CS-KG 2.0: A Large-scale Knowledge Graph of Computer Science

Area of Science:

• Computer Science
• Artificial Intelligence
• Information Science

Background:

• The rapid evolution of Artificial Intelligence (AI) and open access publishing presents both opportunities and challenges in scientific research.
• The sheer volume of published papers annually overwhelms traditional literature review and analysis methods.
• Knowledge graphs offer a structured approach to integrate and formalize information, enhancing AI's ability to process scientific literature.

Purpose of the Study:

• To present the latest version of the Computer Science Knowledge Graph (CS-KG 2.0).
• To provide a structured and interconnected knowledge base for AI systems to process scientific literature.
• To facilitate advanced research applications by organizing vast amounts of computer science research data.

Main Methods:

• Generation of an extensive knowledge base from 15 million research papers in computer science.
• Description of 25 million entities and 67 million relationships within the graph.
• Utilizing AI and knowledge graph technologies for information integration and representation.

Main Results:

• The CS-KG 2.0 contains a detailed representation of scientific knowledge in computer science.
• The graph comprises 25 million entities interconnected by 67 million relationships.
• The resource is designed for AI-driven analysis and exploration of the scientific literature.

Conclusions:

• CS-KG 2.0 offers a powerful resource for managing and analyzing the growing body of computer science literature.
• This knowledge graph enhances AI capabilities for tasks like trend analysis, hypothesis generation, and literature review automation.
• The structured data facilitates novel research opportunities and improves scientific question-answering systems.