Knowledge Graph Construction for Enterprise Knowledge Retrieval

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Enterprise knowledge is scattered and dependent on employee experience, leading to inefficiencies in knowledge retrieval and collaboration, resulting in low office efficiency.

Innovation Solution

A method and apparatus for constructing a knowledge graph by acquiring user-related information and knowledge data, determining entities, metadata, and relationships, and outputting a comprehensive knowledge base for smart working.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If enterprise knowledge is stored in traditional scattered formats dependent on employee experience, then knowledge retention is maintained through personal accumulation, but knowledge retrieval efficiency deteriorates and collaboration becomes difficult

Engineering Contradiction:
Improveknowledge retentionVSAvoidknowledge retrieval time
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The patent merges scattered enterprise knowledge from multiple sources (documents, databases, employee expertise) into a unified knowledge graph structure. This integration consolidates previously distributed knowledge into a centralized system that maintains comprehensive knowledge retention while enabling efficient retrieval through structured relationships and semantic connections between knowledge elements.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The knowledge graph serves as an intermediary layer between raw enterprise data and user knowledge queries. It transforms unstructured or semi-structured knowledge into a standardized graph structure with entities, relationships, and attributes, enabling efficient querying and retrieval without requiring direct access to scattered source systems.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Device complexity

If enterprise knowledge is organized in traditional scattered formats, then implementation complexity is reduced, but knowledge completeness and comprehensiveness deteriorate

Engineering Contradiction:
Improvesystem implementation complexityVSAvoidknowledge completeness
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The patent segments enterprise knowledge into discrete knowledge elements (entities, relationships, attributes) that can be independently extracted, stored, and managed. This segmentation allows the system to build comprehensive knowledge coverage by systematically processing individual knowledge units from various sources while maintaining manageable implementation complexity through modular processing steps.

Inventive Principle:
Principle #1Segmentation

3Reliability

If employees rely on personal experience for knowledge management, then knowledge accuracy is maintained through individual expertise, but collaboration efficiency deteriorates and knowledge sharing becomes difficult

Engineering Contradiction:
Improveknowledge accuracyVSAvoidcollaboration efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The knowledge graph system incorporates feedback mechanisms where employee interactions, corrections, and contributions continuously refine and validate knowledge accuracy. The system maintains reliability by allowing expert employees to verify and update knowledge elements while simultaneously making this validated knowledge available to enhance overall collaboration efficiency.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3822875B1Method and apparatus for outputting information, device, storage medium, and computer program product
Publication Date: 2026.01.28 BEIJING BAIDU NETCOM SCI & TECH CO LTD
  • EP3822875B1 patent drawingFigure 1
  • EP3822875B1 patent drawingFigure 2
  • EP3822875B1 patent drawingFigure 3

AI summary

A method and apparatus for outputting information, a device, a storage medium, and a computer program product are provided. An implementation of the method may include: acquiring knowledge data from at least one data source; acquiring information related to at least one user and a preset entity set; determining an entity, metadata of the entity and a relationship between entities, based on the related information, the knowledge data and the preset entity set; creating a knowledge graph based on the entity, the metadata and the relationship; and outputting the knowledge graph.