Knowledge Graph Embedding for Gap Detection and Node Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional data search approaches in enterprises are limited by string matching mechanisms, leading to incomplete and inaccurate results due to 'dark data' and disparate data silos, which hinder efficient information sharing and access across units.
Innovation Solution
A knowledge graph system that structures data as a graph with semantic meaning, using embedding techniques to translate and enhance the graph, identifying gaps, and reconstructing relationships to create new nodes representing new combinations of information, thereby improving data retrieval and recommendation generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If string matching mechanisms are used for data search, then the search process is simple and fast, but the search results are incomplete and inaccurate due to dark data and data silos
Solution Approach 1:
The patent merges multiple disparate data stores from different enterprise units into a unified knowledge graph. The knowledge graph integration layer combines data from various sources, transforming isolated data silos into a connected semantic network that improves search accuracy while managing complexity through standardized integration processes
Solution Approach 2:
The knowledge graph acts as an intermediary layer between traditional data stores and query systems. It transforms unstructured or semi-structured data into structured knowledge with explicit relationships, enabling accurate semantic search without requiring changes to underlying data stores
2Adaptability or versatility
If data is stored in separate data stores managed by different units, then each unit can manage its data independently, but data becomes less accessible across units forming data silos
Solution Approach 1:
The system segments the knowledge graph into domain-specific subgraphs that correspond to different enterprise units. Each subgraph maintains its own data characteristics and management rules, allowing units to manage their data independently while the overall knowledge graph provides cross-unit accessibility through defined relationship pathways
Solution Approach 2:
The knowledge graph platform provides universal access mechanisms that work across all data sources. It implements standardized query interfaces and relationship models that enable different units to access and share data through a common framework, making the system versatile across diverse data types and organizational structures
3Reliability
If conventional knowledge query systems are used, then the system is simple to operate, but computing resources are consumed repeatedly and results are inaccurate or incomplete
Solution Approach 1:
The system performs preliminary processing by pre-computing and storing knowledge relationships, embeddings, and semantic connections in the knowledge graph during off-peak hours. This preliminary action transforms raw data into structured knowledge beforehand, so that actual queries can be executed efficiently with minimal resource consumption while returning complete and accurate results
Solution Approach 2:
The patent replaces traditional mechanical string-matching search mechanisms with semantic search based on knowledge graphs and vector embeddings. This substitution enables the system to understand query intent and retrieve relevant information based on meaning rather than exact string matches, improving result completeness while reducing the need for repeated computational searches
Data Source
AI summary
Knowledge graph systems are disclosed for enhancing a knowledge graph by generating a new node. The knowledge graph system converts a knowledge graph into an embedding space, and selects a region of interest from within the embedding space. The knowledge graph system further identifies, from the region of interest, one or more gap regions, and calculates a center for each gap region. A node is generated for each gap region, and the information represented by the node is added to the original knowledge graph to generate an updated knowledge graph.