def run_turkish():
params = Parameter()
params.UseTransRules = True
params.DoPruning = False
params.DoCompound = False
params.ExcludeUnreliable = False
params.BestNCandSuffix = 150
infile_train = r'data/wordlist.2010.tur.utf8.txt'
infile_test_gold = r'data/mit/gold.tur.txt'
run_experiment(infile_train, infile_test_gold, params)
def run_finnish():
"""Runs an experiment on Finnish data against gold standard results."""
params = Parameter()
params.UseTransRules = False
params.DoPruning = True
params.DoCompound = True
params.ExcludeUnreliable = True
params.BestNCandSuffix = 150
infile_train = r'data/wordlist.2010.fin.utf8.txt'
infile_test_gold = r'data/mit/gold.fin.txt'
run_experiment(infile_train, infile_test_gold, params)
params = Parameter()
arg_parser = argparse.ArgumentParser()
arg_parser.add_argument('infile', help='The input file containing a word list with line format: <word> <freq>')
arg_parser.add_argument('outfile', help='The output file to save the segmentation result')
arg_parser.add_argument('-p', '--prune', help='Whether use pruning (1|0, default:%s)' % params.DoPruning, type=bool, default=params.DoPruning)
arg_parser.add_argument('-t', '--trans', help='Whether use transformation rules (1|0, default:%s)' % params.UseTransRules, type=bool, default=params.UseTransRules)
arg_parser.add_argument('-c', '--comp', help='Whether process compounding (1|0, default:%s)' % params.DoCompound, type=bool, default=params.DoCompound)
arg_parser.add_argument('-e', '--excl', help='Whether exclude unreliable roots (1|0, default:%s)' % params.ExcludeUnreliable, type=bool, default=params.ExcludeUnreliable)
arg_parser.add_argument('-n', '--hyphen', help='Whether explicitly deal with hyphen words (1|0, default:%s)' % params.DoHyphen, type=bool, default=params.DoHyphen)
arg_parser.add_argument('-a', '--apos', help='Whether explicitly deal with apostrophes (1|0, default:%s)' % params.DoApostrophe, type=bool, default=params.DoApostrophe)
arg_parser.add_argument('-r', '--root', help='Minimal length of roots that will be possibly segmented (default:%s)' % params.MinStemLen, type=int, default=params.MinStemLen)
arg_parser.add_argument('-s', '--suff', help='Maximal length of suffixes (default:%s)' % params.MaxSuffixLen, type=int, default=params.MaxSuffixLen)
args = arg_parser.parse_args()
params.DoPruning = args.prune
params.UseTransRules = args.trans
params.DoCompound = args.comp
params.ExcludeUnreliable = args.excl
params.DoHyphen = args.hyphen
params.DoApostrophe = args.apos
params.MinStemLen = args.root
params.MaxSuffixLen = args.suff
params.print_all()
run(args.infile, args.outfile, params)
'--apos',
help='Whether explicitly deal with apostrophes (1|0, default:%s)' %
parameters.DoApostrophe,
type=bool,
default=parameters.DoApostrophe)
arg_parser.add_argument(
'-r',
'--root',
help=
'Minimal length of roots that will be possibly segmented (default:%s)'
% parameters.MinStemLen,
type=int,
default=parameters.MinStemLen)
arg_parser.add_argument('-s',
'--suff',
help='Maximal length of suffixes (default:%s)' %
parameters.MaxSuffixLen,
type=int,
default=parameters.MaxSuffixLen)
args = arg_parser.parse_args()
parameters.DoPruning = args.prune
parameters.UseTransRules = args.trans
parameters.DoCompound = args.comp
parameters.ExcludeUnreliable = args.excl
parameters.DoHyphen = args.hyphen
parameters.DoApostrophe = args.apos
parameters.MinStemLen = args.root
parameters.MaxSuffixLen = args.suff
parameters.print_all()
run(args.infile, args.outfile, parameters)