Virtual Assistant Adaptation Using Knowledge Graph Model Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital content systems suffer from operational inflexibility and inefficiency in selecting large language models, requiring excessive user interactions and computational resources due to separate interfaces for content items and model selection, leading to inaccurate and inefficient model selection processes.
Innovation Solution
A model modification system that utilizes a knowledge graph to determine relationships between user accounts, content items, and large language models, providing personalized recommendations based on contextual data within a content management system, reducing navigational efforts and preserving computational resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If existing systems provide a generic repository of all available large language models and wait for user selection, then users can access any model, but the system suffers from operational inflexibility and requires excessive user interactions to navigate through interfaces and lists of models
Solution Approach 1:
The system performs preliminary actions by automatically determining and recommending appropriate large language models based on user account data, content items, and contextual information before user selection. This eliminates the need for users to navigate through interfaces and lists of models, as the suitable model is already identified and presented to the user.
Solution Approach 2:
The system serves itself by automatically selecting and recommending large language models based on analyzed user account data, content items, and contextual factors. This self-service capability reduces reliance on manual user navigation and selection, thereby improving ease of operation and reducing navigational time.
2Productivity
If existing systems use separate interfaces for content items and for selecting/interacting with large language models, then functionality is separated, but the system becomes navigationally inefficient and consumes excessive computational resources for processing device interactions and caching data
Solution Approach 1:
The system merges the interface for content items with the interface for selecting and interacting with large language models into a unified interface. This integration eliminates the need for separate interfaces, reducing navigational inefficiency and minimizing computational resources required for processing device interactions and caching data across multiple interfaces.
Solution Approach 2:
The unified interface serves multiple functions: displaying content items, recommending large language models, and enabling interaction with selected models. This multi-functional interface improves system efficiency by consolidating operations that previously required separate interfaces, thereby reducing computational resource consumption.
3Measurement precision
If existing systems provide a flat, unguided interaction structure relying on user input and understanding of available models, then user autonomy is maintained, but model selection becomes inaccurate and improper selections occur
Solution Approach 1:
The system incorporates feedback mechanisms by analyzing user account data, content items, and contextual information to determine and recommend appropriate large language models. This feedback loop improves model selection accuracy by using actual user behavior and content characteristics rather than relying solely on user input and understanding.
Solution Approach 2:
The system changes parameters by dynamically determining model recommendations based on varying user account data, content items, and contextual factors. This parameter-based approach improves selection accuracy by adapting recommendations to specific user needs and content characteristics rather than using a static, flat interaction structure.
Data Source
AI summary
This disclosure describes systems that identify one or more models (e.g., large language models and/or virtual assistants) permitted to access content items stored for user accounts within a content management system. The disclosed systems can determine a model available to a user account within the content management system from among the one or more models. For example, the disclosed systems can determine one or more relationships between the user accounts within the content management system, large language models utilized by the user accounts, virtual assistants utilized by the user accounts, and content items accessed by the user accounts. The disclosed systems can determine the model for the user account according to the one or more relationships. The disclosed systems can provide a notification corresponding to the model via a user interface of a client device associated with the user account.