Multi-Toolkit Enterprise Knowledge Graph Mining
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Enterprises face challenges in creating and maintaining accurate and up-to-date knowledge graphs due to unique vocabularies, private project names, and non-standard use of words, leading to inaccurate search results and inefficient use of time and resources, especially when dealing with large datasets.
Innovation Solution
The implementation of a multi-toolkit enterprise mining system that uses neural entity recognition and Bayesian inference techniques to mine and link entities across multiple sources, including internal and external data sources, to generate and update a knowledge graph, ensuring accurate and comprehensive information retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single toolkit is used for knowledge graph mining, then the system is simple to manage, but the coverage and flexibility of information extraction is limited
Solution Approach 1:
The system divides the knowledge graph mining task into multiple specialized toolkits, each responsible for extracting specific types of entities (e.g., projects, companies, products, users) from different data sources. This segmentation allows each toolkit to specialize in particular entity types while collectively providing comprehensive coverage across all enterprise knowledge domains.
Solution Approach 2:
The system creates a universal framework that can handle multiple types of entities and data sources through a common architecture. The framework provides unified entity linking, disambiguation, and knowledge graph construction capabilities that work across all specialized toolkits, enabling the system to be both diverse in functionality and consistent in operation.
2Adaptability or versatility
If multiple toolkits are used for knowledge graph mining, then the coverage and flexibility of information extraction is improved, but the system complexity increases
Solution Approach 1:
The system merges the outputs from multiple specialized toolkits into a unified knowledge graph through a central entity linking and disambiguation module. This module consolidates entity mentions across different toolkits, resolves ambiguities, and creates a cohesive knowledge structure that integrates information from all sources while managing complexity through centralized coordination.
Solution Approach 2:
The system introduces intermediary components including entity linking services, disambiguation modules, and permission management layers that mediate between the multiple specialized toolkits and the final knowledge graph output. These intermediaries handle the complexity of coordinating multiple toolkits, managing entity identity resolution, and enforcing access controls without requiring changes to the individual toolkits themselves.
3Reliability
If proprietary information is stored in the knowledge graph, then the enterprise knowledge is preserved, but the ability to share information with third parties is limited
Solution Approach 1:
The system applies different permission and access control settings to different portions of the knowledge graph based on sensitivity and ownership. Highly proprietary information receives stricter access controls, while less sensitive information can be shared more broadly. This local differentiation of access policies allows the system to maintain accuracy and reliability for proprietary data while enabling selective sharing of appropriate information with third parties.
4Reliability
If content is modified according to permissions, then the security and compliance is improved, but the usability and user experience is worsened
Solution Approach 1:
The system automatically applies permission rules and content modifications without requiring manual user intervention. The permission management system self-services by automatically filtering and modifying content based on user roles, permissions, and compliance rules, thereby maintaining security and compliance while preserving a seamless user experience without obvious restrictions or manual approval processes.
Data Source
AI summary
Examples described herein generally relate to a computer system including a knowledge graph storing a plurality of entities. A mining of a set of enterprise source documents within an enterprise intranet is performed, by a plurality of knowledge mining toolkits, to determine a plurality of entity names. The plurality of entity names are linked based on entity metadata by traversing various relationships between people, files, sites, groups, associated with entities. An entity record is generated within a knowledge graph for a mined entity name from the linked entity names based on an entity schema and ones of the set of enterprise source documents associated with the mined entity name. The entity record includes attributes aggregated from the ones of the set of enterprise source documents associated with the mined entity name.