mssammon|csil-mm24|/scratch/tutorial/ne/analysis|[1]% ls
analyze.pl*  conll03.train.upd.lex  ne.analysis  ne.res  run_ne_analysis.script


mssammon|csil-mm24|/scratch/tutorial/ne/analysis|[2]% ./analyze.pl
Usage: ./analyze.pl snowresults lexicon [> output]


mssammon|csil-mm24|/scratch/tutorial/ne/analysis|[3]% head ne.res
Algorithm information:
Winnow: (1.35, 0.8, 4, 0.3435) Targets: 1-5
Example 1 Label: 5
5:	1*
4:	0
2:	0
1:	0
3:	0

Example 2 Label: 4


mssammon|csil-mm24|/scratch/tutorial/ne/analysis|[4]% head conll03.train.upd.lex
1	label[ORG]
1001	phLen[1]
1002	w[*_German]&t[*_JJ]
1003	w[*___to]&t[*___TO]
1004	w[*__call]&t[*__NN]
1005	w[*rejects]&t[*VBZ]
1006	w[*EU]&t[*NNP]
1007	w[EU]&t[NNP]
1008	t[*VBZ]
1009	t[*_JJ]


mssammon|csil-mm24|/scratch/tutorial/ne/analysis|[5]% ./analyze.pl ne.res conll03.train.upd.lex 
name	prec.	recall	f1	num
ORG	0.860	0.618	0.719	1661
MISC	0.788	0.624	0.696	702
PER	0.931	0.694	0.795	1617
LOC	0.633	0.913	0.748	1668
OTHER	0.926	0.960	0.943	7761
average precision: 0.87480489214304
average recall: 0.862331270042509
average F1: 0.860184218201154


mssammon|csil-mm24|/scratch/tutorial/ne/analysis|[6]%