Table 10. Randomness and variability of the results.

Dataset	Performance metrics with different random seeds
	Accuracy			F1-score
	Mean	Std	Conf. interval (Conf. Level 95%)	Mean	Std	Conf. interval (Conf. Level 95%)
ceylon-ide-eclipse	68.18	6.47	[64.17–72.19]	76.94	7.01	[72.60–81.29]
BroadleafCommerce	75.57	1.58	[74.60–76.55]	84.45	1.32	[83.63–85.27]
hazelcast	61.36	1.32	[60.54–62.18]	65.17	4.57	[62.34–68.00]
elasticsearch	65.24	1.51	[64.31–66.18]	74.02	3.56	[71.82–76.23]
MapDB	69.21	4.86	[66.20–72.23]	74.44	5.27	[71.17–77.70]
netty	65.03	5.39	[61.69–68.37]	73.38	5.98	[69.67–77.08]
orientdb	66.05	3.59	[63.83–68.28]	73.86	4.45	[71.10–76.62]
neo4j	66.56	4.44	[63.81–69.32]	74.67	5.02	[71.55–77.78]
titan	70.19	2.99	[68.34–72.04]	79.78	2.64	[78.14–81.42]
mcMMO	56.68	4.54	[53.87–59.50]	64.09	6.21	[60.24–67.94]
Android-Universal-Image-Loader	64.70	3.59	[62.47–66.92]	74.23	2.28	[72.81–75.65]
antlr4	71.25	1.25	[70.48–72.02]	81.68	0.97	[81.08–82.28]
junit	70.96	2.90	[69.16–72.75]	80.64	2.35	[79.19–82.10]
mct	93.18	5.47	[89.79–96.57]	94.39	5.79	[90.80–97.98]
oryx	88.77	1.72	[87.70–89.83]	93.80	1.10	[93.12–94.48]