Generation support system, generation support method, and generation support program

Patent metadata

Field	Value
Publication number	US-2025363298-A1
Application number	US-202519187018-A
Country	US
Kind code	A1
Filing date	Apr 23, 2025
Priority date	May 24, 2024
Publication date	Nov 27, 2025
Grant date	—

Abstract

When a natural sentence that is text data of a natural language is inputted, the generation support system can access a language model that interprets the natural sentence and that probabilistically predicts an answer sentence to the natural sentence. The processor outputs a generation-source ticket of a natural sentence related to an incident, a check item group to be checked for the incident, and a first prompt that is the natural sentence requesting that an answer sentence about each check item of the check item group be generated with reference to the generation-source ticket, to the language model, causes the language model to acquire a postmortem including a result of a first trial of generation of an answer sentence about each check item in the first prompt, as a result of output of the generation-source ticket, the check item group, and the first prompt, and outputs the postmortem.

First claim

1 . A generation support system comprising:
a processor that executes a program; and a storage device that stores the program, wherein when a natural sentence that is text data of a natural language is inputted, the generation support system accesses a language model that interprets the natural sentence and that probabilistically predicts an answer sentence to the natural sentence, and wherein the processor executes: a first requesting process of outputting a generation-source ticket of the natural sentence related to an incident, a check item group to be checked for the incident, and a first prompt that is the natural sentence requesting that an answer sentence about each check item of the check item group be generated with reference to the generation-source ticket, to the language model; a first acquiring process of causing the language model to acquire a postmortem including a result of a first trial of generation of an answer sentence about the each check item in the first prompt, as a result of output of the generation-source ticket, the check item group, and the first prompt by the first requesting process; and a first output process of outputting the postmortem acquired by the first acquiring process.

2 . The generation support system according to claim 1 , wherein
the processor executes a first determining process of determining whether an insufficient item is present in a first postmortem acquired by the first acquiring process, the insufficient item being a check item among the check item group that gives the result of the first trial that is not an answer sentence allowing check of the check item, and wherein in the first output process, the processor outputs the first postmortem for which whether the insufficient item is present therein is determined by the first determining process.

3 . The generation support system according to claim 2 , wherein
the processor executes: a search process of retrieving a related ticket related to the generation-source ticket, from a ticket group, based on a search query about the insufficient item; a second requesting process of outputting the related ticket retrieved by the search process and a second prompt requesting generation of the answer sentence about the insufficient item using the related ticket, to the language model; a second acquiring process of causing the language model to acquire a result of a second trial of generation of an answer sentence about the insufficient item in the second prompt, as a result of output of the related ticket and the second prompt by the second requesting process; and a second output process of outputting the result of the second trial acquired by the second acquiring process.

4 . The generation support system according to claim 3 , wherein
each ticket of the ticket group includes a ticket summary vector that is an embedded expression made by vectorizing a ticket summary text summarizing the ticket, and wherein in the search process, the processor retrieves a ticket that corresponds to a search query about the insufficient item and that has a ticket summary vector whose distance to a ticket summary vector of the generation-source ticket is equal to or less than a given distance, from the ticket group, as the related ticket.

5 . The generation support system according to claim 3 , wherein
in the second acquiring process, the processor acquires a second postmortem including the check item group, the result of the first trial, and the result of the second trial, and wherein in the second output process, the processor outputs the second postmortem acquired by the second acquiring process.

6 . The generation support system according to claim 3 , wherein
the processor executes a second determining process of determining whether the insufficient item is present, based on the result of the second trial acquired by the second acquiring process.

7 . The generation support system according to claim 6 , wherein
the processor repeatedly executes the search process, the second requesting process, the second acquiring process, and the second output process until determining by the second determining process that the insufficient item is no longer present.

8 . The generation support system according to claim 1 , wherein
the generation-source ticket includes a task memo summary text summarizing a task memo describing task content on the incident.

9 . A generation support method executed by a generation support system comprising a processor that executes a program, and a storage device that stores the program, wherein
when a natural sentence that is text data of a natural language is inputted, the method allows the generation support system to access a language model that interprets the natural sentence and that probabilistically predicts an answer sentence to the natural sentence, and wherein the method causes the processor to execute: a first requesting process of outputting a generation-source ticket of the natural sentence related to an incident, a check item group to be checked for the incident, and a first prompt that is the natural sentence requesting that an answer sentence about each check item of the check item group be generated with reference to the generation-source ticket, to the language model; a first acquiring process of causing the language model to acquire a postmortem including a result of a first trial of generation of an answer sentence about the each check item in the first prompt, as a result of output of the generation-source ticket, the check item group, and the first prompt by the first requesting process; and a first output process of outputting the postmortem acquired by the first acquiring process.

10 . A generation support program causing a processor that, when a natural sentence that is text data of a natural language is inputted, can access a language model that interprets the natural sentence and that probabilistically predicts an answer sentence to the natural sentence, to execute:
a first requesting process of outputting a generation-source ticket of the natural sentence related to an incident, a check item group to be checked for the incident, and a first prompt that is the natural sentence requesting that an answer sentence about each check item of the check item group be generated with reference to the generation-source ticket, to the language model; a first acquiring process of causing the language model to acquire a postmortem including a result of a first trial of generation of an answer sentence about the each check item in the first prompt, as a result of output of the generation-source ticket, the check item group, and the first prompt by the first requesting process; and a first output process of outputting the postmortem acquired by the first acquiring process.

Assignees

• Hitachi Ltd

Inventors

• Wada Kiyomi
• Masuda Mineyoshi

Classifications

• G06F40/30

  Semantic analysis · CPC title

• G06F40/56

  Natural language generation · CPC title

• G06F40/20 Primary

  Natural language analysis (semantic analysis of natural language G06F40/30) · CPC title