Utterance Meaning Analysis Using Keyword Extraction and Topic Derivation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies for processing utterance meaning struggle to efficiently derive and analyze the underlying topics and intentions from voice signals in various environments such as interviews and conferences, lacking effective methods to extract and present relevant information in real-time.
Innovation Solution
An apparatus and method that collect voice signals, convert them into text form, extract keywords, and derive utterance topics based on these keywords, using preset reference values, user selection, correlation analysis, and information about the utterer or multiple utterers, to analyze and display the utterance meaning efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If voice signals are converted into text form and keywords are extracted to derive utterance topics, then the precision of utterance meaning analysis is improved, but the complexity of the processing system increases
Solution Approach 1:
The system segments the utterance processing into distinct stages: voice signal collection, text conversion, keyword extraction, and topic derivation. Each stage handles a specific aspect of the processing, making the overall complex system manageable through modular segmentation of functions.
Solution Approach 2:
Keywords serve as an intermediary element between the raw text information and the derived utterance topics. The keyword extraction process identifies important terms that mediate between the original speech content and the final topic classification, simplifying the analysis by focusing on key representative terms.
2Reliability
If multiple analysis methods including correlation analysis and user selection are implemented, then the reliability of utterance topic derivation is improved, but the time required for processing increases
Solution Approach 1:
The system allows for selective application of analysis methods based on requirements. User selection enables partial action where only necessary verification steps are performed, while correlation analysis provides excessive action for enhanced reliability when needed. This flexibility allows balancing reliability against processing time by adjusting the extent of analysis performed.
Solution Approach 2:
User selection mechanisms provide feedback loops where users can review and adjust derived topics. This feedback enables refinement of topic derivation by incorporating human judgment, improving reliability while allowing the system to learn and adapt to user preferences over time, potentially reducing processing time for future similar cases.
3Quantity of substance
If detailed topic analysis and result visualization are performed, then the quantity of useful information is improved, but the complexity of the output system increases
Solution Approach 1:
The system transforms one-dimensional text keywords into two-dimensional topic structures with hierarchical relationships. Detailed topic analysis adds another dimension by organizing topics into categories and subcategories, while visualization presents this multi-dimensional information in a graphical format, increasing information quantity through dimensional transformation rather than simple addition.
Solution Approach 2:
The system creates simplified copies or representations of the complex topic structure through visualization. Instead of presenting raw processed data, it generates visual copies such as graphs or diagrams that represent the underlying topic relationships, making complex information more accessible while maintaining the full detail of the analysis in the background data structure.
Data Source
AI summary
Disclosed herein is a method for analyzing an utterance meaning. The method includes collecting a voice signal of an utterer; converting the collected voice signal into information in a text form, extracting a keyword of the text information from the text information, and deriving at least one utterance topic on the basis of the extracted keywords of the text information.