• How do I find the configuration parameters for the UIMA annotators and analysis engines found within ClearTK?
• How do I setup and install different classifiers used by ClearTK?
• What are the training arguments for the different classifiers used by ClearTK?

How do I find the configuration parameters for the UIMA annotators and analysis engines found within ClearTK?

The simplest way to find the parameters is to look for fields that start with PARAM.

Examples:

• org.cleartk.syntax.opennlp.PosTaggerAnnotator.PARAM_POSTAG_MODEL_FILE
• org.cleartk.classifier.CleartkAnnotator.PARAM_CLASSIFIER_FACTORY_CLASS_NAME
• org.cleartk.classifier.CleartkAnnotator.PARAM_DATA_WRITER_FACTORY_CLASS_NAME

For more information on how to set configuration parameters, refer to the !uimaFIT documentation for ConfigurationParameterFactory

How do I setup and install different classifiers used by ClearTK?

ClearTK makes use of several machine learning executables that need to be installed separately from ClearTK. Depending on your needs, you should only need to install a subset of these binaries.

LIBLINEAR

Download: The installation file can be downloaded from the LIBLINEAR home page. Check the relevant files in doc/dependency under the ClearTK project root for which version of LIBLINEAR to use.

Installation: Follow the instructions that come with the LIBLINEAR download. You also need to make sure that the LIBLINEAR binaries (train / predict) are in the system path when running ClearTK.

LIBSVM

Download: The installation file can be downloaded from the LIBSVM home page. Check the relevant files in doc/dependency under the ClearTK project root for which version of LIBSVM to use.

Installation: Follow the instructions that come with the LIBSVM download. You also need to make sure that the LIBSVM binaries (svm-train / svm-predict) are in the system path when running ClearTK.

SVMLight

Download: The installation file can be downloaded from the SVMLight home page. Check the relevant files in doc/dependency under the ClearTK project root for which version of SVMLight to use.

Installation Follow the instructions that come with the SVMLight download. Make sure that the SVMlight binaries (svm_learn / svm_classify) are in the system path when running ClearTK.

SVMRank

SVMrank is a different implementation of an SVMLight training algorithm, for training Ranking SVMs.

Download: The installation file can be downloaded from the SVMlight SVMrank page

Installation Follow the instructions that come with the SVMrank download. Make sure that the SVMlight binaries (svm_rank_learn / svm_rank_classify) are in the system path when running ClearTK.

TK-SVMlight

This project is not yet ready for use.

What are the training arguments for the different classifiers used by ClearTK?

Training arguments depend on the specific classifier library. For detailed argument information refer to the links below. Note: ClearTK will provide the arguments for the input training file.

Liblinear

Refer to the LIBLINEAR home page.

LibSVM

Refer to the the LIBSVM home page.

Mallet Classifier

The first argument is a factory class found in org.cleartk.classifier.factory:

• C45
• MaxEnt
• MCMaxEnt
• NaiveBayes

Each of these take their own parameters. Refer to the factory methods for details.

OpenNLP MaxEnt

OpenNLP MaxEnt takes two optional parameters iterations and cutoff. The eventfie is provided by cleartk. For more information visit the RealValueFileEventStream documentation

SVMLight

Refer to the the SVMLight home page

SVMrank

Refer to the the SVMrank home page