (S509). As described above, the business operator may appropriately apply known methods such as credit card settlement or account transfer for the deposit process of the usage fee. The management device 100 confirms the deposit of the usage fee from the business operator (S138), generates a deposit history (payment history), and updates the usage fee billing information (S139).

[0080] FIG. 6 is a diagram for explaining the distribution process flow and the distribution mechanism to the training data provider in the right management system.

[0081] As shown in FIG. 6, the management device 100 performs usage fee distribution processing at regular intervals (S151). Specifically, the management device 100 aggregates usage fee billing information (usage fee billing log) (S152), and distributes the aggregated usage fees by learning data ID used in the generation AI model (S153). That is, the distribution amount is calculated for each learning data ID. The management device 100 performs a process of paying the calculated distribution amount to the transfer destination of the corresponding learning data provider (S154). After the payment process is completed, a deposit notification is sent to the user terminal of the learning data provider (S155).

[0082] As a rights management platform, the management device 100 identifies the learning data of the learning data provider and associates it with the generation AI model generated using the learning data. Then, by associating the business operator who actually conducts business using the generation AI with the generation AI model, the business conducted by the business operator using the generation AI model can be associated with the learning data provider. Thereby, a route for distributing the business income of the business operator to the learning data provider can be formed.

[0083] As shown in FIG. 6, there may be a case where the model developer and the business operator belong to the same organization. In this case, user registration is performed individually as the model developer and the business operator, and after the above-mentioned learning data use permission and model use permission, the business operator and the learning data provider are associated with each other.

[0084] The rights management system of the present embodiment can provide a rights management platform that can connect relevant parties well when creating new content or business using generation AI.

[0085] Through the rights management platform, the learning data provider, the model developer, and the business operator who conducts business using the generation AI model can be managed in a unified manner, and the learning data that contributed to the generation AI model that brings business income can be distributed to the learning data provider.

[0086] Although the embodiments have been described above, the rights management system is applicable not only to generative AI models but also to the first machine learning model described above. That is, not only business profits from new content or new data generated by a generative AI model, but also business profits from predictions or judgments made by machine learning models such as conventional image recognition AI models and voice recognition AI models, the learning data may contribute to the machine learning model. Therefore, regardless of whether the machine learning model generates new content or new data, it is possible to protect the person who has the rights to the learning data.

[0087] In addition, each function constituting the rights management system can be realized by one computer device or distributed and realized by a plurality of computer devices. Also, the storage area that can store learning data and provide the learning data to the user terminal of the model developer may be configured to be provided separately from this rights management system. For example, it is also possible to construct a rights management system in cooperation with an external storage device that provides a data service for storing and downloading data.

[0088] In addition, the learning data identity management unit 1233 and the model identity management unit 1242 may be configured as one identity management unit to provide functions for generating identity check information, providing an identity check information generation program, and collating with registered identity check information for both the learning data and the machine learning model including the generative AI model.

[0089] In addition, each function constituting the rights management system can be realized by a program. A computer program prepared in advance to realize each function is stored in an auxiliary storage device, and a control unit such as a CPU reads the program stored in the auxiliary storage device into the main storage device, and the control unit executes the program read into the main storage device, whereby the functions of each part can be operated.

[0090] In addition, the above program can also be provided to a computer in a state recorded on a computer-readable recording medium. Examples of computer-readable recording media include optical discs such as CD-ROMs and Blu-ray (registered trademark) Disc Rewritable, phase change optical discs such as DVD-ROMs, magneto-optical discs such as MO (Magneto Optical), magnetic discs such as floppy (registered trademark) discs and hard discs, memory cards such as SD memory cards and USB flash drives. Also included as a recording medium are hardware devices such as integrated circuits (IC chips such as ROMs and RAMs) specially designed and configured for the purpose of the present invention. Furthermore, the present invention including the above program is not limited to being executed on the architecture of a von Neumann type computer, and may be executed on the architecture of a so-called non-von Neumann type computer such as a neurocomputer based on the mechanism of a brain nerve circuit or a quantum computer applying quantum mechanics to information processing.

[0091] Although the embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. This novel embodiment can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are included in the invention described in the claims and its equivalent scope.

Explanation of Reference Numerals

[0092] 100 Management device 110 Communication device 120 Control device 121 User interface control unit 122 User management unit 1221 Registration control unit 123 Learning data management unit 1231 Learning data registration control unit 1232 Learning data right information management unit 1233 Learning Data Identity Management Department 1234 Learning Data Usage Permission Management Department 124 Model Management Department 1241 Model Registration Control Department 1242 Model Identity Management Department 1243 Model Usage Permission Management Department 125 Usage Fee Management Department 1251 Business Information Reception Department 1252 Usage Fee Calculation Department 1253 Usage Fee Payment Management Department 1254 Usage Fee Allocation Management Department 130 Memory Device 1301 User Information 1302 Learning Data 1303 Right Information 1304 Learning Data Attribute Information 1305 Learning Data Usage Permission Information 1306 Model Registration Information 1307 Model Usage Permission Information 1308 Business Information 1309 Usage Fee Billing Information 1310 Usage Fee Allocation Information

Claims

1. A rights management system for learning data used in a generative AI model that generates new content or new data, comprising: a learning data management unit that assigns a learning data ID to the learning data of a learning data provider, accepts registration of the rights information of the learning data to which the learning data ID is assigned, and accepts a license application for the learning data from a model developer; a model management unit that accepts registration of a generative AI model generated using the licensed learning data, generates model registration information associated with the learning data ID of the used learning data, and accepts a license application for the registered generative AI model from a business operator; a royalty management unit that calculates a royalty according to the rights information of the learning data associated with the licensed generative AI model for a business operator who obtains business revenue through the content or data generated by the generative AI model, aggregates the royalties received from the business operator, and calculates a distribution amount for each learning data provider; A rights management system, characterized by comprising the above.

2. The learning data management unit: A learning data registration control unit that assigns a learning data ID to the learning data uploaded from the terminal of the learning data provider and registers it in a predetermined storage area; A learning data rights information management unit that controls the input of rights information through the terminal of the learning data provider and registers the input rights information in a predetermined storage area in association with the learning data ID; A learning data license management unit that accepts a license application for the registered learning data from the terminal of the model developer and performs license processing for each learning data ID. The rights management system according to claim 1, wherein the learning data management unit controls the licensed learning data to be downloadable to the terminal of the model developer.

3. The learning data management unit: A learning data identity management unit that executes an identity check information generation program based on a hash function composed of a predetermined algorithm for the uploaded learning data to generate identity check information; The rights management system according to claim 2, wherein the learning data registration control unit registers the learning data by associating the learning data ID with the identity check information.

4. The learning data management unit: ​ A learning data identity management unit that provides an identity check information generation program based on a hash function configured with a predetermined algorithm to the terminal of a learning data provider, A learning data registration control unit that receives identity check information generated by executing the identity check information generation program for learning data from the terminal of the learning data provider, assigns a learning data ID, and registers it in a predetermined storage area, A learning data rights information management unit that controls the input of rights information through the terminal of the learning data provider and registers the input rights information in a predetermined storage area in association with the learning data ID, A learning data use permission management unit that receives a use permission application for the registered learning data from the terminal of the model developer and performs use permission processing for each learning data ID, and the rights management system according to claim 1, characterized by comprising the same.

5. The learning data identity management unit receives learning data or identity check information of learning data from each terminal of a learning data provider, a model developer, and / or a business operator, executes an identity check information generation program for the received learning data, and collates the generated identity check information or the received identity check information with the registered identity check information, and outputs an identity check result. The rights management system according to claim 3 or 4, characterized by the above.

6. The model management unit, A model registration control unit that assigns a model ID for each generation AI model generated using learning data and generates model registration information including the learning data ID and the model ID of the used learning data, A model use permission management unit that receives a use permission application for the registered generation AI model from the terminal of the business operator and performs use permission processing for each model ID, The rights management system according to claim 1 or 2, characterized by comprising the same.

7. The model management unit, Further comprising a model identity management unit that provides an identity check information generation program based on a hash function configured with a predetermined algorithm to the terminal of the model developer, The model registration control unit receives identity check information generated by executing the identity check information generation program for the source data of the generation AI model from the terminal of the model developer, and generates model registration information including the learning data ID, the model ID, and the identity check information. The rights management system according to claim 6, characterized by the above.

8. The generated AI model holds model property information including the learning data ID and the number of data of the learning data used. The model registration control unit receives, from the terminal of the model developer, identity check information generated for the source data of the generated AI model including the model property information, and generates model registration information. The rights management system according to claim 7, characterized in that.

9. The model identity management unit receives, from the terminals of the learning data provider, the model developer, and / or the business operator, identity check information generated for the source data of the generated AI model, collates the received identity check information with the registered identity check information, and outputs an identity check result. The rights management system according to claim 7, characterized in that.

10. The royalty management unit A business information reception unit that receives, from the terminal of the business operator, revenue-related information related to the business revenue obtained through the content or data generated by the licensed generated AI model, A royalty calculation unit that calculates a royalty according to the rights information of the learning data associated with the licensed generated AI model based on the revenue-related information, A royalty payment management unit that performs a billing process for the royalty paid by the business operator, A royalty distribution management unit that aggregates the royalties received from the business operator, calculates distribution amounts for each learning data provider, and pays the distribution amounts to the transfer destinations of the learning data providers. The rights management system according to claim 1, characterized by having.

11. A program executed by a computer that performs rights management of learning data used for a generated AI model that generates new content or new data, A first function that assigns a learning data ID to the learning data of the learning data provider, receives registration of rights information of the learning data to which the learning data ID is assigned, and receives a license application for the learning data from the model developer, A second function that receives registration of a generated AI model generated using the licensed learning data, generates model registration information associated with the learning data ID of the used learning data, and receives a license application for the registered generated AI model from the business operator. For a business operator who obtains business revenue through the content or data generated by a generative AI model, calculate a usage fee according to the right information of the training data associated with the licensed generative AI model, aggregate the usage fees received from the business operator, and calculate the distribution amount for each training data provider. A third function to do so, A program for realizing the above.

12. A rights management system for training data, A training data management unit that assigns a training data ID to the training data of a training data provider, accepts registration of the right information of the training data to which the training data ID is assigned, and accepts a license application for the training data from a model developer. A model management unit that accepts registration of a machine learning model generated using the licensed training data, generates model registration information associated with the training data ID of the used training data, and accepts a license application for the registered machine learning model from a business operator. A usage fee management unit that calculates the usage fee to be paid by the business operator according to the right information of the training data associated with the licensed machine learning model, aggregates the usage fees received from the business operator, and calculates the distribution amount for each training data provider. A rights management system characterized by having the above.

Images