Human-Computer Interaction System for Flexible Topic Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current human-computer interaction technologies, such as key-phrase matching, finite-state machines, and slot-filling, are inflexible and limited in handling complex tasks and topic switching in various interaction scenarios.
Innovation Solution
A human-computer interaction processing system and method that describes interaction tasks, performs interaction process control for current inputs, and determines expected next inputs based on interaction scenarios, allowing for flexible dialogue flow management and topic switching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If key-phrase matching is used for interaction management, then implementation is simple, but flexibility and handling capability are insufficient
Solution Approach 1:
The interaction management function is segmented into multiple independent modules: interaction task description module, interaction process control module, and expected interaction input determination module. Each module handles specific aspects of the interaction, allowing independent optimization and improving overall flexibility while maintaining implementation clarity.
Solution Approach 2:
The system transitions from traditional single-dimension keyword matching to multi-dimensional interaction management by incorporating interaction task descriptions, process control states, and expected input predictions. This dimensional expansion enables handling of complex interaction scenarios while maintaining systematic organization.
2Reliability
If finite-state machine is used to cover all interaction flows, then completeness is improved, but complexity increases and topic switching becomes difficult
Solution Approach 1:
The interaction process control module dynamically adjusts the interaction flow based on current state and task requirements, rather than following rigid predetermined state transitions. This dynamic approach maintains complete interaction coverage while enabling flexible topic switching and reducing system complexity.
Solution Approach 2:
The expected interaction input determination module acts as an intermediary that predicts upcoming user inputs and prepares appropriate responses in advance. This mediator enables smooth topic transitions and maintains interaction completeness without requiring complex state machine reconfiguration.
3Productivity
If slot-filling is used for interaction management, then task execution control is improved, but application scope is limited to scenarios with small amount of slot information
Solution Approach 1:
The interaction process control module is designed with universal functionality that can handle various interaction patterns including but not limited to slot-filling. It can manage complex tasks with high amounts of slot information, multi-topic conversations, and diverse interaction scenarios, making the system broadly applicable across different domains.
Solution Approach 2:
The system dynamically adjusts its parameter requirements based on the interaction scenario. Rather than requiring a fixed set of slots, it adapts the amount and type of information needed based on task complexity and scenario requirements, enabling efficient task execution across both simple and complex interactions.
Data Source
AI summary
Embodiments of the specification provide a human-computer interaction processing system, method, storage medium, and electronic device thereof. The method comprises: describing an interaction task in an interaction scenario; performing interaction process control for a current interaction input in the interaction scenario based on the interaction task; and determining an expected next interaction input in the interaction scenario corresponding to the current interaction input based on the interaction process control.