Construct a Knowledge Graph for China Coronavirus (COVID-19) Patient Information Tracking

Area of Science:

• Epidemiology
• Data Science
• Network Analysis

Background:

• The Coronavirus disease (COVID-19) pandemic caused significant global losses.
• Conventional epidemiological studies often focus on macro-level transmission trends.
• In-depth analysis of patient data is crucial for improving social isolation and public health interventions.

Purpose of the Study:

• To propose a method for analyzing the COVID-19 epidemic situation using knowledge graphs and interactive visual analysis.
• To develop a structured knowledge system for tracking confirmed patients and their contacts.
• To enhance understanding of disease transmission dynamics and identify key factors.

Main Methods:

• Constructed a knowledge graph of COVID-19 patient activity tracks.
• Utilized Conditional Random Field (CRF) algorithm for entity and knowledge extraction.
• Employed PageRank and Label Propagation algorithms for community detection in patient networks.
• Integrated semantic relationship analysis and deep learning for model verification.
• Applied multi-view visualizations including map distribution, knowledge graph, and track visualization.

Main Results:

• Demonstrated the feasibility of constructing a knowledge graph for COVID-19 patient activity.
• Successfully analyzed epidemic situations using confirmed patient data and multi-view interactions.
• Identified key nodes and traced activity tracks effectively through the knowledge graph.
• The proposed method provides a feasible approach for understanding disease transmission processes.

Conclusions:

• Knowledge graphs combined with interactive visualization offer a powerful tool for analyzing infectious disease outbreaks like COVID-19.
• This approach facilitates in-depth data mining, patient tracking, and understanding of transmission dynamics.
• The findings support the use of knowledge graphs for enhanced public health strategies and interventions during epidemics.