. Author manuscript; available in PMC: 2009 Jan 1.

Published in final edited form as: Artif Intell Med. 2007 Nov 28;42(1):13–35. doi: 10.1016/j.artmed.2007.10.001

Table 16.

Comparison of features for random corpus. For all pairs of features, the differences between F-measures for PHI and the differences between F-measures for non-PHI are significant at α = 0.05. The only exceptions are the difference of F-measures in non-PHI of lexical bigrams and POS information (marked by †), the difference in F-measures in PHI of MeSH and orthographic features (marked by ‡), and the difference in F-measures in non-PHI of MeSH and orthographic features (marked by •). Best F-measures are in bold.

Feature	Class	Precision	Recall	F-measure
Target words	Non-PHI	91.61%	98.95%	95.14%
Target words	PHI	86.26%	42.03%	56.52%
Lexical bigrams	Non-PHI	95.61%	98.10%	96.84%†
Lexical bigrams	PHI	85.43%	71.14%	77.63%
Syntactic bigrams	Non-PHI	96.96%	98.72%	97.83%
Syntactic bigrams	PHI	90.76%	80.20%	85.15%
POS information	Non-PHI	94.85%	98.38%	96.58%†
POS information	PHI	86.38%	65.84%	74.73%
Dictionary	Non-PHI	88.99%	99.26%	93.85%
Dictionary	PHI	81.92%	21.41%	33.95%
MeSH	Non-PHI	86.49%	100%	92.75%•
MeSH	PHI	0%	0%	0%‡
Orthographic	Non-PHI	86.49%	100%	92.75%•
Orthographic	PHI	0%	0%	0%‡