User-Specific Knowledge Graph for Context-Aware Electronic Services
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic communication often requires repetitive input from users, is disruptive, and lacks flexibility, particularly in scenarios involving language barriers and non-standardized addresses, leading to frustration and inefficiencies in delivery and hospitality services.
Innovation Solution
A user-specific knowledge graph is created using a virtual assistant to identify communication parties, transcribe and recognize named entities, determine contexts, and embed communications and contexts in an LLM, allowing for context-aware responses without user input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If electronic communication requires input and attention from all communicating parties, then communication can be established, but it becomes disruptive and repetitive
Solution Approach 1:
The system performs preliminary actions by automatically transcribing communications, identifying named entities, and building knowledge graphs without requiring user input. This preliminary processing enables the system to retrieve and utilize contextual information proactively, eliminating the need for repetitive user responses in subsequent communications.
Solution Approach 2:
The patent introduces an intermediary system comprising automatic speech recognition modules, named entity recognition modules, and knowledge graph databases that mediate between communication parties. This intermediary automatically processes and stores communication data, allowing the system to answer queries without direct user involvement and reducing repetitive interactions.
2Speed
If prerecorded responses are used, then response time is reduced, but the responses become inflexible and not targeted to specific incoming communications
Solution Approach 1:
The system dynamically changes parameters by retrieving specific contextual information from knowledge graphs based on the incoming communication content. Instead of using fixed prerecorded responses, the system adjusts its responses by querying stored communications, named entities, and contextual data, enabling both fast response times and high adaptability to specific situations.
Solution Approach 2:
The system implements feedback mechanisms by continuously querying the knowledge graph with information from incoming communications and adjusting responses based on retrieved contextual data. This feedback loop ensures responses are both timely and specifically tailored to the current communication context rather than being generic prerecorded messages.
3Measurement precision
If automatic speech recognition and named entity recognition are implemented, then communication accuracy is improved, but system complexity increases
Solution Approach 1:
The patent segments the complex processing system into distinct functional modules: automatic speech recognition modules for transcription, named entity recognition modules for identifying key information, and knowledge graph databases for storage. This segmentation allows each module to specialize in specific tasks, improving overall accuracy while making the system more manageable and maintainable despite its complexity.
Data Source
AI summary
In one embodiment, a method includes accessing a communication between a user and an entity and identifying (1) each party to the communication and (2) one or more portions of the communication made by each identified party. The method further includes determining one or more contexts associated with the communication; generating an embedding of (1) the communication and (2) the one or more contexts in an embedding space; and updating a knowledge graph specific to the user, and accessible by an LLM associated with the user, with the embedding.