Virtual Agent Social Graph Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtual assistants lack the ability to take initiative and coordinate with other users or assistants without user input, and they often have distinct voices and personalities that may not align with the user they are representing, which can be undesirable in proxy scenarios.
Innovation Solution
Virtual agents that can access user information to perform tasks autonomously and adapt their voice and behavior to resemble the user, allowing them to interact with other users or agents on behalf of the user, while maintaining privacy controls and personalization based on social network data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If virtual assistants perform predefined tasks with distinct voices and personalities, then they can provide specialized services, but they cannot take initiative or coordinate with other users without user input
Solution Approach 1:
The virtual agent accesses and utilizes the user's social graph information, communication patterns, and preferences stored in the system to autonomously perform tasks without requiring continuous user input. The agent serves itself by leveraging pre-stored user data to make decisions and execute actions independently.
Solution Approach 2:
The system pre-processes and stores user information including social graph data, communication patterns, and preferences before tasks need to be executed. This preliminary preparation enables the virtual agent to act autonomously when tasks arise, without needing real-time user input.
2Adaptability or versatility
If virtual assistants have distinct voices and personalities, then they can provide specialized services, but their voice may not align with the user they are representing in proxy scenarios
Solution Approach 1:
The virtual agent dynamically adjusts its voice characteristics by modifying parameters such as pitch, tone, and speech patterns based on the user's communication style extracted from social graph data. This allows the agent to adapt its voice to match different users while maintaining a functional personalization system.
Solution Approach 2:
The system creates a virtual representation of the user's communication style by analyzing and copying patterns from the user's social interactions. The virtual agent then uses this copied communication pattern to generate responses that align with the user's typical voice and mannerisms.
3Productivity
If virtual agents access user information to perform tasks autonomously, then they can reduce user input requirements, but user privacy may be compromised
Solution Approach 1:
The system introduces a privacy management layer that acts as an intermediary between the virtual agent and user data. This mediator controls and regulates access to social graph information, ensuring that only necessary and appropriate data is used for task execution while maintaining user privacy protections.
Solution Approach 2:
The system applies different levels of privacy protection to different types of user information. Sensitive data receives higher protection while less sensitive data used for task execution receives appropriate but lighter protection, creating a differentiated privacy approach that enables productivity while managing risks.
Data Source
AI summary
Exemplary embodiments relate to improvements in digital assistants incorporating personalization based on social network data. Various aspects of the agent, such as the agent's voice, language style, and avatar may be personalized. Personalization may be applied to components of an agent's architecture (e.g., the virtual agent's language model, natural language generator, voice generation component, etc.). Moreover, by interfacing with the social network's social graph, the agent may be provided with information useful to performing certain tasks (e.g., a calendar for scheduling, food preferences for ordering tasks, etc.). An agent may be provided (and personalized) for a single user, or a group of users (e.g., a family). The agent can be personalized to anyone, which may allow (e.g.) for the agent to represent a celebrity or a person who is not currently available in interactions with others. Different agents can talk to each other, e.g. for purposes of scheduling meetings.