Conversation Controller Plan-Based Flow Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional conversation controllers are unable to output reply sentences in sequence to maintain a predetermined conversation flow, as the flow is determined by user utterances, making it difficult to continue a previously prepared conversation.
Innovation Solution
A conversation controller with a plan memory unit to store plans containing reply sentences and next candidate designation information, allowing it to select and output reply sentences in a predetermined order based on user utterances, while withholding or resuming responses according to user interest.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the conversation controller determines the conversation flow according to user utterances, then it can respond flexibly to user input, but it cannot output reply sentences in a predetermined sequence to maintain a previously prepared conversation flow
Solution Approach 1:
The conversation controller dynamically switches between two operational modes: plan-based mode for maintaining predetermined conversation flows and discourse space mode for flexible user response. This dynamic adaptability allows the system to transition between structured and unstructured conversation modes based on the current interaction context, resolving the contradiction between flow stability and user flexibility.
Solution Approach 2:
The system changes the control parameter from purely user-driven to plan-driven by introducing a flag that indicates whether to follow a predetermined plan or respond freely to user input. This parameter change enables the conversation controller to switch between maintaining a prepared conversation flow and adapting to user utterances, thus resolving the contradiction.
2Stability of the object's composition
If the conversation controller continuously outputs reply sentences according to a predetermined plan, then it maintains conversation flow, but it cannot respond to user utterances about unrelated topics
Solution Approach 1:
The conversation controller dynamically adjusts its behavior based on the relationship between user utterances and the current plan. When a user utterance is related to the current plan, the system continues following the predetermined flow. When unrelated, it switches to discourse space mode to handle the new topic, thus maintaining both flow continuity and adaptability to unrelated topics.
Solution Approach 2:
The system continuously monitors user utterances to determine their relationship to the current conversation plan. This feedback mechanism allows the controller to detect when the user has shifted to an unrelated topic and switch modes accordingly, enabling the system to maintain predetermined flow when appropriate while responding to unrelated topics when necessary.
3Duration of action of stationary object
If the conversation controller withholds next candidate reply sentences, then it can resume conversations later, but it increases memory requirements and processing complexity
Solution Approach 1:
The system extracts and stores only the essential elements needed for conversation resumption (the flag indicating plan status and the current plan identifier) rather than storing entire conversation histories or multiple candidate replies. This extraction approach enables conversation resumption while minimizing memory requirements and processing complexity.
Data Source
AI summary
A conversation controller outputs a reply sentence according to a user utterance. The conversation controller comprises a conversation database and a conversation control unit. The conversation database stores a plurality of plans. Each plan has a reply sentence and one or more pieces of next plan designation information for designating a next candidate reply sentence to be output following the reply sentence. The conversation control unit selects one of the plans stored in the conversation database according to a user utterance and outputs a reply sentence which the selected plan has. Then, the conversation control unit selects one piece of the next plan designation information which the plan has according to a next user utterance and outputs a next candidate reply sentence on the basis of the selected piece of the next plan designation information. Some plans have a plurality of reply sentences into which one explanatory sentence is divided.