Enhancing the Efficiency of a Decision Support System through the Clustering of Complex Rule-Based Knowledge Bases and Modification of the Inference Algorithm

Table 4

Description of inference efficiency vs. representative generation methods.

Representative generation method	Fired rules		Empty representative		New facts		Searched clusters
	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Threshold	5.31	21.69	0.0	0.0	0.92	1.95	54.53	102.35
LowerApp	5.65	23.31	71.13	60.55	0.79	1.67	62.05	111.50
UpperApp	11.53	31.68	0.0	0.0	1.32	2.63	95.14	121.03
Weighted	4.64	19.93	30.45	48.67	0.82	1.69	52.52	101.17