Table 2.

Table 2 Summary of Terminology Network Properties and Their Descriptions as Reported in the UMLS Documentation(24)

Source Abbreviation, 2007 AA	Official Name	Category	Statistical Classification	Scale-Free	At least 95% Hierarchical	Small-World
GO	Gene Ontology, 2005	Gene names	no	yes	yes	yes
HL7V3.0	Health Level Seven Vocabulary, 1998–2006	Miscellaneous	no	yes	yes	yes
NCBI	NCBI Taxonomy, 2006	Gene names	no	yes	yes	no
CPT	Current Procedural Terminology, 4th ed., 2006	Procedures only	no	yes	no	no
RCD	Clinical Terms Version 3 (Read Codes), 1999	“Comprehensive” clinical vocabularies	no	yes	no	yes
SNOMEDCT	SNOMED Clinical Terms, 2006	“Comprehensive” clinical vocabularies	no	yes	no	yes
SNMI	Systematized Nomenclature of Human and Veterinary Medicine, 1998	“Comprehensive” clinical vocabularies	no	yes	no	yes
NCI	NCI Thesaurus, 2006	Diseases	no	yes	no	yes
MSH	Medical Subject Headings, 2007	Organisms ∗	no	yes	no	yes
LNC	Logical Observation Identifier Names and Codes, 2.17,2006	Thesaurus (used for indexing and retrieval of biomedical literature)†	no	no	no	yes
NOC	Nursing Outcomes Classification, 1997	Nursing (currently used primarily for clinical documentation and research)	yes	no	no	no
DSM4	Diagnostic and Statistical Manual of Mental Disorders. 4th ed., 1994	Diagnoses/clinical problems/signs and symptoms	yes	no	yes	no
ICD10	International Statistical Classification of Diseases and Related Health Problems, 1998	Diagnoses only	yes	no	yes	no
ICD9CM	International Classification of Diseases, 9th Revision, Clinical Modification, 2007	Diagnoses and procedures	yes	no	yes	no
NIC	Nursing Interventions Classification, 2005	Nursing (currently used primarily for clinical documentation and research)	yes	no	yes	no
ICPC	International Classification of Primary Care, 1993	Diagnoses/clinical problems/signs and symptoms	yes	no	yes	no

^∗ In UMLS documentation, ²⁴ NCBI was listed under the category gene names.

^† In UMLS documentation, ²⁴ MSH was listed under the category “Comprehensive” vocabularies.

Scale-free networks, which are characterized by a power law distribution in average node degree, have a small number of highly-connected hubs. Small-world networks have a short average path length and strong clustering at the neighborhood level.