úÎeÑ]l      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVW X Y Z [ \ ] ^ _ ` a b c d e f g h i j k None9An atomic part of morphosyntactic tag with optional POS. A tier description. $Does it include the part of speech? Tier grammatical attributes. Select tier attributes. Split the positional tag. &Default tiered tagging configuration.  lm     lmNone A weighted collection of type a elements. %A word parametrized over a tag type. Orthographic form. 5Set of word interpretations. To each interpretation  a (weight of correctness within the context is assigned. 7Out-of-vocabulary (OOV) word, i.e. word unknown to the  morphosyntactic analyser. A sentence of s. Map function over word tags. !Map function over sentence tags. Interpretations of the word. Interpretations of the word. Make a weighted collection. 1Map function over weighted collection elements. n nNone Document format. 2Parse textual interpretations into a functor with  sentence elements. +Show textual reprezentation of a document. Sentence handler. !Sentence handler. #%Split sentence into a list of words. $Merge words with a sentence. %Words handler. &Words handler. (*Extract information relevant for tagging. )3Select the set of morphosyntactic interpretations. *0Textual representation of morphposyntactic tag.  !"#$%&'()* !"#$%&'()**&'()!"#$%  !"#$%&'()*None o A token. p<Interpretations of the token, each interpretation annotated  with a disamb Boolean value (if q, the interpretation ! is correct within the context). rNo space, space or newline. ++Create document handler given value of the ignore tag. sSentence handler. tWord handler. u*Extract information relevant for tagging. vSelect interpretations. w Printing. An infix synonym for x. !yz{|o}~€pr‚ƒ„+stuv…†‡ˆ‰Šw‹ŒŽ++yz{|o}~€prƒ‚„+stuv…†‡ˆ‰Šw‹ŒŽNone,@An observation consist of an index (of list type) and an actual  observation value. -?A schema is a block of the Ox computation performed within the = context of the sentence and the absolute sentence position. .CThe Ox monad specialized to word token type and text observations. /Default guessing schema. 0Default disambiguation schema. ,-./0,-./0-.,/0,-./0None 17A guesser represented by the conditional random field. ‘0Schematize the input sentence with according to schema rules. 4Determine the k5 most probable labels for each word in the sentence. ’0Include guessing results into weighted tag maps  assigned to individual words. 5,Include guessing results into the sentence. 6;Tag sentence in external format. Selected interpretations 6 (tags correct within the context) will be preserved. 7 Tag file. 8Train guesser. “&Schematized data from the plain file. 123‘4’567Document format handler Guesser argument Observation schema Guesser itself Input Output 8Document format handler Observation schema SGD parameters Training file Maybe eval file “1234567812345678 123‘4’5678“None 9CRF tagging function. :CRF training function. ;3Split is just a function from an original tag form  to a complex tag form. ”0Schematize the input sentence with according to schema rules. •2Unsplit the complex tag (assuming, that it is one & of the interpretations of the word). <*Perform context-sensitive disambiguation. =Tag the sentence. >Disambiguate document. ?Train disamb model. –&Schematized data from the plain file. 9:Training data — action Maybe evalation data Resulting model ;”•<=>Document format handler Observation schema Tiered tagging CRF tagging function Input Output ?Document format handler Observation schema Tiered tagging CRF training function Training file Maybe eval file Resultant model data –9:;<=>?;:9<=>? 9:;”•<=>?–None @CRF model data. ˜CRF feature map. ™9Transition map restricted to a particular tagging layer. š5Codec internal data. The first component is used to : encode observations of type a, the second one is used to  encode labels of type [b]. E Feature. O Sublabel. R Observation. ›"Feature generation for complex [Lb] label type. œ)Codec dependes on the number of layers. Dummy feature index. ž0Codec specification given the number of layers. U<Train the CRF using the stochastic gradient descent method. @ Use the provided feature selection function to determine model  features. V'Find the most probable label sequence. *@ABCD˜Ÿ ¡™šEFGHIJKLMNOPQRST¢£›¤¥œ¦žUNumber of tagging layers Feature selection  Args for SGD V§¨©ª@ABCDEFGHIJKLMNOPQRSTUV!RSTOPQEJIHFKLMNKLNKNGKN@ABCDUV@ABCD˜Ÿ ¡™šEJIHFKLMNKLNKNGKNOPQRST¢£›¤¥œ¦žUV§¨©ª None W-Training disambiguation model configuration. X+Tagging with disambiguation configuration. Y%Disambiguation configuration with... ]Disambiguation configuration. a'Guessing configuration and model data. eGuessing configuration. i?Perform disambiguation preceded by context-sensitive guessing. «Tag the sentence. jTag document. k*Train guessing and disambiguation models. WXYZ[\]^_`abcdefghiGuessing configuration Disambiguation configuration Input Output «jDocument format handler Guessing configuration Disambiguation configuration Input Output kDocument format handler Guessing configuration SGD params for guesser Disambiguation configuration Training file Maybe eval file Resultant models ¬Document format handler Guesser !Template for temporary file name File to guess Handler WXYZ[\]^_`abcdefghijkefghabcd]^_`YZ[\XWijk WXYZ[\]^_`abcdefghi«jk¬­  !"#$%%&'()*+,-./01234456789:;<=>?@:AABCDEFGHIJKLMNOOP//QRS T U V V W X Y Y  Z [ [ \ ] ^ ^ _ ` > @ :abc"defg(+,hijklliScmnogpqr)stuvwxyz{|}~}€de‚ƒ„…†‡ˆ‚‰Š‹ŒŽ‘’“ ? ”•concraft-0.3.0NLP.Concraft.Disamb.TieredNLP.Concraft.Disamb.PositionalNLP.Concraft.MorphosyntaxNLP.Concraft.FormatNLP.Concraft.Format.PlainNLP.Concraft.SchemaNLP.Concraft.GuessNLP.Concraft.Disamb NLP.Concraftcrf-chain2-generic-0.3.0Data.CRF.Chain2.Generic.Model selectHidden selectPresentFeatSelPartposattsTierwithPoswithAttsselectsplittierConfDefaultWMapunWMapWordorthtagWMapoovSentmapWordmapSent interpsSetinterpsmkWMapmapWMapDocparseDocshowDoc sentHandler parseSent mergeSent wordHandlerextractTag plainFormatObSchemaOxguessSchemaDefaultdisambSchemaDefaultGuessercrfguessinclude guessSentguessDoctrainOnTagCRFTrainCRFSplitdisamb disambSent disambDocCRF numOfLayers codecDatamodelFeatOFeatobTFeat1TFeat2TFeat3x1x2x3lnLbunLbunObtraintag DisambTrain DisambTag DisambWith disambConf disambWith DisambConf disambSchema GuessData guessConfguesser GuessConfguessNum guessSchema $fBinaryPart $fBinaryTierTokenghc-prim GHC.TypesTrueSpace<>base Data.MonoidmappendInterpspaceknownNewLineNonenoneBase parsePlain parseWord parseInterp parseHeader parseSpace showPlain buildSent buildWord buildInterps buildSpace buildKnown schematize includeWMapsschemedunSplitIOFeatMapTransMap CodecDatafeatGencodecdummy codecSpec transMapsotherMapputIgetIobLenslbLens!? $fBinaryCRF$fFeatMapFeatMapFeat$fBinaryFeatMap $fBinaryFeat guessFile