Missing values compensation in duplicates detection using hot deck method

Journal of Big Data

Table 2 Experiment results

Data set	Method	Declared duplicates	True positives	False positives	False negatives	True negatives	Precision	Recall	F-measure
Restaurant	DuDe	97	78	19	34	16114	0.80	0.70	0.75
Restaurant	DDID	97	97	0	15	16133	1.00	0.87	0.93
CD	DuDe	306	205	101	94	476762	0.67	0.69	0.68
CD	DDID	262	236	26	63	476837	0.90	0.79	0.84
MusicBrainz (A)	DuDe	97090	33103	63987	110647	28649838	0.34	0.23	0.27
MusicBrainz (A)	DDID	79818	52185	27633	91565	28686192	0.65	0.36	0.46
MusicBrainz (B)	DuDe	224125	46414	177711	216280	118748569	0.20	0.17	0.18
MusicBrainz (B)	DDID	216722	53757	162965	208937	118763315	0.25	0.20	0.22