Knowledge Graph Construction via Sub-Graph Fusion

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing information processing systems face challenges in efficiently capturing, analyzing, and presenting large volumes of data from multiple sources, which hinders the construction of comprehensive knowledge graphs.

Innovation Solution

The method involves collecting data from various sources, extracting structured and unstructured data using unsupervised machine learning, forming sub-graph structures for each data source, and combining them into a comprehensive knowledge graph structure.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If data from multiple sources is collected and integrated into a comprehensive knowledge graph, then the completeness and comprehensiveness of the knowledge graph is improved, but the complexity of data processing and system architecture increases

Engineering Contradiction:

Improvecomprehensiveness of knowledge graphVSAvoidsystem architecture complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent divides the knowledge graph construction process into distinct modules: data collection module, data processing module (with structured and unstructured data extraction), sub-graph generation module, and fusion module. Each module handles specific tasks independently, making the overall system more manageable despite processing multi-source data.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediate sub-graph structures as mediators between raw multi-source data and the final comprehensive knowledge graph. Each data source first generates its own sub-graph, which then serves as an intermediate representation that is easier to fuse than raw data, reducing the complexity of direct multi-source integration.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If unsupervised machine learning is used to extract unstructured data, then the ability to process diverse data formats is improved, but the computational resources and processing time required increase

Engineering Contradiction:

Improvedata format processing capabilityVSAvoidcomputational resource consumption

Core Design Contradiction:

Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

The data processing module is segmented into two parallel pathways: one for structured data extraction and another for unstructured data extraction using unsupervised machine learning. This segmentation allows the system to apply computationally intensive unsupervised learning only where necessary (for unstructured data) while using simpler methods for structured data, optimizing resource usage.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies unsupervised machine learning selectively only to unstructured data portions rather than processing all data through the same intensive method. This partial application of complex processing reduces overall computational resource consumption while maintaining the ability to handle diverse data formats.

Inventive Principle:

Principle #16Partial or excessive action

3Reliability

If sub-graph structures are formed for each data source and then combined, then the traceability and data quality are improved, but the processing time and computational overhead increase

Engineering Contradiction:

Improvedata traceabilityVSAvoidprocessing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent performs preliminary actions by generating sub-graph structures for each data source separately before fusion. This preliminary organization of data into source-specific sub-graphs establishes traceability early in the process and prepares data in a format that facilitates efficient fusion, reducing the time penalty of maintaining traceability.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

By segmenting the knowledge graph into source-specific sub-graphs, the patent enables parallel processing of individual sub-graphs before fusion. This segmentation maintains data traceability through clear source attribution while allowing computational operations to be performed independently and in parallel, reducing overall processing time.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS12265577B2Knowledge graph management based on multi-source data

Publication Date: 2025.04.01 EMC IP HLDG CO LLC

AI summary

Techniques for constructing and otherwise managing knowledge graphs in information processing system environments are disclosed. For example, a method comprises the following steps. The method collects data from a plurality of data sources. The method extracts structured data and unstructured data from the collected data, wherein unstructured data is extracted using an unsupervised machine learning process. The method forms a plurality of sub-graph structures comprising a sub-graph structure for each of the data sources based on at least a portion of the extracted structured data and unstructured data. The method combines the plurality of sub-graph structures to form a combined graph structure representing the collected data from the plurality of data sources. The resulting combined graph structure is a comprehensive knowledge graph.