Omnichannel Virtual Agent Using Knowledge Graph
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conversational agents lack the ability to effectively communicate and understand user intents across multiple channels and domains, limiting their ability to provide comprehensive and context-aware responses.
Innovation Solution
An omnichannel, intelligent, proactive virtual agent system that utilizes a world model represented as a knowledge graph to facilitate semantic understanding and response generation, integrating with IoT systems and enabling natural language speech processing for human interaction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a conversational agent uses traditional communication methods, then the implementation is simple, but the ability to understand user intents across multiple channels and domains is limited
Solution Approach 1:
The patent introduces a knowledge graph as an intermediary layer between the conversational agent and enterprise data sources. This knowledge graph serves as a mediator that translates diverse user queries across multiple channels into unified semantic representations, enabling the agent to understand intents accurately without directly complexifying the communication interfaces themselves.
Solution Approach 2:
The conversational agent is designed with multi-functionality to operate across multiple communication channels (voice, text, graphical user interface) and domains simultaneously. The agent can process queries from various channels through a unified architecture that leverages the knowledge graph for semantic understanding, making the system universally applicable across different interaction modes without requiring separate specialized systems for each channel.
2Measurement precision
If the agent provides comprehensive and context-aware responses, then the quality of user interaction improves, but the processing time and computational resources increase
Solution Approach 1:
The system performs preliminary actions by pre-processing and indexing enterprise data into the knowledge graph before actual user queries arrive. This pre-organization of information in semantic relationships allows the agent to quickly retrieve relevant context during real-time interactions, achieving comprehensive and context-aware responses without significant processing delays.
Solution Approach 2:
The patent extracts only the necessary contextual information from the vast enterprise data sources and represents it in the knowledge graph. By selectively extracting and pre-organizing only the relevant semantic relationships and facts needed for context-aware responses, the system reduces the computational burden during query processing while maintaining high accuracy in understanding user intents and providing appropriate context.
3Adaptability or versatility
If the system integrates with multiple enterprise systems and data sources, then the versatility and data accessibility improve, but the system integration complexity increases
Solution Approach 1:
The knowledge graph functions as a universal intermediary layer that sits between the conversational agent and multiple enterprise system data sources. Rather than requiring the agent to directly integrate with each individual system, the knowledge graph mediates by providing a unified semantic representation that can be queried consistently across all data sources, thereby reducing integration complexity while maintaining broad data accessibility.
Solution Approach 2:
The system segments the complex integration task by separating data access from query processing. Different enterprise systems can be independently connected to the knowledge graph using standardized protocols, and once integrated, all data is represented uniformly in the graph. This segmentation allows new data sources to be added without affecting the core agent architecture, reducing overall integration complexity while enhancing versatility.
Data Source
AI summary
An omni-channel, intelligent, proactive virtual agent system and method of use are provided by which a user may engage in a conversation with the agent to interact with structured and unstructured data of an enterprise that is stored in a domain-specific world model for the enterprise.