Enterprise Knowledge Graph Curation via Intranet Mining
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Proprietary knowledge graphs for enterprises face challenges in creating accurate, up-to-date, and complete knowledge stores while maintaining confidentiality, as they need to manage sensitive information without sharing it with third parties.
Innovation Solution
A computer system and method for generating and managing an enterprise knowledge graph by mining enterprise source documents within an intranet to identify entity names, generating entity records based on schemas, allowing user curation, and displaying attributes based on permissions, combining machine-learned and user-curated knowledge.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If enterprise source documents are mined to generate entity records automatically, then productivity and completeness of the knowledge graph improve, but the risk of information leakage and loss of control over confidential data increases
Solution Approach 1:
The system segments the knowledge graph construction process into automated mining operations and manual curation stages. The processor automatically mines entity names and generates entity records from enterprise source documents, then presents these records to users for review and curation. This segmentation allows high-productivity automated processing while maintaining control through staged human oversight, preventing premature exposure of confidential information.
Solution Approach 2:
The system introduces an intermediary curation layer between automated document mining and final knowledge graph deployment. Users act as intermediaries who review mined entity records, verify their accuracy, and control what information is finalized in the knowledge graph. This intermediary stage prevents direct exposure of confidential source documents while maintaining data quality and security.
2Reliability
If users are given access to curate entity records, then accuracy and reliability of the knowledge graph improve, but system complexity and operational overhead increase
Solution Approach 1:
The system enables users to perform self-service curation of entity records through an intuitive interface. Users can review mined entity records, make corrections, add missing information, and validate data accuracy without requiring complex administrative procedures. This self-service approach improves reliability through user expertise while keeping operational complexity manageable through automated assistance.
Solution Approach 2:
The system implements feedback loops where users curate entity records and their corrections are fed back into the mining process. The processor learns from user curation actions to improve future automated entity extraction accuracy. This feedback mechanism enhances knowledge graph reliability over time while reducing the need for manual intervention in subsequent mining operations.
3Loss of information
If all mined entity records are displayed to users, then completeness of information improves, but access control and permission management become more difficult
Solution Approach 1:
The system applies local quality control by displaying different subsets of mined entity records to different users based on their permissions and roles. Each user sees only the entity records relevant to their responsibilities and security clearance level. This approach maintains information completeness for each user's context while simplifying permission management through role-based access control.
Solution Approach 2:
The system enhances entity records with localized metadata indicating their sensitivity level, source document classification, and applicable access permissions. This metadata allows the display mechanism to automatically filter and present appropriate records to each user without requiring complex manual permission management, maintaining both information completeness and security.
Data Source
AI summary
Examples described herein generally relate to a computer system for generating a knowledge graph storing a plurality of entities and to displaying a topic page for an entity in the knowledge graph. The computer system performs a mining of source documents within an enterprise intranet to determine a plurality of entity names. The computer system generates an entity record within the knowledge graph for a mined entity name based on an entity schema and the source documents. The entity record includes attributes aggregated from the source documents. The computer system receives a curation action on the entity record from a first user. The computer system updates the entity record based on the curation action. The computer system displays an entity page including at least a portion of the attributes to a second user based on permissions of the second user to view the source documents.