Back to results
US 2026/0291870 A1Published application
System
AssigneeSoftBank Group Corp.
InventorToshiharu SASAKI
Life of this patent
Estimating the term…
Abstract
A system includes a processor that is configured to obtain and analyze information from various databases on an information network to identify globally assigned identification information, generate and analyze a prompt sentence that is to be provided to a generative AI model, the prompt sentence being generated based on information received from a user and instructing the generative AI model to perform analysis, and classify information into an appropriate service or department based on a result of the analysis performed by the generative AI model.
Claims20
A system comprising:
- circuitry configured to:
- obtain, via a communication interface coupled to a packet-switched network, identification information assigned on a global scale by querying information sources on an information network, and analyze the obtained identification information;
- collect user information including communication information, device information, and location information from a user device via the communication interface, normalize the user information into a fixed internal representation, and store the normalized user information in a storage device;
- generate, based on the normalized user information, a prompt sentence that instructs a generative neural network model to perform analysis, transmit the prompt sentence and context data to the generative neural network model, receive an analysis result, and determine a usage policy for the user information and the identification information from the analysis result;
- combine at least a portion of the normalized user information into a fixed-format character string in accordance with the usage policy, apply a cryptographic transformation function to the character string to generate unique identification information that does not require network address translation, and store in the storage device an association among the generated unique identification information, the normalized user information, and the acquired identification information;
- assign and deliver the unique identification information to the user device via the communication interface;
- and
- acquire acoustic information from the user device, recognize an emotional state of a user by applying signal processing to the acoustic information, and adjust at least one communication control parameter to optimize a communication path or communication resource between the circuitry and the user device based on the recognized emotional state and the unique identification information.
5 more elements
A system comprising:
- circuitry configured to:
- obtain, via a communication interface coupled to a packet-switched network, globally-assigned identification information from information sources on an information network, and collect user information including communication information, device information, and location information from a user device;
- normalize the user information into a fixed internal representation and store the normalized user information in a storage device;
- generate a prompt sentence based on the normalized user information, transmit the prompt sentence and context data to a generative neural network model, receive an analysis result, and determine a usage policy from the analysis result;
- combine a portion of the normalized user information into a fixed-format character string in accordance with the usage policy, apply a cryptographic transformation function to generate unique identification information not requiring network address translation, and store an association among the unique identification information, the normalized user information, and the obtained identification information in the storage device;
- acquire acoustic information from the user device, apply signal processing to recognize an emotional state of a user, and adjust at least one communication control parameter for a communication path of the user device based on the recognized emotional state and the unique identification information;
- and
- classify the analysis result into a service category, route a communication session of the user device to a service endpoint associated with the service category, and transmit the unique identification information to the user device via the communication interface.
5 more elements
A method comprising:
- obtaining, via a communication interface coupled to a packet-switched network, identification information assigned on a global scale by querying information sources on an information network, and analyzing the obtained identification information;
- collecting user information including communication information, device information, and location information from a user device via the communication interface, normalizing the user information into a fixed internal representation, and storing the normalized user information in a storage device;
- generating, based on the normalized user information, a prompt sentence that instructs a generative neural network model to perform analysis, transmitting the prompt sentence and context data to the generative neural network model, receiving an analysis result, and determining a usage policy from the analysis result;
- combining at least a portion of the normalized user information into a fixed-format character string in accordance with the usage policy, applying a cryptographic transformation function to generate unique identification information that does not require network address translation, and storing in the storage device an association among the generated unique identification information, the normalized user information, and the acquired identification information;
- assigning and delivering the unique identification information to the user device via the communication interface;
- and
- acquiring acoustic information from the user device, recognizing an emotional state of a user by applying signal processing to the acoustic information, and adjusting at least one communication control parameter to optimize a communication path or communication resource based on the recognized emotional state and the unique identification information.
4 more elements
Dependencies are inferred from the claim text.