48 lines
1.0 KiB
Plaintext
48 lines
1.0 KiB
Plaintext
|
|
trainFileList = /u/nlp/data/spanish/ner/ancora.ner.train.tsv,/u/nlp/data/spanish/ner/ldc-DF.ner.train.tsv,/u/nlp/data/spanish/ner/ldc-NW.ner.train.tsv
|
|
testFiles = /u/nlp/data/spanish/ner/ancora.ner.dev.tsv,/u/nlp/data/spanish/ner/ldc-DF.ner.dev.tsv,/u/nlp/data/spanish/ner/ldc-NW.ner.dev.tsv
|
|
serializeTo = spanish.ancora2.crf.ser.gz
|
|
|
|
useDistSim = false
|
|
|
|
map = word=0,answer=1
|
|
|
|
useTitle = true
|
|
useSpanishTitle = true
|
|
useClassFeature=true
|
|
useWord=true
|
|
useNGrams=true
|
|
noMidNGrams=true
|
|
usePrev=true
|
|
useNext=true
|
|
useLongSequences=true
|
|
useSequences=true
|
|
usePrevSequences=true
|
|
maxLeft=3
|
|
useTypeSeqs=true
|
|
useTypeSeqs2=true
|
|
useTypeySequences=true
|
|
useOccurrencePatterns=true
|
|
useLastRealWord=true
|
|
useNextRealWord=true
|
|
normalize=true
|
|
wordShape=dan2uselC
|
|
useDisjunctive=true
|
|
disjunctionWidth=5
|
|
#useDisjunctiveShapeInteraction=true
|
|
|
|
type=crf
|
|
|
|
saveFeatureIndexToDisk = true
|
|
|
|
readerAndWriter=edu.stanford.nlp.sequences.ColumnDocumentReaderAndWriter
|
|
|
|
useObservedSequencesOnly=true
|
|
|
|
sigma = 1
|
|
useQN = true
|
|
QNsize = 25
|
|
|
|
# makes it go faster
|
|
featureDiffThresh=0.05
|