0
|
1 trainFile = /u/nlp/data/ner/goodClassifiers/data/muc67.jenny.train
|
|
2 testFile = /u/nlp/data/ner/goodClassifiers/data/muc67.jenny.test
|
|
3 serializeTo = english.muc.7class.distsim.crf.ser.gz
|
|
4
|
|
5 distSimLexicon = /u/nlp/data/pos_tags_are_useless/egw4-reut.512.clusters
|
|
6 useDistSim = true
|
|
7
|
|
8 map = word=0,answer=1
|
|
9
|
|
10 saveFeatureIndexToDisk = true
|
|
11
|
|
12 useClassFeature=true
|
|
13 useWord=true
|
|
14 #useWordPairs=true
|
|
15 useNGrams=true
|
|
16 noMidNGrams=true
|
|
17 maxNGramLeng=6
|
|
18 usePrev=true
|
|
19 useNext=true
|
|
20 #useTags=true
|
|
21 #useWordTag=true
|
|
22 useLongSequences=true
|
|
23 useSequences=true
|
|
24 usePrevSequences=true
|
|
25 maxLeft=1
|
|
26 useTypeSeqs=true
|
|
27 useTypeSeqs2=true
|
|
28 useTypeySequences=true
|
|
29 useOccurrencePatterns=true
|
|
30 useLastRealWord=true
|
|
31 useNextRealWord=true
|
|
32 #useReverse=false
|
|
33 normalize=true
|
|
34 # normalizeTimex=true
|
|
35 wordShape=chris2useLC
|
|
36 useDisjunctive=true
|
|
37 disjunctionWidth=5
|
|
38 #useDisjunctiveShapeInteraction=true
|
|
39
|
|
40 type=crf
|
|
41
|
|
42 readerAndWriter=edu.stanford.nlp.sequences.ColumnDocumentReaderAndWriter
|
|
43
|
|
44 useObservedSequencesOnly=true
|
|
45
|
|
46 useQN = true
|
|
47 QNsize = 25
|
|
48
|
|
49 # makes it go faster
|
|
50 featureDiffThresh=0.05
|