Dialog Device Attribute Matching for Natural Conversation Flow
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing dialog devices struggle to maintain a natural conversation with users when changing information sources, leading to artificiality in conversational sentences due to weak linkage between sentences generated from different news information.
Innovation Solution
A dialog device and method that includes an information acquisition unit, conversation execution unit, attribute setting unit, and acquisition control unit to ensure continuity by matching attribute information between content information sources, allowing for seamless transitions in conversation topics while maintaining user engagement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of moving object
If content information is changed frequently to maintain continuous conversation with the user, then the conversation can continue longer, but the linkage between conversational sentences becomes weak and the user feels artificiality
Solution Approach 1:
The system extracts attribute information from content information and uses it as feedback to select subsequent content information that shares common attributes, ensuring coherent conversation flow while enabling continuous topic changes
Solution Approach 2:
The system changes the parameter of content information selection from random or fixed to dynamic selection based on attribute information matching, allowing flexible topic transitions while maintaining conversational coherence
2Productivity
If content information is changed without discretion to enable continuous conversation, then the conversation can proceed without repetition, but the conversational sentences lack natural linkage and appear artificial
Solution Approach 1:
Attribute information serves as an intermediary between different content information pieces, bridging the gap between unrelated topics and enabling smooth transitions that maintain conversational naturalness
Solution Approach 2:
The attribute information extraction and matching mechanism provides a universal framework that works across different content domains and conversation types, enabling both topic diversity and conversational coherence
Data Source
AI summary
A dialog device for executing a conversation with a user using a conversational sentence generated from content information is provided. The dialog device includes an information acquisition unit that acquires the content information from an information source, and changes the content information used in generation of the conversational sentence, and continues the conversation with the user. The dialog device sets multiple attribute information pieces with respect to specific content information used in generation of the conversational sentence. The dialog device controls the information acquisition unit to acquire other content information, having at least a part of attribute information pieces equal to the specific content information, as a candidate used in generation of the conversational sentence next to the specific content information.


