The impact of near domain transfer on biomedical named entity recognition

s 110 80 Tokens 27,421 26,578 Av. length 32.57 29.93 ANA 194 195 0.33 0.26 (138) (133) CHE 44 147 0.08 0.07 (33) (75) DIS 892 955 0.39 0.27 (282) (442) GGP 1663 754 0.41 0.45 (928) (511) ORG 799 770 0.56 0.67 (429) (323) PHE 507 1430 0.52 0.33 (423) (1113) Table 2: Characteristics of the C1 auto-immune and C2 cardiovascular corpora: number of abstracts, number of tokens, average sentence length, frequency of each entity type. Figures in parentheses represent counts after removing duplication. a: probability that a word in an entity class X in C1 is also a word in entity class X in C2. b: probability that a word in an entity class X in C2 is also a word in entity class X in C1 (3) We calculated from Table 2 the average number of mentions for each entity form by class and noted that this is relatively stable across corpora, except for DIS which has less variation in C2 than C1 and CHE which has more variation in C2 than C1. When combining evidence from both corpora the approximate order of type/token ratio are PHE < ANA < CHE,GGP < ORG < DIS indicating that on average PHE entities have the greatest variation. Average entity lengths in tokens (not shown) indicate that PHE are significantly longer than other entity mentions; and (4) We calculated the probability that a word token in an entity class from one corpus would appear in an instance of the same entity class in the other corpus, reported as columns a and b. Although the probability of an exact match in instances between entities in the two corpora is generally quite low (below 20% data not shown) there appears to be significant vocabulary overlap in most classes except for chemicals. 3.2 Conditional Random Fields As in (Finkel and Manning, 2009) we apply our approach to a linear chain conditional random field (CRF) model (Lafferty et al., 2001; McCallum and Wei, 2003; Settles, 2004; Doan et al., 2012) using the Mallet toolkit1 with default parameters. CRFs have been shown consistently to be among the highest performing bioNER learners. The data selection strategies employed here though are neutral and could have been applied to any other fully supervised learner model.