Skip to main content
. Author manuscript; available in PMC: 2019 Dec 10.
Published in final edited form as: JAMIA Open. 2019 Mar 28;2(2):246–253. doi: 10.1093/jamiaopen/ooz007

Table 2.

Results of comparison between word embedding expanded queries and baseline queries (task 1) for 14 dietary supplements

Queries Number of clinical notes Number of patients
Dietary supplements Number of word embedding expanded terms Base query Word embedding query Additional records found Percentage increase (%) Base query Word embedding query Additional patients found Percentage increase (%)
Calcium 12 7 450 261 7 543 569 93 308 1.25 1 000 561 1 002 211 1650 0.16
Chamomile 3 5221 6120 899 17.22 3504 4146 642 18.32
Cranberry 3 196 862 198 625 1763 0.90 76 664 77 327 663 0.86
Dandelion 2 4468 4564 96 2.15 2377 2419 42 1.77
Flaxseed 2 104 007 169 343 65 336 62.82 25 136 45 222 20 086 79.91
Garlic 1 92 803 93 941 1138 1.23 31 273 31 400 127 0.41
Ginger 1 96 438 96 452 14 0.01 59 693 59 698 5 0.01
Ginkgo 3 20 259 28 093 7834 38.67 5854 7791 1937 33.09
Ginseng 2 9926 11 277 1351 13.61 4023 4469 446 11.09
Glucosamine 5 466 617 467 758 1141 0.24 70 842 70 938 96 0.14
Lavender 3 18 793 20 667 1874 9.97 11 855 13 011 1156 9.75
Melatonin 1 753 511 753 753 242 0.03 118 846 118 896 50 0.04
Turmeric 3 33 573 48 749 15 176 45.20 8379 13 486 5107 60.95
Valerian 2 15 883 16 219 336 2.12 7051 7207 156 2.21