Virtual Assistant Proactive Action Recommendations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users may not be aware of all the capabilities of virtual assistants, leading to inefficiencies in discovering and utilizing their features, as they often require explicit user requests for information or actions.
Innovation Solution
The virtual assistant is configured to proactively output information about additional actions related to an existing conversation, such as rescheduling dinner or ground transportation, without explicit user input, by determining events and selecting relevant actions based on user information and interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the assistant waits for explicit user requests to provide information or perform actions, then the system complexity remains low and operations are simple, but the usability deteriorates because users are not aware of all capabilities and must explicitly request each function
Solution Approach 1:
The assistant proactively determines and outputs information about additional actions that can be performed in furtherance of an existing conversation before the user explicitly requests them. This preliminary action enables users to discover capabilities they weren't aware of, improving usability without requiring complex interaction patterns
Solution Approach 2:
The assistant monitors the existing conversation context and provides feedback about relevant capabilities and actions that align with the user's current goals. This feedback mechanism helps users understand what the assistant can do without requiring them to explicitly ask, thereby improving ease of operation
2Productivity
If the assistant proactively outputs information about additional actions without explicit user input, then the productivity improves by reducing the number of user inputs required, but the device complexity increases due to the need for event determination and action selection mechanisms
Solution Approach 1:
The assistant performs preliminary determination of events and selection of relevant actions based on the existing conversation context before presenting information to the user. This advance preparation reduces the number of interaction turns needed, improving productivity while managing complexity through structured processing
Solution Approach 2:
The assistant autonomously monitors conversation state, determines relevant events, and selects appropriate actions without requiring continuous user guidance. This self-service capability improves task completion speed by reducing back-and-forth interactions, while the complexity is managed through automated decision-making processes
3Adaptability or versatility
If the assistant provides comprehensive information about all available actions, then the adaptability improves as users can discover more capabilities, but the loss of information increases due to information overload that may overwhelm users
Solution Approach 1:
The assistant filters and presents only those additional actions that are locally relevant to the existing conversation context and user goals, rather than listing all available capabilities. This selective approach enables capability discovery while preventing information overload by tailoring the information to the specific situation
Data Source
AI summary
A system is described that determines, based on information associated with a user of a computing device, an event for initiating an interaction between the user and an assistant executing at the computing device. The system selects, based on the event and from a plurality of actions performed by the assistant, at least one action associated with the event. The system determines, based on the at least one action, whether to output a notification of the event which includes an indication of the event and a request to perform the at least one action associated with the event. Responsive to determining to output the notification of the event, the system sends, to the assistant, the notification of the event for output during the interaction between the user and the assistant