packages feed

morfette 0.3 → 0.3.1

raw patch · 4 files changed

+70/−5 lines, 4 files

Files

GramLab/Morfette/Utils.hs view
@@ -209,7 +209,7 @@ train (prepr,_) fspecs flags [dat,modeldir] = do   toks <- fmap (map parseToken . lines) (readFile dat)   let tokSet = Set.fromList [ s | t@(s,_,_) <- toks ]-  dict <- getDict flags $ Just tokSet+  dict <- getDict flags Nothing   let langConf = case [f | Lang f <- flags ] of { [] -> "xx" ; [f] -> f }       lex = Conf { dictLex = dict                  , lang = langConf }
+ Makefile view
@@ -0,0 +1,63 @@+defaults: all+++#program name+MORFETTE=../src/dist/build/morfette/morfette+++#parameters for training and eval+# modify to suit your needs+# for french TYPE can be either ftb4 or ftbmax++TYPE=ftb4+TRAINDATADIR=../DATA/+PREF=${TRAINDATADIR}/${TYPE}+TRAINSET=${PREF}/ftb_1.pos.utf8.morpheteready+DEVSET=${PREF}/ftb_2.pos.utf8.morpheteready+GOLDSET=${PREF}/ftb_3.pos.utf8.morpheteready++LEXICON=${TRAINDATADIR}/lexicon/lexicon.ftb4.4morfette.csv+ITERPOS=5+ITERLEMMA=2+MODELNAME=${PREF}/${TYPE}.${ITERPOS}x${ITERLEMMA}.model++all: configure build+		+	+check: train eval_dev+++configure:+	runghc Setup.lhs configure --user++build: configure+	runghc Setup.lhs build+	+install:+	runghc Setup.lhs install+	+clean:+	runghc Setup.lhs clean+ +++++train:+	${MORFETTE} train --iter-pos=${ITERPOS} --iter-lemma=${ITERLEMMA} \+	${TRAINSET} ${MODELNAME} \+	--dict=${LEXICON} +RTS -K100m -sstderr+	+	+eval_dev:+	${MORFETTE} predict  ${MODELNAME} < ${DEVSET} > ${DEVSET}.tagged+	${MORFETTE}  eval --ignore-case ${TRAINSET} ${DEVSET} ${DEVSET}.tagged |\+	tee ${DEVSET}.tagged.result.${TYPE}.${ITERPOS}x${ITERLEMMA}.model+	+	++eval_gold:+	${MORFETTE} predict  ${MODELNAME} < ${GOLDSET} > ${GOLDSET}.tagged+	${MORFETTE}  eval --ignore-case ${TRAINSET} ${GOLDSET} ${GOLDSET}.tagged |\+	tee ${GOLDSET}.tagged.result.${TYPE}.${ITERPOS}x${ITERLEMMA}.model+
README view
@@ -1,9 +1,11 @@ =INTRODUCTION= +Morfette website: http://sites.google.com/site/morfetteweb/+ Morfette is a tool for supervised learning of inflectional morphology. Given a corpus of sentences annotated with lemmas  and morphological labels, and optionally a lexicon, morfette -learns how to morphologically analyse new sentences.+learns how to morphologically analyse new sentences.   In the learning stage Morfette fits two separate logistic regression models: one for morphological tagging and one for lemmatization. The@@ -55,7 +57,7 @@  =EXAMPLE USAGE= To train a new model:-morfette train --dict-file=DICT TRAINING-FILE MODEL-DIR+morfette train --dict-file=DICT TRAINING-FILE MODEL-DIR +RTS -K100m  To use the model in MODEL-DIR to analyze new data: morfette predict MODEL-DIR < TEST-DATA > ANALYZED-TEST-DATA
morfette.cabal view
@@ -1,5 +1,5 @@ Name:		morfette-Version:	0.3+Version:	0.3.1 Homepage:	http://sites.google.com/site/morfetteweb/ Synopsis:	A tool for supervised learning of morphology Description:	Morfette is a tool for supervised learning of inflectional@@ -12,7 +12,7 @@ Maintainer:	Grzegorz Chrupała <gchrupala@lsv.uni-saarland.de> build-type:     Simple category:       Natural Language Processing-Extra-source-files: README, INSTALL+Extra-source-files: README, INSTALL, Makefile  Executable:     morfette Main-Is:        Main.hs