Dialog Topic Graph Visualization via Utterer-Edge Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech recognition technologies face difficulties in visualizing the relationships between multiple topics in dialogues, making it hard to immediately understand connections from textual information.
Innovation Solution
A relation visualizing apparatus that generates a graph using dialog content data, where topic items are represented as nodes and utterer IDs are used as edges to connect nodes, allowing for the visualization of topic relationships through a Sankey diagram or similar graphical representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If dialog content is represented as textual information, then the content can be stored and processed, but it is difficult to immediately read the relation between topics
Solution Approach 1:
The patent transforms one-dimensional textual dialog content into a two-dimensional graphical representation (graph with nodes and edges). Topic items are represented as nodes and their relationships as edges, allowing visual perception of topic connections that cannot be immediately detected in text form. This dimensional transformation enables simultaneous presentation of multiple topic relationships in a spatial layout that human vision can process efficiently.
2Ease of operation
If a graph is generated to visualize topic relationships, then the relations become visible, but additional processing steps are required
Solution Approach 1:
The patent segments dialog content into discrete topic items and their associated relationships, representing each topic as a separate node and each relationship as a separate edge in the graph. This segmentation allows the complex task of visualizing topic relationships to be broken down into manageable components: extracting topics, identifying relationships, and rendering them as a network graph. The modular structure simplifies both the processing logic and the visualization implementation.
Data Source
AI summary
A relation visualizing apparatus to an embodiment includes a memory; and a processor configured to: receive, as an input, dialog content data representing a content on which two or more persons have had a dialog with respect to predetermined two or more topic items and generate a graph, composed of nodes and edges, by using the topic items as the nodes and utterer IDs identifying persons having given utterance to two topic items different from each other as the edges, the edges connecting nodes corresponding to the two topic items; and display the generated graph.