Text generation method, text generation device

The text generation method and device address the limitation of existing technologies by generating summaries based on specified logical structures, ensuring high accuracy and adaptability across various content types through user-selectable templates and optimization, enhancing the flexibility and relevance of generated text.

JP7875796B2Active Publication Date: 2026-06-18HITACHI LTD
3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
HITACHI LTD
Filing Date
2022-12-14
Publication Date
2026-06-18

AI Technical Summary

⚠Technical Problem

Existing text generation technologies, such as those described in Patent Document 1, are unable to generate text based on a specified logical structure, limiting their applicability and accuracy in generating summaries for diverse content types.

⚗Method used

A text generation method and device that utilizes a pre-trained model to process input text and logical structures, incorporating a language information acquisition process, logical text acquisition, and text output process to generate summaries that adhere to a specified logical structure, using models like GPT, GPT-2, GPT-3, T5, or BART, and employing a stack-based algorithm to convert logical templates into logical texts.

🎯Benefits of technology

Enables high-accuracy text generation that reflects the logical structure and author's intent, allowing for customizable summaries that can handle diverse content types and logical structures, including sustainability reports and academic papers, with user-selectable templates and optimization processes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

To generate a text based on a designated logical structure.SOLUTION: The present invention is directed to a text generating method executed by a computer having a storage unit for storing a text generation model previously trained. The method has a language information acquiring process of acquiring language information indicating a processing-target sentence, a logical text acquiring process of acquiring a logical text, and a text output process of inputting the processing target sentence and the logical text into the text generation model to obtain an output text. The text generation model uses, as an input, the logical text and the processing target sentence which indicate a specified logical structure as logical structure between sentences and uses, as an output, text obtained from the processing target sentence and changed in accordance with the specified logical structure.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art