Voice Input Structuring for Accurate Information Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Voice recordings often lack grammatical correctness and explicit subjects, making them inaccurate for information retrieval, and require repeated access to obtain desired information, which can disrupt research and development or diagnostic processes.
Innovation Solution
A recording and organizing method that structures and records input items using a tree structure, allowing for accurate information retrieval by inputting user comments through an input unit, structuring them, and presenting structured records, with options for text and voice input, time and non-time record division, user discrimination, and comparison with other information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If voice recording is used for information input, then hands-free operation and continuous work are enabled, but grammatical correctness and information accuracy deteriorate
Solution Approach 1:
The patent introduces an intermediary processing system between voice input and information storage. The voice recording device captures speech, which then passes through automatic speech recognition and natural language processing intermediaries that convert colloquial speech into structured, accurate information suitable for storage and retrieval, thereby maintaining both hands-free operation and information accuracy
Solution Approach 2:
The patent replaces the mechanical system of manual information entry with an automated voice-based system. By substituting keyboard typing and manual device operation with voice commands processed through recognition software, the system enables hands-free operation while maintaining information accuracy through automated processing
2Loss of information
If repeated information retrieval is performed through terminal input and reading, then desired information can be obtained, but work interruption and time loss increase
Solution Approach 1:
The patent applies preliminary action by pre-structuring and organizing information during the initial recording phase using a hierarchical classification system. Information is categorized and tagged in advance, enabling rapid retrieval through voice commands without requiring repeated terminal input and reading operations, thus reducing work interruption time while maintaining complete information access
Solution Approach 2:
The patent creates structured copies of information in multiple formats and organizational levels. Instead of requiring users to repeatedly access and read original terminal data, the system generates organized copies indexed by category, time, and relevance, allowing quick voice-based retrieval that minimizes work disruption
3Ease of operation
If simple voice recording is used, then operation simplicity is maintained, but information structure and usability deteriorate
Solution Approach 1:
The patent segments information into hierarchical categories and subcategories during automated processing. The system divides unstructured voice input into organized segments with clear relationships, creating a structured information architecture that remains simple to access through voice commands while providing comprehensive organization internally
Solution Approach 2:
The patent creates a universal information management system that handles multiple information types (voice, text, data) through a single integrated platform. The hierarchical structure serves multiple functions including storage, retrieval, analysis, and cross-referencing, maintaining operational simplicity while providing complex organizational capabilities
Data Source
AI summary
People's statements are not always grammatically correct. The subjects thereof are not specified since they are expressed by pronouns, or are omitted in Japanese and the like. Therefore, the record of speech and the like in unchanged form is incorrect as information, and in many cases, it is difficult to utilize the record in unchanged form. A method for structuring and recording an inputted item comprises: an input step for inputting a user's statement by an input unit; a structuring step for, in a structuring unit, creating a structured record in which the user's statement inputted in the input step is applied to a predetermined structure, and storing the structured record in a recording unit; and a presentation step for presenting, to the user, at least part of the structured record by a presentation unit, the structure having a tree structure composed of a plurality of hierarchies.


