blob: 6d54c3c4073ef7a700c5660feaf852ed83536068 [file] [log] [blame]
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE chapter PUBLIC "-//OASIS//DTD DocBook XML V4.4//EN"
"http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd"[
]>
<!-- Licensed to the Apache Software Foundation (ASF) under one or more contributor
license agreements. See the NOTICE file distributed with this work for additional
information regarding copyright ownership. The ASF licenses this file to
you under the Apache License, Version 2.0 (the "License"); you may not use
this file except in compliance with the License. You may obtain a copy of
the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required
by applicable law or agreed to in writing, software distributed under the
License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS
OF ANY KIND, either express or implied. See the License for the specific
language governing permissions and limitations under the License. -->
<chapter id="tools.tokenizer">
<title>Tokenizer</title>
<section id="tools.tokenizer.introduction">
<title>Tokenization</title>
<para>
The OpenNLP Tokenizers segment an input character sequence into
tokens. Tokens are usually
words, punctuation, numbers, etc.
<screen>
<![CDATA[
Pierre Vinken, 61 years old, will join the board as a nonexecutive director Nov. 29.
Mr. Vinken is chairman of Elsevier N.V., the Dutch publishing group.
Rudolph Agnew, 55 years old and former chairman of Consolidated Gold Fields
PLC, was named a director of this British industrial conglomerate.
]]>
</screen>
The following result shows the individual tokens in a whitespace
separated representation.
<screen>
<![CDATA[
Pierre Vinken , 61 years old , will join the board as a nonexecutive director Nov. 29 .
Mr. Vinken is chairman of Elsevier N.V. , the Dutch publishing group .
Rudolph Agnew , 55 years old and former chairman of Consolidated Gold Fields PLC ,
was named a nonexecutive director of this British industrial conglomerate .
A form of asbestos once used to make Kent cigarette filters has caused a high
percentage of cancer deaths among a group of workers exposed to it more than 30 years ago ,
researchers reported .
]]>
</screen>
OpenNLP offers multiple tokenizer implementations:
<itemizedlist>
<listitem>
<para>Whitespace Tokenizer - A whitespace tokenizer, non whitespace
sequences are identified as tokens</para>
</listitem>
<listitem>
<para>Simple Tokenizer - A character class tokenizer, sequences of
the same character class are tokens</para>
</listitem>
<listitem>
<para>Learnable Tokenizer - A maximum entropy tokenizer, detects
token boundaries based on probability model</para>
</listitem>
</itemizedlist>
Most part-of-speech taggers, parsers and so on, work with text
tokenized in this manner. It is important to ensure that your
tokenizer
produces tokens of the type expected by your later text
processing
components.
</para>
<para>
With OpenNLP (as with many systems), tokenization is a two-stage
process:
first, sentence boundaries are identified, then tokens within
each
sentence are identified.
</para>
<section id="tools.tokenizer.cmdline">
<title>Tokenizer Tools</title>
<para>The easiest way to try out the tokenizers are the command line
tools. The tools are only intended for demonstration and testing.
</para>
<para>There are two tools, one for the Simple Tokenizer and one for
the learnable tokenizer. A command line tool the for the Whitespace
Tokenizer does not exist, because the whitespace separated output
would be identical to the input.</para>
<para>
The following command shows how to use the Simple Tokenizer Tool.
<screen>
<![CDATA[
$ opennlp SimpleTokenizer]]>
</screen>
To use the learnable tokenizer download the english token model from
our website.
<screen>
<![CDATA[
$ opennlp TokenizerME en-token.bin]]>
</screen>
To test the tokenizer copy the sample from above to the console. The
whitespace separated tokens will be written back to the
console.
</para>
<para>
Usually the input is read from a file and written to a file.
<screen>
<![CDATA[
$ opennlp TokenizerME en-token.bin < article.txt > article-tokenized.txt]]>
</screen>
It can be done in the same way for the Simple Tokenizer.
</para>
<para>
Since most text comes truly raw and doesn't have sentence boundaries
and such, its possible to create a pipe which first performs sentence
boundary detection and tokenization. The following sample illustrates
that.
<screen>
<![CDATA[
$ opennlp SentenceDetector sentdetect.model < article.txt | opennlp TokenizerME tokenize.model | more
Loading model ... Loading model ... done
done
Showa Shell gained 20 to 1,570 and Mitsubishi Oil rose 50 to 1,500.
Sumitomo Metal Mining fell five yen to 692 and Nippon Mining added 15 to 960 .
Among other winners Wednesday was Nippon Shokubai , which was up 80 at 2,410 .
Marubeni advanced 11 to 890 .
London share prices were bolstered largely by continued gains on Wall Street and technical
factors affecting demand for London 's blue-chip stocks .
...etc...]]>
</screen>
Of course this is all on the command line. Many people use the models
directly in their Java code by creating SentenceDetector and
Tokenizer objects and calling their methods as appropriate. The
following section will explain how the Tokenizers can be used
directly from java.
</para>
</section>
<section id="tools.tokenizer.api">
<title>Tokenizer API</title>
<para>
The Tokenizers can be integrated into an application by the defined
API.
The shared instance of the WhitespaceTokenizer can be retrieved from a
static field WhitespaceTokenizer.INSTANCE. The shared instance of the
SimpleTokenizer can be retrieved in the same way from
SimpleTokenizer.INSTANCE.
To instantiate the TokenizerME (the learnable tokenizer) a Token Model
must be created first. The following code sample shows how a model
can be loaded.
<programlisting language="java">
<![CDATA[
try (InputStream modelIn = new FileInputStream("en-token.bin")) {
TokenizerModel model = new TokenizerModel(modelIn);
}]]>
</programlisting>
After the model is loaded the TokenizerME can be instantiated.
<programlisting language="java">
<![CDATA[
Tokenizer tokenizer = new TokenizerME(model);]]>
</programlisting>
The tokenizer offers two tokenize methods, both expect an input
String object which contains the untokenized text. If possible it
should be a sentence, but depending on the training of the learnable
tokenizer this is not required. The first returns an array of
Strings, where each String is one token.
<programlisting language="java">
<![CDATA[
String tokens[] = tokenizer.tokenize("An input sample sentence.");]]>
</programlisting>
The output will be an array with these tokens.
<programlisting>
<![CDATA[
"An", "input", "sample", "sentence", "."]]>
</programlisting>
The second method, tokenizePos returns an array of Spans, each Span
contain the begin and end character offsets of the token in the input
String.
<programlisting language="java">
<![CDATA[
Span tokenSpans[] = tokenizer.tokenizePos("An input sample sentence.");]]>
</programlisting>
The tokenSpans array now contain 5 elements. To get the text for one
span call Span.getCoveredText which takes a span and the input text.
The TokenizerME is able to output the probabilities for the detected
tokens. The getTokenProbabilities method must be called directly
after one of the tokenize methods was called.
<programlisting language="java">
<![CDATA[
TokenizerME tokenizer = ...
String tokens[] = tokenizer.tokenize(...);
double tokenProbs[] = tokenizer.getTokenProbabilities();]]>
</programlisting>
The tokenProbs array now contains one double value per token, the
value is between 0 and 1, where 1 is the highest possible probability
and 0 the lowest possible probability.
</para>
</section>
</section>
<section id="tools.tokenizer.training">
<title>Tokenizer Training</title>
<section id="tools.tokenizer.training.tool">
<title>Training Tool</title>
<para>
OpenNLP has a command line tool which is used to train the models
available from the model download page on various corpora. The data
can be converted to the OpenNLP Tokenizer training format or used directly.
The OpenNLP format contains one sentence per line. Tokens are either separated by a
whitespace or by a special &lt;SPLIT&gt; tag.
The following sample shows the sample from above in the correct format.
<screen>
<![CDATA[
Pierre Vinken<SPLIT>, 61 years old<SPLIT>, will join the board as a nonexecutive director Nov. 29<SPLIT>.
Mr. Vinken is chairman of Elsevier N.V.<SPLIT>, the Dutch publishing group<SPLIT>.
Rudolph Agnew<SPLIT>, 55 years old and former chairman of Consolidated Gold Fields PLC<SPLIT>,
was named a nonexecutive director of this British industrial conglomerate<SPLIT>.]]>
</screen>
Usage of the tool:
<screen>
<![CDATA[
$ opennlp TokenizerTrainer
Usage: opennlp TokenizerTrainer[.namefinder|.conllx|.pos] [-abbDict path] \
[-alphaNumOpt isAlphaNumOpt] [-params paramsFile] [-iterations num] \
[-cutoff num] -model modelFile -lang language -data sampleData \
[-encoding charsetName]
Arguments description:
-abbDict path
abbreviation dictionary in XML format.
-alphaNumOpt isAlphaNumOpt
Optimization flag to skip alpha numeric tokens for further tokenization
-params paramsFile
training parameters file.
-iterations num
number of training iterations, ignored if -params is used.
-cutoff num
minimal number of times a feature must be seen, ignored if -params is used.
-model modelFile
output model file.
-lang language
language which is being processed.
-data sampleData
data to be used, usually a file name.
-encoding charsetName
encoding for reading and writing text, if absent the system default is used.]]>
</screen>
To train the english tokenizer use the following command:
<screen>
<![CDATA[
$ opennlp TokenizerTrainer -model en-token.bin -alphaNumOpt -lang en -data en-token.train -encoding UTF-8
Indexing events using cutoff of 5
Computing event counts... done. 262271 events
Indexing... done.
Sorting and merging events... done. Reduced 262271 events to 59060.
Done indexing.
Incorporating indexed data for training...
done.
Number of Event Tokens: 59060
Number of Outcomes: 2
Number of Predicates: 15695
...done.
Computing model parameters...
Performing 100 iterations.
1: .. loglikelihood=-181792.40419263614 0.9614292087192255
2: .. loglikelihood=-34208.094253153664 0.9629238459456059
3: .. loglikelihood=-18784.123872910015 0.9729211388220581
4: .. loglikelihood=-13246.88162585859 0.9856103038460219
5: .. loglikelihood=-10209.262670265718 0.9894422181636552
...<skipping a bunch of iterations>...
95: .. loglikelihood=-769.2107474529454 0.999511955191386
96: .. loglikelihood=-763.8891914534009 0.999511955191386
97: .. loglikelihood=-758.6685383254891 0.9995157680414533
98: .. loglikelihood=-753.5458314695236 0.9995157680414533
99: .. loglikelihood=-748.5182305519613 0.9995157680414533
100: .. loglikelihood=-743.5830058068038 0.9995157680414533
Wrote tokenizer model.
Path: en-token.bin]]>
</screen>
</para>
</section>
<section id="tools.tokenizer.training.api">
<title>Training API</title>
<para>
The Tokenizer offers an API to train a new tokenization model. Basically three steps
are necessary to train it:
<itemizedlist>
<listitem>
<para>The application must open a sample data stream</para>
</listitem>
<listitem>
<para>Call the TokenizerME.train method</para>
</listitem>
<listitem>
<para>Save the TokenizerModel to a file or directly use it</para>
</listitem>
</itemizedlist>
The following sample code illustrates these steps:
<programlisting language="java">
<![CDATA[
Charset charset = Charset.forName("UTF-8");
ObjectStream<String> lineStream = new PlainTextByLineStream(new FileInputStream("en-sent.train"),
charset);
ObjectStream<TokenSample> sampleStream = new TokenSampleStream(lineStream);
TokenizerModel model;
try {
model = TokenizerME.train("en", sampleStream, true, TrainingParameters.defaultParams());
}
finally {
sampleStream.close();
}
OutputStream modelOut = null;
try {
modelOut = new BufferedOutputStream(new FileOutputStream(modelFile));
model.serialize(modelOut);
} finally {
if (modelOut != null)
modelOut.close();
}]]>
</programlisting>
</para>
</section>
</section>
<section id="tools.tokenizer.detokenizing">
<title>Detokenizing</title>
<para>
Detokenizing is simple the opposite of tokenization, the original non-tokenized string should
be constructed out of a token sequence. The OpenNLP implementation was created to undo the tokenization
of training data for the tokenizer. It can also be used to undo the tokenization of such a trained
tokenizer. The implementation is strictly rule based and defines how tokens should be attached
to a sentence wise character sequence.
</para>
<para>
The rule dictionary assign to every token an operation which describes how it should be attached
to one continuous character sequence.
</para>
<para>
The following rules can be assigned to a token:
<itemizedlist>
<listitem>
<para>MERGE_TO_LEFT - Merges the token to the left side.</para>
</listitem>
<listitem>
<para>MERGE_TO_RIGHT - Merges the token to the right side.</para>
</listitem>
<listitem>
<para>RIGHT_LEFT_MATCHING - Merges the token to the right side on first occurrence
and to the left side on second occurrence.</para>
</listitem>
</itemizedlist>
The following sample will illustrate how the detokenizer with a small
rule dictionary (illustration format, not the xml data format):
<programlisting>
<![CDATA[
. MERGE_TO_LEFT
" RIGHT_LEFT_MATCHING]]>
</programlisting>
The dictionary should be used to de-tokenize the following whitespace tokenized sentence:
<programlisting>
<![CDATA[
He said " This is a test " .]]>
</programlisting>
The tokens would get these tags based on the dictionary:
<programlisting>
<![CDATA[
He -> NO_OPERATION
said -> NO_OPERATION
" -> MERGE_TO_RIGHT
This -> NO_OPERATION
is -> NO_OPERATION
a -> NO_OPERATION
test -> NO_OPERATION
" -> MERGE_TO_LEFT
. -> MERGE_TO_LEFT]]>
</programlisting>
That will result in the following character sequence:
<programlisting>
<![CDATA[
He said "This is a test".]]>
</programlisting>
TODO: Add documentation about the dictionary format and how to use the API. Contributions are welcome.
</para>
<section id="tools.tokenizer.detokenizing.api">
<title>Detokenizing API</title>
<para>TODO: Write documentation about the detokenizer api. Any contributions
are very welcome. If you want to contribute please contact us on the mailing list
or comment on the jira issue <ulink url="https://issues.apache.org/jira/browse/OPENNLP-216">OPENNLP-216</ulink>.</para>
</section>
<section id="tools.tokenizer.detokenizing.dict">
<title>Detokenizer Dictionary</title>
<para>TODO: Write documentation about the detokenizer dictionary. Any contributions
are very welcome. If you want to contribute please contact us on the mailing list
or comment on the jira issue <ulink url="https://issues.apache.org/jira/browse/OPENNLP-217">OPENNLP-217</ulink>.</para>
</section>
</section>
</chapter>