Exploring semantic deep learning for building reliable and reusable one health knowledge from PubMed systematic reviews and veterinary clinical notes

Table 10 Locality-based modules extracted from the SNOMED CT ontology for the 11 well-known medical conditions

Target term (n-gram) lower case	Number of SNOMED CT concept identifiers for the signature	Total number of axioms	Number of OWL Classes	Number of OWL object properties	Number of SubClassOf axioms	Number of EquivalentClass axioms
anaemia	34	105,205	22,227	46	7092	15,134
arthritis	33	74,398	16,180	31	7470	8709
asthma	37	97,647	20,804	39	7611	13,192
ckd	19	51,890	10,929	40	3899	7029
diabetes	29	463,437	100,119	44	49,818	50,300
epilepsy	19	10,072	2085	23	1055	1029
glaucoma	39	101,997	21,278	44	11,047	10,230
heart_failure	37	68,006	14,634	44	5490	9143
hypertension	31	152,283	32,270	48	16,425	15,844
obesity	28	90,224	18,994	40	12,376	6617
osteoarthritis	38	73,496	15,903	37	8922	6980

The second column just reports the total number of SNOMED CT identifiers for the ontological signature. The worksheet “signatures” within the Additional file 3 contains the list of SNOMED CT identifiers (as signature) for each target term. From the third to the last column ontology metrics information for the locality-based module created per target term is provided. The last two columns indicate the number of descriptions and definitions extracted from the SNOMED CT ontology for each locality-based module, respectively

ISSN: 2041-1480