Automated Assistant Dialog Transfer for Multi-Assistant Query Handoff
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated assistants require explicit invocation and may not be able to process user queries efficiently when multiple assistants are present, leading to resource wastage and prolonged dialog interactions.
Innovation Solution
Techniques for transferring dialog data from an initially invoked automated assistant to a second automated assistant, allowing seamless interaction and processing of queries without explicit reinvocation, leveraging capabilities of multiple assistants in the environment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a user explicitly invokes an automated assistant before each query, then the assistant can process the query, but computational resources are wasted and dialog duration is prolonged
Solution Approach 1:
The system performs preliminary actions by maintaining dialog state information in memory during the assistant's active period. When a user speaks a query, the system already has the dialog state ready, so it only needs to process the new query rather than reprocessing the entire dialog history, thus reducing computational resources while maintaining reliable query processing.
Solution Approach 2:
The system extracts only the necessary dialog state information from the complete dialog history and transfers it to the second automated assistant. This selective extraction avoids the need to transfer or reprocess the entire dialog, reducing computational overhead while preserving the essential context needed for reliable query processing.
2Reliability
If a user explicitly invokes an automated assistant for each query, then the assistant can process the query, but dialog duration increases
Solution Approach 1:
The first automated assistant maintains dialog state information in memory during its active period. When the user invokes the second automated assistant, the dialog state is already prepared and can be immediately transferred, eliminating the need for the user to restate previous context and reducing dialog duration while maintaining processing reliability.
Solution Approach 2:
The system uses dialog state information as an intermediary that bridges the two automated assistants. This intermediary contains the essential context needed for query processing, allowing seamless transition between assistants without requiring the user to repeat information, thus reducing dialog duration while maintaining reliable processing.
3Adaptability or versatility
If multiple automated assistants are present in the environment, then more capabilities are available, but it becomes unclear which assistant should process the query
Solution Approach 1:
The system implements feedback by having the first automated assistant detect when a user invokes a second assistant and automatically transfer dialog state information. This feedback mechanism resolves the ambiguity of which assistant should process the query by enabling seamless collaboration between multiple assistants, maintaining ease of operation while leveraging extended capabilities.
4Productivity
If dialog state information is transferred between assistants, then query processing efficiency improves, but data transfer complexity increases
Solution Approach 1:
The system extracts only the essential dialog state information from the complete dialog history and transfers it between assistants. This selective extraction minimizes the amount of data transferred while maintaining processing efficiency, reducing the complexity of the data transfer mechanism compared to transferring entire dialog histories.
Solution Approach 2:
The system creates a copy of the necessary dialog state information and transfers it to the second automated assistant. This copying approach allows the second assistant to process queries independently without requiring complex real-time data sharing mechanisms, improving productivity while keeping the transfer mechanism relatively simple.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems and methods for providing dialog data, from an initially invoked automated assistant to a subsequently invoked automated assistant. A first automated assistant may be invoked by a user utterance, followed by a dialog with the user that is processed by the first automated assistant. During the dialog, a request to transfer dialog data to a second automated assistant is received. The request may originate with the user, by the first automated assistant, and/or by the second automated assistant. Once authorized, the first automated assistant provides the previous dialog data to the second automated assistant. The second automated assistant performs one or more actions based on the dialog data.