53 lines
1.1 KiB
Plaintext
53 lines
1.1 KiB
Plaintext
|
trainFile = /u/nlp/data/ner/goodClassifiers/data/muc67.jenny.train
|
||
|
testFile = /u/nlp/data/ner/goodClassifiers/data/muc67.jenny.test
|
||
|
serializeTo = english.muc.7class.nodistsim.crf.ser.gz
|
||
|
|
||
|
wordFunction = edu.stanford.nlp.process.AmericanizeFunction
|
||
|
|
||
|
distSimLexicon = /u/nlp/data/pos_tags_are_useless/egw.bnc.200
|
||
|
useDistSim = false
|
||
|
|
||
|
map = word=0,answer=1
|
||
|
|
||
|
saveFeatureIndexToDisk = true
|
||
|
|
||
|
useClassFeature=true
|
||
|
useWord=true
|
||
|
#useWordPairs=true
|
||
|
useNGrams=true
|
||
|
noMidNGrams=true
|
||
|
maxNGramLeng=6
|
||
|
usePrev=true
|
||
|
useNext=true
|
||
|
#useTags=true
|
||
|
#useWordTag=true
|
||
|
useLongSequences=true
|
||
|
useSequences=true
|
||
|
usePrevSequences=true
|
||
|
maxLeft=1
|
||
|
useTypeSeqs=true
|
||
|
useTypeSeqs2=true
|
||
|
useTypeySequences=true
|
||
|
useOccurrencePatterns=true
|
||
|
useLastRealWord=true
|
||
|
useNextRealWord=true
|
||
|
#useReverse=false
|
||
|
normalize=true
|
||
|
# normalizeTimex=true
|
||
|
wordShape=chris2useLC
|
||
|
useDisjunctive=true
|
||
|
disjunctionWidth=5
|
||
|
#useDisjunctiveShapeInteraction=true
|
||
|
|
||
|
type=crf
|
||
|
|
||
|
readerAndWriter=edu.stanford.nlp.sequences.ColumnDocumentReaderAndWriter
|
||
|
|
||
|
useObservedSequencesOnly=true
|
||
|
|
||
|
useQN = true
|
||
|
QNsize = 25
|
||
|
|
||
|
# makes it go faster
|
||
|
featureDiffThresh=0.05
|