Digital Data Tagging Apparatus Resolving Homonym Ambiguity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital data tagging apparatuses face difficulties in discriminating homonyms and synonyms, leading to ineffective search results due to the inability to accurately assign tags from audio data.
Innovation Solution
A digital data tagging apparatus that includes a processor configured to extract words/phrases from audio data, determine tag candidates based on association thresholds, and assign tags by considering homonyms and synonyms, with options for user selection and display of tag candidates, including synonyms with high pronunciation and meaning similarity, and homonyms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If audio data is used to extract words/phrases for tagging, then tagging efficiency is improved, but discrimination accuracy between homonyms deteriorates
Solution Approach 1:
The patent introduces context information as an intermediary element between the extracted word/phrase and the tag assignment. By analyzing surrounding words, sentence structure, and semantic context, the system mediates the ambiguous relationship between homonyms and their meanings, enabling accurate tag selection without manual intervention while maintaining high tagging efficiency
Solution Approach 2:
The system changes the parameter of tag candidate selection by considering multiple factors including context relevance, frequency of occurrence, and semantic similarity. By dynamically adjusting which tag candidates are presented based on these parameters, the system resolves homonym ambiguity while maintaining efficient automated tagging
2Adaptability or versatility
If multiple tag candidates including synonyms are assigned, then search coverage is improved, but tag selection complexity increases
Solution Approach 1:
The patent applies local quality by differentiating the presentation and weighting of tag candidates based on their type. Synonyms are presented with visual indicators showing their relationship to the extracted term, while homonyms are distinguished through context-based ranking. This localized differentiation helps users quickly identify the most appropriate tag without being overwhelmed by the full set of candidates
Solution Approach 2:
The system performs preliminary filtering and ranking of tag candidates before presentation to the user. By pre-processing the tag candidate list to highlight the most relevant options based on context analysis, the system reduces the effective selection complexity while maintaining comprehensive search coverage through the inclusion of multiple synonym variants
Data Source
AI summary
It is possible for a user to easily assign a desired tag by using a voice, regardless of homonyms and synonyms having different expressions. In a digital data tagging apparatus, a tagging method, a program, and a recording medium according to an embodiment of the present invention, a digital data acquisition unit acquires digital data to which a tag is assigned, and an audio data acquisition unit acquires audio data related to the digital data. A word/phrase extraction unit extracts a word/phrase from the audio data, a tag candidate determination unit determines one or more tag candidates of which a degree of association with the word/phrase is equal to or more than a first threshold value from among a plurality of tag candidates, which are stored in advance in a tag candidate storage unit, as a first tag candidate, and a tag assignment unit assigns at least one of a tag candidate group including the word/phrase and the first tag candidate to the digital data as the tag.