An AI Workflow Combining Bidirectional Encoder Representations from Transformers (BERT) and Graph Neural Networks...

Area of Science:

• Artificial Intelligence
• Knowledge Management
• Data Science

Background:

• Enterprise knowledge management (KM) systems struggle with large volumes of unstructured data, hindering information extraction and decision-making.
• Inefficient knowledge sharing and delayed decisions are common issues in current KM systems.
• Existing methods lack the ability to effectively process and structure diverse enterprise data.

Purpose of the Study:

• To propose a unified artificial intelligence-driven framework to enhance enterprise knowledge management.
• To improve the extraction of relevant information from unstructured organizational data.
• To accelerate decision-making processes and knowledge sharing.

Main Methods:

• A hybrid AI framework combining Graph Neural Networks (GNNs) for ontology alignment and semantic reasoning.
• Utilizing refined Bidirectional Encoder Representations from Transformers (BERT) for domain-specific entity and relation extraction.
• Systematic data collection, preprocessing, knowledge graph construction, and GNN-based ontology alignment.

Main Results:

• A 35% decrease in decision-making latency and a 21% gain in knowledge retrieval precision compared to baseline methods.
• Experimental validation across two industry applications demonstrated significant improvements.
• User feedback highlighted increased satisfaction due to semantic search and contextual tagging features.

Conclusions:

• The proposed framework enables reproducible knowledge graph building from unstructured enterprise data.
• It effectively fuses graph-based reasoning with deep learning for information extraction.
• The study concludes that aligned knowledge representations improve both strategic and operational KM outcomes, offering a scalable solution.