|
||||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |
java.lang.Objectde.fu_berlin.ties.extract.amend.BeginEndReextractor
public class BeginEndReextractor
A second level for the BeginEndStrategy
,
similar to the ELIE/2 system by Finn and Kushmerick.
Nested Class Summary | |
---|---|
static class |
BeginEndReextractor.PositivePredictionsMap
An inner class used for communication with the first level ( BeginEndStrategy ). |
Field Summary | |
---|---|
static String |
CONTEXT_BEGIN_MAP
Context key for the map from token indices to positive predictions of the begin classifier. |
static String |
CONTEXT_END_MAP
Context key for the map from token indices to positive predictions of the end classifier. |
static String |
SUFFIX_AMEND
Optional suffix that can be appended to the Classifier.CONFIG_CLASSIFIER configuration key to modify
the type of classifiers used by this instance. |
Constructor Summary | |
---|---|
BeginEndReextractor(Set<String> classNames,
TiesConfiguration config,
LengthFilter myLengthFilter)
Creates a new instance. |
Method Summary | |
---|---|
ExtractionContainer |
reextract(ExtractionContainer originalExtractions,
List<ContextDetails> contextDetails,
ContextMap context)
Re-extract all extractions in a document at the end of the document, considering the complete states of all tokens. |
String |
toString()
Returns a string representation of this object. |
void |
train(ExtractionContainer answerKeys,
List<ContextDetails> contextDetails,
ContextMap context)
Trains the re-extractor on a document. |
Methods inherited from class java.lang.Object |
---|
clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait |
Field Detail |
---|
public static final String CONTEXT_BEGIN_MAP
public static final String CONTEXT_END_MAP
public static final String SUFFIX_AMEND
Classifier.CONFIG_CLASSIFIER
configuration key to modify
the type of classifiers used by this instance.
Constructor Detail |
---|
public BeginEndReextractor(Set<String> classNames, TiesConfiguration config, LengthFilter myLengthFilter) throws ProcessingException
classNames
- a set of valid class names (String)config
- used to configure this instancemyLengthFilter
- used to determine the maximum tolerated length of
extractions
ProcessingException
- if an error occurs during initialization
of the used classifeirsMethod Detail |
---|
public ExtractionContainer reextract(ExtractionContainer originalExtractions, List<ContextDetails> contextDetails, ContextMap context) throws ProcessingException
reextract
in interface FinalReextractor
originalExtractions
- the original extractions predicted for this
documentcontextDetails
- a list of context details representing all
tokens in the documentcontext
- a map of context object provided by the combination
strategy
ProcessingException
- if an error occurs during processingpublic String toString()
toString
in class Object
public void train(ExtractionContainer answerKeys, List<ContextDetails> contextDetails, ContextMap context) throws ProcessingException
train
in interface FinalReextractor
answerKeys
- the true extractions to train for this documentcontextDetails
- a list of context details representing all
tokens in the documentcontext
- a map of context object provided by the combination
strategy
ProcessingException
- if an error occurs during processing
|
||||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |