Table 1 Description of general feature types with examples
From: Machine learning with naturally labeled data for identifying abbreviation definitions
FC | A character of a SF matches the 1st character of a word in a LF |
| Â | C erebrospinal fluid ( C SF) |
FC-ST | A character of a SF matches the 1st character of a stop-word in a LF |
| Â | people w ith AIDS coalition (P W A) |
FCG | A character of a SF matches the character following a non-alphanumeric non-space character in a LF |
| Â | GH- r eleasing peptide (GH R P) |
SBW | A character of a SF matches a character within a token in a LF such that token splits at that character into two substrings, one or both of which are defined words. |
| Â | Cerebro s pinal fluid ( C SF) |
LS | The last character of a SF is ‘s’ and last token in a LF ends in ‘s’ or ‘i’ |
| Â | plasma concentration s (PC s ) |
ALC | A letter of a SF matches a capital non-1st character letter in a LF |
| Â | gamma-vinyl G ABA (GV G ) |
ALS | All characters of a SF appear anywhere in a single token in a LF in the correct order |
| Â | br omo d eoxy u ridine ( BrdU ) |
LT | Look-up table match between a character in a SF and a token in a LF |
| Â | Current ( I ) |
CL | A substring of a SF matches two or more consecutive characters of a token in a LF |
| Â | methyl- beta -cyclodextrin (M beta CD) |