Table 1. Experimental results: Baseline vs. Case sensitive vs. Information gain.

Ontology	Baseline			Case sensitive (CS)			Information Gain (IG)			CS + IG
	P	R	F1	P	R	F1	P	R	F1	P	R	F1
CL	84.95	72.22	78.07	84.95	72.22	78.07	83.02	71.67	76.93	83.02	71.67	76.93
GO_CC	80.29	71.70	75.75	83.32	72.95	77.79	80.93	73.19	76.86	83.56	73.18	78.02
GO_BPMF	55.48	23.55	33.06	55.48	23.55	33.06	47.67	34.04	39.72	47.67	34.04	39.72
ChEBI	47.83	54.54	50.97	47.34	53.28	50.13	44.35	51.97	47.86	43.83	50.82	47.06
PRO	28.59	63.53	39.43	41.72	52.72	46.58	27.48	63.96	38.45	40.39	53.33	45.96
SO	50.68	52.39	51.52	50.50	51.76	51.12	45.60	50.43	47.89	43.71	48.73	46.08

Bolded values denote an increase in F-Score against the baseline. Using a case-sensitive processing approach leads to an increase in F-Score ranging from 0.02 to 0.07 on three of the six ontologies—i.e., CL, GO_CC and PRO. Similarly, information gain improves the F-Score in two ontologies—i.e., GO_CC and GO_BPMF. The efficiency of the combined approaches mirrors proportionally their individual behaviour.