Tackling the challenges of matching biomedical ontologies

Table 1 Overview of the ontology matching systems that participated in OAEI biomedical tracks

System	Size	Lexicon	Relations	Repair	Background knowledge	OAEI Bio tracks
AgrMaker [19]	+	weights	part of	-	Bio; Man; Med	A
AML [11]	+++	WN; weights	all	Logic	Bio; Auto; M/E	all
Anchor-Flood [40]	+	WN	-	-	Man; Exp	A
Aroma [41]	+++	-	-	-	-	A, LB
ASMOV [42]	+	WN	-	-	Bio; Man; Exp	A
AUTOMSv2[43]	++	WN; weights	-	-	Man; Exp	LB-
BLOOMS [20]	+	WN	part of	-	Bio; Man; Med	A
COMMAND [44]	+	-	-	Logic	-	A
CroMatcher [45]	+	WN	-	-	Man; Exp	A
DKP-AOM [46]	+	WN	-	-	Man; Exp	A, LB-
DSSim [47]	+	WN	-	-	Man; Exp	A
FCA-Map [16]	++	UMLS	-	Logic	Man; Exp	all-
GMap [48]	+	external?	-	Logic	Man; Exp	A
GOMMA [49]	+++	-	-	-	Bio; Auto; Med	A, LB
kosimap [50]	+	-	-	-	-	A
LogMap [13]	+++	WN; UMLS	-	Logic	Bio; Auto; M/E	all
LP HOM [51]	+	-	-	-		A
Lyam++[52]	++	BabelNet	-	-	Man; Exp	all-
MapPSO [53]	+	-	-	-	-	A
OACAS [54]	+	-	-	-	-	A
PhenomeNET [21]	++	(AML)	part of	-	Bio; Man; Med	DP
SAMBO [22]	+	WN	part of	-	Bio; Man; Exp	A
ServOMap [55]	+++	WN	-	Logic	Man; Exp	A, LB
TaxoMap [56]	+	-	-	-	-	A
TOAST [57]	+	-	-	-	-	A
WikiMatch [58]	++	Wikipedia	-	-	Man; Exp	A, LB-
YAM++[59]	+++	WN	-	Rules	Man; Exp	A, LB

Size details the systems’ capability of handling medium-sized (+), large (++) or very large (+++) ontologies; Lexicon lists their use of lexical tools such as WordNet (WN) or the UMLS SPECIALIST Lexicon (UMLS), as well as whether they use synonym weights; Relations lists the types of relations they contemplate in addition to subclass relations; Repair details whether they perform alignment repair based on logic or rules; Background Knowledge describes whether they use biomedical ontologies as background knowledge (Bio), whether the process of background knowledge selection is manual (Man) or automatic (Auto), and whether background knowledge is used as a mediator (Med) or for lexical expansion (Exp); OAEI Bio Tracks lists the tracks in which the system successfully competed in out of Anatomy (A), Large Biomedical Ontologies (LB), and Disease & Phenotype (DP), with - indicating that the system did not complete the largest LB tasks

ISSN: 2041-1480