/***************************************************************************/ /* How to use the TreeTagger */ /* Author: Helmut Schmid, University of Stuttgart, Germany */ /***************************************************************************/ The TreeTagger consists of two programs: train-tree-tagger is used to create a parameter file from a lexicon and a handtagged corpus. tree-tagger expects a parameter file and a text file as arguments and annotates the text with part-of-speech tags. The file formats are described below. By default, the programs are located in the ./bin sub-directory. If either of the programs is called without arguments, it will print information about its usage. Tagging ------- Tagging is done with the tree-tagger program. It requires at least one command line argument, the parameter file. If no input file is specified, input will be read from stdin. If neither an input file nor an output file is specified, the tagger will print to stdout. tree-tagger {-options-} { {}} Description of the command line arguments: * : Name of a parameter file which was created with the train-tree-tagger program. * : Name of the file which is to be tagged. Each token in this file has to be on a separate line. Tokens may contain blanks. It is possible to override the lexical information contained in the parameter file of the tagger by specifying a list of possible tags after a token. This list has to be preceded by a tab character and the elements are separated by tab characters. This pretagging feature could be used e.g. to ensure that certain text-specific expressions are tagged properly. Punctuation marks must be on separate lines as well. Clitics (like "'s", "'re", and "'d" in English or "-la" and "-t-elle" in French) should be separated if they were separated in the training data. (The French and English parameter files available by ftp expect separation of clitics). Sample input file: He moved to New York City NP . * : Name of the file to which the tagger should write its output. Further optional command line arguments: * -token: The words/tokens are printed in addition to the POS tags * -lemma: Lemmas are printed as well. * -sgml: This option instructs the tagger to ignore tokens which start with '<' and end with '>' (SGML tags). * -lex : The file contains additional lexicon entries to be used by the tagger. The file format is identical to the format of the lexicon argument of the training program (see below). * -no-unknown: If an unknown word is encountered, emit the word form as lemma. This was previously the default behaviour. Now, the default behaviour is to print "" as lemma. * -threshold

: This option tells the tagger to print all tags of a word with a probability higher than

times the largest probability. (The tagger will use a different algorithm in this case and the set of best tags might be different from the tags generated without this option.) * -prob: Print tag probabilities (in combination with option -threshold) * -pt-with-prob: If this option is specified, then each pretagging tag (see above) has to be followed by a whitespace and a tag probability value. * -pt-with-lemma: If this option is specified, then each pretagging tag (see above) has to be followed by a whitespace and a lemma. Lemmas may contain blanks. If both -pt-with-prob and -pt-with-lemma have been specified, then each pretagging tag is followed by a probability and a lemma in that order. * -hyphen-heuristics: needed for chunking. See below for more information about how to train a chunk parameter file with hyphen-heuristics. The options below are for advanced users. Please, read the papers on the TreeTagger to fully understand their meaning. * -proto: If this option is specified, the tagger creates a file named "lexicon-protocol.txt", which contains information about the degree of ambiguity and about the other possible tags of a word form. The part of the lexicon in which the word form has been found is also indicated. 'f' means fullform lexicon and 's' means affix lexicon. 'h' means that the word contains a hyphen and that the part of the word following the hyphen has been found in the fullform lexicon. * -eps : Value which is used to replace zero lexical frequencies. This is the case if a word/tag pair is contained in the lexicon but not in the training corpus. The choice of this parameter has only minor influence on the tagging accuracy. * -base: If this option is specified, only lexical information is used for tagging but no contextual information about the preceding tags. This option is only useful in order to obtain a baseline result to which to compare the actual tagger output. Training -------- Training is done with the *train-tree-tagger* program. It expects at least four command line arguments which are described below. train-tree-tagger {options} Description of the command line arguments: * : name of a file which contains the fullform lexicon. Each line of the lexicon corresponds to one word form and contains the word form and a sequence of tag-lemma pairs. Each tag is preceded by a tab character and each lemma is preceded by a blank or tab character. Example: aback RB aback abacuses NNS abacus abandon VB abandon VBP abandon abandoned JJ abandoned VBD abandon VBN abandon abandoning VBG abandon Remark: The tagger doesn't need the lemmas for tagging actually. If you do not have the lemma information or if you do not plan to annotate corpora with lemmas, you can replace the lemma with a dummy value, e.g. "-". You can use the Perl script make-lex.perl as follows in order to create a tagger lexicon from the training corpus: cmd/make-lex.perl corpus > lexicon If you have additional lexicon entries stored in a separate file "lex" with entries like this (The POS tag is preceded by a tab character.) aback RB aback aback RP aback abacs NNS abac you can include them as follows: cmd/make-lex.perl corpus lex > lexicon If train-tree-tagger complains about unknown tags, just add another entry to the lexicon with the respective POS tag. * : name of a file which contains a list of open class tags i.e. possible tags of unknown word forms separated by whitespace. The tagger will use this information when it encounters unknown words, i.e. words which are not contained in the lexicon. Example: (for Penn Treebank tagset) FW JJ JJR JJS NN NNS NP NPS RB RBR RBS VB VBD VBG VBN VBP VBZ * : name of a file which contains tagged training data. The data must be in one-word-per-line format. This means that each line contains one token and one tag in that order separated by a tabulator. Punctuation marks are considered as tokens and must be tagged as well. The file should neither contain empty lines nor untagged SGML markup. Example: Pierre NP Vinken NP , , 61 CD years NNS * : name of the file in which the resulting tagger parameters are stored. The following parameters are optional. Read the papers on the TreeTagger to fully understand their meaning. * -st : the end-of-sentence part-of-speech tag, i.e. the tag which is assigned to sentence punctuation like ".", "!", "?". Default is "SENT". You have to use this option, if your tag for sentence punctuation is not "SENT". If you have more than one such tag, choose the most frequent one. * -utf8 assume that the data is encoded with UTF8 * -cl : number of preceding words forming the statistical context. The default is 2 which corresponds to a trigram context. For small training corpora and/or large tagsets, it could be useful to reduce this parameter to 1. * -dtg : Threshold - If the information gain at a leaf node of the decision tree is below this threshold, the node is deleted. * -sw : A smoothing parameter, which determines how much the probability distribution of some decision tree node is smoothed with the probability distribution of the parent node. * -ecw : weight of the equivalence class based probability estimates. * -atg Threshold - If the information gain at a leaf of an affix tree is below this threshold, it is deleted. The default is 1.2. The accuracy of the TreeTagger usually improves, if different settings of the above parameters are tested and the best combination is chosen. Caveat: Make sure that the lexicon and the training corpus contain no extra blanks. If the word form, for instance, is followed by a blank and a tab character, the blank will be considered part of the word. The script 'cmd/create-pos-parameter-file' can be used to train a parameter file, provided the file: 'lib/open-class-tags' exists. The script creates a lexicon and a parameter file that both will be stored in the lib-directory. Training a parameter file for chunking --------------------------------------- (This section of the README file was created by Wiebke Wagner.) Training is done with the *train-tree-tagger* program just like the training of part-of-speech parameter files. The input files differ (see below). train-tree-tagger {options} Description of the command line arguments: * : name of a file which contains the fullform lexicon. Each line of the lexicon corresponds to one word form and contains the word form with its pos-tag and a sequence of chunktag-lemma pairs. Since there is no lemma for a string containing a word and its pos, lemma is just a dummy-placeholder. Each chunktag is preceded by a tab character and each dummy-lemma is preceded by a blank or tab character. Example: Abs-NP NP/I-NC # NP/B-NC # Academic-NP NP/I-NC # Activated-VVN VBN/B-VC # VBN/B-NC # Activation-NN NN/I-NC # NN/B-NC # VB VB/I-VC # The lexicon must contain specific entries that contain a hypen ('WORD-POS POS/IOB DUMMY-LEMMA') and general entries without a hypen ('POS POS/IOB DUMMY-LEMMMA)'. The hypen-heuristics enables the program to select the general entry if no specific entry is available. You can use the Perl script make-chunk-lex.perl as follows in order to create a tagger-chunker lexicon from the training corpus: 'cmd/make-chunk-lex.perl' corpus > chunker-lexicon Attention: if 'train-tree-tagger' shows the error message: 'ERROR: Sentence punctuation tag "SENT" is not in lexicon!' add the option: [-st 'SENT/O'] to the system call. If train-tree-tagger complains about unknown tags, just add another entry to the lexicon with the respective POS tag. * : name of a file which contains only the dummy entry: NN/B-NC. If there are general entries in the lexicon for every word class, and if the hyphen heuristic is activated no tags have to be guessed. * : name of a file which contains training data annotated with part-of-speech tags and chunk tags. The data must be in one-word-per-line format. This means that each line contains one token with its mark-up: WORD-POS POS/IOB DUMMY-LEMMA The file should neither contain empty lines nor untagged SGML markup. Example: Activation-NN NN/I-NC of-IN IN/B-PC the-DT DT/B-NC CD28-NP NP/I-NC surface-NN NN/I-NC * : name of the file in which the resulting tagger parameters are stored. The optional parameters like for training a part-of-speech parameter file The accuracy of the TreeTagger usually improves, if different settings of the above parameters are tested and the best combination is chosen. Caveat: Make sure that the lexicon and the training corpus contain no extra blanks. If the word form, for instance, is followed by a blank and a tab character, the blank will be considered part of the word. The script 'cmd/create-chunk-parameter-file' can be used to train a chunk parameter file, provided the file: 'lib/open-class-chunks' exists and contains the dummy entry: NN/B-NC The script creates a lexicon and a parameter file that both will be stored in the lib-directory.