Virtual Assistant Notification Context Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Virtual assistant systems face challenges in maintaining conversational context when proactively triggering notifications, leading to unintelligible conversations and errors due to changing contexts and active conversational parts, especially when users ignore or revisit notifications.
Innovation Solution
A system and method for a virtual assistant that proactively generates notifications with a clean conversational context, allowing re-evaluation when the user consumes the notification, ensuring the correct conversational part is active to understand user responses, and using a history log memory to manage and update notifications effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If the virtual assistant system proactively triggers notifications proactively, then the system can provide timely information to users, but the conversational context may change or become unintelligible by the time the notification is delivered
Solution Approach 1:
The system performs preliminary actions by generating notifications with embedded conversational context before the user actually consumes the notification. The notification includes a close-ended type question and relevant context information is prepared in advance, so when the user responds, the conversational part is already activated and ready to understand the user's response without requiring re-evaluation of the entire conversation history.
Solution Approach 2:
The system implements feedback mechanisms where the terminal client detects when the user consumes the notification and sends a retransmission request to the terminal server. The server then re-evaluates the notification with updated context and sends a new version, ensuring the conversational context is current and accurate at the time of user interaction.
2Measurement precision
If the notification is generated with current context, then the notification is accurate, but the conversational part may not be active when the user responds
Solution Approach 1:
The notification is generated with the conversational context embedded in advance, including the close-ended type question and relevant information. This preliminary preparation ensures that when the user responds, the context is already in place, eliminating the need for re-evaluation and ensuring the conversational part is activated and ready to handle the user's response accurately.
Solution Approach 2:
The system uses feedback from the terminal client to trigger retransmission requests when the user consumes the notification. This feedback loop allows the terminal server to re-evaluate and send updated notifications, ensuring both accuracy and reliability of the conversational context at the time of user interaction.
3Loss of information
If the system re-evaluates notifications upon user response, then the context is updated and accurate, but additional processing time and resources are required
Solution Approach 1:
The system performs re-evaluation of notifications in advance, before the user actually responds. The terminal server prepares updated notifications with current context and stores them, so when the user consumes the notification and sends a retransmission request, the updated version is already ready to be sent immediately, minimizing processing time.
Solution Approach 2:
The system implements periodic re-evaluation of notifications at predetermined intervals or triggered by specific events (such as user consumption). This periodic action ensures context remains accurate without requiring continuous re-evaluation, balancing accuracy with processing efficiency.
4Quantity of substance
If multiple notifications are sent to users, then comprehensive information is provided, but the conversational context becomes complex and difficult to manage
Solution Approach 1:
The system segments the notification delivery process into distinct components: the terminal server generates notifications with embedded context, the terminal client manages delivery and user interaction, and the history log memory stores contextual information. This segmentation allows multiple notifications to be managed independently, reducing the complexity of context management even when multiple notifications are sent.
Solution Approach 2:
The history log memory acts as an intermediary that stores and manages conversational context for multiple notifications. This intermediary component handles the complexity of context management by maintaining separate context records for each notification, allowing the system to send multiple notifications while keeping context organization simple and manageable.
Data Source
AI summary
A terminal server of a virtual assistant system for proactively triggering notifications is disclosed. The terminal server is configured to: receive data indicative of a change of a service related state associated with a user of at least one terminal client; generate accordingly a close-ended type question; instruct a transmission of the close-ended type question to the at least one terminal client; in response to a retransmission request, received from the at least one terminal client in relation to the transmission: not perform the close-ended type question, access a storage of the service related state to generate accordingly a new close-ended type question, instruct a transmission of the new close-ended type question to the at least one terminal client, analyze a closed type answer provided by the at least one terminal client, and instruct transmission of a current response to the answer provided by the user.