Fig 1.
Five-layer architecture diagram for knowledge graph-based library intelligent data management.
The architectural structure consists of five layers (color-coded): Data Layer (light green) for integration of multiple data sources, Data Processing Layer (dark green) for ETL processes, Knowledge Graph Layer (gold) for ontology-based knowledge development; Intelligent Service Layer (purple) for AI-based services, Application Layer (coral) for user interfaces. Black arrows denote data flow; purple demonstrates feedback for improvement.
Fig 2.
System operation example: Illustration of personalized recommendation service scenarios.
The graph shows personalized recommendation processes designed for a graduate student (top path) and an undergraduate (bottom path) having varying levels of expertise because of differences in their academic backgrounds. The books (dark green), journals (blue), and attributes (orange ovals) are interlinked through using semantic links (edge labels). The recommended items have confidence scores (88%−96%) calculated through using paths in the knowledge graph. The recommendations vary depending upon historical usage, level of expertise, as well as research interest of the user.
Table 1.
System evaluation index system and measurement methods.
Table 2.
User experience feedback data.
Fig 3.
Comprehensive comparison of multi-dimensional evaluation results.
Fig 4.
Performance comparison between traditional systems and knowledge graph systems.
Table 3.
Analysis of system performance and technical indicator achievement.