Skip to content

Concept Taxonomy for Modeling Healthcare Data with Graphs

This document defines the categorical taxonomy for organizing the 513 concepts in the learning graph. Each category has a TaxonomyID (3-5 letter abbreviation) and a description of the concepts it contains.

Taxonomy Categories

1. Foundation Concepts

TaxonomyID: FOUND

Description: Fundamental concepts in graph theory and database systems that serve as prerequisites for understanding healthcare graph modeling. Includes basic graph structures, database concepts, and data modeling principles.


2. Graph Technologies

TaxonomyID: GTECH

Description: Technical concepts related to graph databases, query languages, and graph database implementation, including Cypher, GQL, GSQL, indexing, scalability, distributed graph storage, and performance tuning.


3. Graph Analytics and Algorithms

TaxonomyID: ANAL

Description: Graph algorithms and analytical techniques including pathfinding, centrality measures, community/cluster detection, graph embeddings, and graph neural networks.


4. Healthcare Domain Fundamentals

TaxonomyID: HCARE

Description: Core healthcare system concepts, terminology, and organizational structures, including healthcare economics, care models, medical coding systems, and healthcare interoperability standards.


5. Patient Data and Clinical Concepts

TaxonomyID: PAT

Description: Patient-centric concepts covering clinical data, diagnoses, treatments, medications, and patient care management, including patient records, care plans, outcomes, and patient journey mapping.


6. Provider Operations

TaxonomyID: PROV

Description: Healthcare provider concepts including hospitals, clinics, physicians, scheduling, referrals, and provider networks, covering credentials, performance metrics, and care team coordination.


7. Payer and Insurance

TaxonomyID: PAYER

Description: Insurance and payer-related concepts including claims processing, policies, coverage, benefits, and reimbursement, including pharmacy benefit management and plan types.


8. Financial and Business Operations

TaxonomyID: FIN

Description: Healthcare business and financial concepts covering revenue cycles, cost analysis, profitability, contracts, and value-based payment models.


9. Fraud, Waste, and Abuse

TaxonomyID: FRAUD

Description: Concepts related to detecting and preventing fraud, waste, and abuse in healthcare, including fraud detection techniques, anomaly detection, and specific fraud patterns.


10. AI and Machine Learning

TaxonomyID: AI

Description: Artificial intelligence and machine learning concepts applied to healthcare, including LLMs, vector stores, token efficiency, retrieval-augmented generation, context graphs, enterprise knowledge graphs, the enterprise nervous system concept, agentic workflows, and responsible AI practices.


11. Clinical Decision Support, FHIR and CQL

TaxonomyID: CDS

Description: Modeling of clinical decision support systems using the HL7 FHIR standard, including FHIR resources, the Four Levels of Knowledge Representation (Narrative, Semi-Structured, Structured, Executable), the three Tiers of Functionality (Data, Logic, Forms/UI), Clinical Quality Language (CQL) and the Expression Logical Model (ELM), Clinical Quality Measures (CQMs), CDS Hooks, and the CMS-sponsored tooling (MADiE, CQL Runner, Bonnie, Cypress) used to author, test, and certify CQL-based measures.


12. Security and Privacy

TaxonomyID: SEC

Description: Healthcare data security and privacy concepts including HIPAA compliance, access control, authentication, authorization, encryption, and de-identification of protected health information.


13. Data Governance

TaxonomyID: GOV

Description: Data management, governance, and quality concepts including metadata management, data lineage, provenance, traceability, and explainability, along with master data management and data stewardship.


14. Capstone and Career

TaxonomyID: CAP

Description: Final capstone projects, presentations, career development, and real-world implementation concepts, representing the culmination of learning and application to practical scenarios.


Distribution Guidelines

Each category should ideally contain:

  • Minimum: 5-10 concepts
  • Target: 10-20 concepts
  • Maximum: 30% of total concepts

No category in this graph exceeds 12% of the total 513 concepts.


Generated for Modeling Healthcare Data with Graphs course