mirror of
https://github.com/moses-smt/mosesdecoder.git
synced 2024-09-11 19:27:11 +03:00
8fdec9bf30
For storing the word vocabulary used in computation of BLEU scores. This change will reduce the running time of extractor about 2-3 seconds (9% reduction).
92 lines
2.5 KiB
C++
92 lines
2.5 KiB
C++
#ifndef MERT_BLEU_SCORER_H_
|
|
#define MERT_BLEU_SCORER_H_
|
|
|
|
#include <ostream>
|
|
#include <string>
|
|
#include <vector>
|
|
|
|
#include "Types.h"
|
|
#include "ScoreData.h"
|
|
#include "StatisticsBasedScorer.h"
|
|
#include "ScopedVector.h"
|
|
|
|
namespace MosesTuning
|
|
{
|
|
|
|
const int kBleuNgramOrder = 4;
|
|
|
|
class NgramCounts;
|
|
class Reference;
|
|
|
|
/**
|
|
* Bleu scoring
|
|
*/
|
|
class BleuScorer: public StatisticsBasedScorer
|
|
{
|
|
public:
|
|
enum ReferenceLengthType {
|
|
AVERAGE,
|
|
CLOSEST,
|
|
SHORTEST
|
|
};
|
|
|
|
explicit BleuScorer(const std::string& config = "");
|
|
~BleuScorer();
|
|
|
|
static std::vector<float> ScoreNbestList(const std::string& scoreFile, const std::string& featureFile);
|
|
static float sentenceLevelBleuPlusOne(const std::vector<float>& stats);
|
|
|
|
virtual void setReferenceFiles(const std::vector<std::string>& referenceFiles);
|
|
virtual void prepareStats(std::size_t sid, const std::string& text, ScoreStats& entry);
|
|
virtual statscore_t calculateScore(const std::vector<int>& comps) const;
|
|
virtual std::size_t NumberOfScores() const { return 2 * kBleuNgramOrder + 1; }
|
|
|
|
int CalcReferenceLength(std::size_t sentence_id, std::size_t length);
|
|
|
|
ReferenceLengthType GetReferenceLengthType() const { return m_ref_length_type; }
|
|
void SetReferenceLengthType(ReferenceLengthType type) { m_ref_length_type = type; }
|
|
|
|
const std::vector<Reference*>& GetReferences() const { return m_references.get(); }
|
|
|
|
/**
|
|
* Count the ngrams of each type, up to the given length in the input line.
|
|
*/
|
|
std::size_t CountNgrams(const std::string& line, NgramCounts& counts, unsigned int n, bool is_testing=false);
|
|
|
|
void DumpCounts(std::ostream* os, const NgramCounts& counts) const;
|
|
|
|
bool OpenReference(const char* filename, std::size_t file_id);
|
|
|
|
// NOTE: this function is used for unit testing.
|
|
bool OpenReferenceStream(std::istream* is, std::size_t file_id);
|
|
|
|
private:
|
|
ReferenceLengthType m_ref_length_type;
|
|
|
|
// reference translations.
|
|
ScopedVector<Reference> m_references;
|
|
|
|
// no copying allowed
|
|
BleuScorer(const BleuScorer&);
|
|
BleuScorer& operator=(const BleuScorer&);
|
|
};
|
|
|
|
/** Computes sentence-level BLEU+1 score.
|
|
* This function is used in PRO.
|
|
*/
|
|
float sentenceLevelBleuPlusOne(const std::vector<float>& stats);
|
|
|
|
/** Computes sentence-level BLEU score given a background corpus.
|
|
* This function is used in batch MIRA.
|
|
*/
|
|
float sentenceLevelBackgroundBleu(const std::vector<float>& sent, const std::vector<float>& bg);
|
|
|
|
/**
|
|
* Computes plain old BLEU from a vector of stats
|
|
*/
|
|
float unsmoothedBleu(const std::vector<float>& stats);
|
|
|
|
}
|
|
|
|
#endif // MERT_BLEU_SCORER_H_
|