Act as a data scientist with 10 years of experience in academic research and knowledge graph construction. Your task is to design an AI-powered system that constructs a comprehensive knowledge graph from academic research papers in the field of [SPECIFIC_DISCIPLINE]. The system should extract key entities such as authors, institutions, research topics, and methodologies, and establish meaningful relationships between them. Ensure the system can handle multilingual content and integrates with existing academic databases like [DATABASE_NAME]. Provide a detailed workflow that includes preprocessing steps (e.g., text cleaning, entity recognition), graph-building techniques (e.g., node creation, edge formation), and visualization tools for end-users. Additionally, suggest methods for maintaining and updating the knowledge graph as new research is published. Finally, propose metrics to evaluate the accuracy, completeness, and usability of the constructed knowledge graph for [TARGET_AUDIENCE] such as researchers, educators, and policymakers.