Package de.fu_berlin.ties.filter

This packages provides generic filtering and rewriting functionality.

See:
          Description

Interface Summary
DocumentRewriter An interface for classes that rewrite documents, for example for adding semantic information.
ElementFilter Interface for classes that decide whether or not to accept elements in XML documents.
Oracle An oracle decides whether ElementFilters should match elements.
SkipHandler A handler that is called by FilteringTokenWalker whenever some tokens are skipped.
 

Class Summary
EmbeddingElements Determines which elements in a document contain extractions (or parts of extractions).
FilterEvaluator A static class that allows to evaluate the precision, recall, and F1 measure of an element filter.
FilteringTokenWalker A token walker that only invokes a provided TokenProcessor on the subset of tokens that are children of an element accepted by a provided ElementFilter.
PredictionRewriter A prediction rewriter uses predictions from another process (e.g. named entities) and stores them as XML elements to provide additional semantic information.
PredictionRewriter2 A variant of the prediction rewriter that uses predictions from another process (e.g. named entities) to provide additional semantic information.
RepresentationFilter A trainable filter that uses a Representationto convert elements into feature vectors.
TrainableFilter An abstract filter that uses a trainable classifier for training.
TrainableFilteringTokenWalker A filtering token walker that can be trained.
 

Package de.fu_berlin.ties.filter Description

This packages provides generic filtering and rewriting functionality. It also provides implementations for sentence filtering (the first step of a double classifiation approach to information extraction) and prediction-based document rewriting ("weakly hierarchical mode").



Copyright © 2003-2007 Christian Siefkes. All Rights Reserved.