Skip to main content

Table 4 Experimental results on real data sets (N = 1,083,878)

From: Efficient algorithms for fast integration on large data sets from multiple sources

   Time - TPA Time – TPA(FCED) Clusters Acc. Clusters Individuals Acc. Com.
ED name constant t = 1 1:52:41 0:27:29 94,381 87,756 108,800 93.0% 80.7%
ED all t = 1 3:11:17 0:29:33 101,864 99,562 108,800 97.8% 91.6%
PD name - 1:06:04 1:04:13 90,950 83,270 108,800 91.6% 76.5%
PDED t = 1 2:04:09 1:06:04 101,344 99,711 108,800 98.4% 91.6%
ED name proportional t = 0.1 1:55:24 0:30:56 94,521 87,966 108,800 93.1% 80.9%
ED all t = 0.1 3:14:37 0:44:05 101,254 99,346 108,800 98.1% 91.3%
PD name - 1:04:32 1:05:41 90,950 83,270 108,800 91.6% 76.5%
PDED t = 0.1 2:06:16 1:09:02 100,896 98,949 108,800 98.1% 90.9%