Voice Interaction System Using Episode Data for User Affinity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional voice recognition systems struggle to provide interactions that evoke a sense of affinity with users due to field-specialized interactions recorded in dictionaries, leading to a lack of originality and freshness.
Innovation Solution
An information processing apparatus and system that acquires episode data from user utterances to control interactions, incorporating episodes based on this data to enhance user engagement and affinity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If field-specialized interactions recorded in a dictionary for each field are used, then interaction accuracy in specialized domains is improved, but user affinity and sense of originality deteriorate
Solution Approach 1:
The patent segments the interaction system into multiple components: a dictionary-based specialized interaction module for accurate domain-specific responses, and an episode-based personalized interaction module for generating original, context-aware responses. These segments work together to balance accuracy and user affinity.
Solution Approach 2:
The patent merges two previously separate interaction approaches (dictionary-based specialized interactions and episode-based personalized interactions) into a unified system. The interaction control unit coordinates both modules to provide responses that are both accurate in specialized domains and original in personalization.
2Reliability
If conventional dictionary-based field-specialized interactions are used, then domain expertise is improved, but interaction originality and freshness deteriorate
Solution Approach 1:
The system performs preliminary actions by pre-storing episode data (user preferences, conversation history, personal information) in advance. This allows the episode-based interaction module to quickly generate original responses without complex real-time processing, maintaining both domain expertise and interaction originality.
3Measurement precision
If interactions are based on recorded dictionary data for specialized fields, then response accuracy is improved, but user engagement and sense of affinity deteriorate
Solution Approach 1:
The system implements feedback mechanisms where the interaction control unit monitors user responses and adjusts the balance between dictionary-based and episode-based interactions. This feedback loop ensures that response accuracy is maintained while progressively enhancing user engagement through personalized episodes.
Data Source
AI summary
An information processing apparatus includes an acquisition unit that acquires, from a storage unit that stores episode data of a speaker, the episode data regarding topic information included in utterance data of the speaker, and an interaction control unit that controls an interaction with the speaker so as to include an episode based on the episode data.


