mosesdecoder/phrase-extract/ScoreFeature.h
2013-06-10 18:11:55 +01:00

143 lines
4.0 KiB
C++

/***********************************************************************
Moses - factored phrase-based language decoder
Copyright (C) 2012- University of Edinburgh
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
This library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
***********************************************************************/
/**
* This contains extra features that can be added to the scorer. To add a new feature:
* 1. Implement a subclass of ScoreFeature
* 2. Updated ScoreFeatureManager.configure() to configure your feature, and usage() to
* display usage info.
* 3. Write unit tests (see ScoreFeatureTest.cpp) and regression tests
**/
#pragma once
#include <string>
#include <map>
#include <vector>
#include <boost/shared_ptr.hpp>
#include "util/exception.hh"
#include "PhraseAlignment.h"
namespace MosesTraining
{
struct MaybeLog {
MaybeLog(bool useLog, float negativeLog):
m_useLog(useLog), m_negativeLog(negativeLog) {}
inline float operator() (float a) const {
return m_useLog ? m_negativeLog*log(a) : a;
}
float m_useLog;
float m_negativeLog;
};
class ScoreFeatureArgumentException : public util::Exception
{
public:
ScoreFeatureArgumentException() throw() {
*this << "Unable to configure features: ";
}
~ScoreFeatureArgumentException() throw() {}
};
/** Passed to each feature to be used to calculate its values */
struct ScoreFeatureContext {
ScoreFeatureContext(
const PhraseAlignmentCollection &thePhrasePair,
float theCount, /* Total counts of all phrase pairs*/
const MaybeLog& theMaybeLog
) :
phrasePair(thePhrasePair),
count(theCount),
maybeLog(theMaybeLog) {
}
const PhraseAlignmentCollection& phrasePair;
float count;
MaybeLog maybeLog;
};
/**
* Abstract base class for extra features that can be added to the phrase table
* during scoring.
**/
class ScoreFeature
{
public:
/** Add the values for this feature function. */
virtual void add(const ScoreFeatureContext& context,
std::vector<float>& denseValues,
std::map<std::string,float>& sparseValues) const = 0;
/** Return true if the two phrase pairs are equal from the point of this feature. Assume
that they already compare true according to PhraseAlignment.equals()
**/
virtual bool equals(const PhraseAlignment& lhs, const PhraseAlignment& rhs) const = 0;
virtual ~ScoreFeature() {}
};
typedef boost::shared_ptr<ScoreFeature> ScoreFeaturePtr;
class ScoreFeatureManager
{
public:
ScoreFeatureManager():
m_includeSentenceId(false) {}
/** To be appended to the score usage message */
const std::string& usage() const;
/** Pass the unused command-line arguments to configure the extra features */
void configure(const std::vector<std::string> args);
/** Add all the features */
void addFeatures(const ScoreFeatureContext& context,
std::vector<float>& denseValues,
std::map<std::string,float>& sparseValues) const;
/**
* Used to tell if the PhraseAlignment should be considered the same by all
* extended features.
**/
bool equals(const PhraseAlignment& lhs, const PhraseAlignment& rhs) const;
const std::vector<ScoreFeaturePtr>& getFeatures() const {
return m_features;
}
/** Do we need to include sentence ids in phrase pairs? */
bool includeSentenceId() const {
return m_includeSentenceId;
}
private:
std::vector<ScoreFeaturePtr> m_features;
bool m_includeSentenceId;
};
}