mosesdecoder/moses/TargetNgramFeature.cpp

444 lines
15 KiB
C++

#include "TargetNgramFeature.h"
#include "Phrase.h"
#include "TargetPhrase.h"
#include "Hypothesis.h"
#include "ScoreComponentCollection.h"
#include "ChartHypothesis.h"
#include "util/string_piece_hash.hh"
namespace Moses {
using namespace std;
int TargetNgramState::Compare(const FFState& other) const {
const TargetNgramState& rhs = dynamic_cast<const TargetNgramState&>(other);
int result;
if (m_words.size() == rhs.m_words.size()) {
for (size_t i = 0; i < m_words.size(); ++i) {
result = Word::Compare(m_words[i],rhs.m_words[i]);
if (result != 0) return result;
}
return 0;
}
else if (m_words.size() < rhs.m_words.size()) {
for (size_t i = 0; i < m_words.size(); ++i) {
result = Word::Compare(m_words[i],rhs.m_words[i]);
if (result != 0) return result;
}
return -1;
}
else {
for (size_t i = 0; i < rhs.m_words.size(); ++i) {
result = Word::Compare(m_words[i],rhs.m_words[i]);
if (result != 0) return result;
}
return 1;
}
}
TargetNgramFeature::TargetNgramFeature(const std::string &line)
:StatefulFeatureFunction("TargetNgramFeature", 0, line)
{
std::cerr << "Initializing target ngram feature.." << std::endl;
vector<string> tokens = Tokenize(line);
//CHECK(tokens[0] == m_description);
CHECK(tokens.size() == 4);
m_factorType = Scan<FactorType>(tokens[1]);
m_n = Scan<size_t>(tokens[2]);
m_lower_ngrams = Scan<bool>(tokens[3]);
}
bool TargetNgramFeature::Load(const std::string &filePath)
{
if (filePath == "*") return true; //allow all
ifstream inFile(filePath.c_str());
if (!inFile)
{
return false;
}
std::string line;
m_vocab.insert(BOS_);
m_vocab.insert(EOS_);
while (getline(inFile, line)) {
m_vocab.insert(line);
}
inFile.close();
return true;
}
const FFState* TargetNgramFeature::EmptyHypothesisState(const InputType &/*input*/) const
{
vector<Word> bos(1,m_bos);
return new TargetNgramState(bos);
}
FFState* TargetNgramFeature::Evaluate(const Hypothesis& cur_hypo,
const FFState* prev_state,
ScoreComponentCollection* accumulator) const
{
const TargetNgramState* tnState = static_cast<const TargetNgramState*>(prev_state);
assert(tnState);
// current hypothesis target phrase
const Phrase& targetPhrase = cur_hypo.GetCurrTargetPhrase();
if (targetPhrase.GetSize() == 0) return new TargetNgramState(*tnState);
// extract all ngrams from current hypothesis
vector<Word> prev_words = tnState->GetWords();
stringstream curr_ngram;
bool skip = false;
// include lower order ngrams?
size_t smallest_n = m_n;
if (m_lower_ngrams) smallest_n = 1;
for (size_t n = m_n; n >= smallest_n; --n) { // iterate over ngram size
for (size_t i = 0; i < targetPhrase.GetSize(); ++i) {
// const string& curr_w = targetPhrase.GetWord(i).GetFactor(m_factorType)->GetString();
const StringPiece& curr_w = targetPhrase.GetWord(i).GetString(m_factorType);
if (m_vocab.size() && (FindStringPiece(m_vocab, curr_w) == m_vocab.end())) continue; // skip ngrams
if (n > 1) {
// can we build an ngram at this position? ("<s> this" --> cannot build 3gram at this position)
size_t pos_in_translation = cur_hypo.GetSize() - targetPhrase.GetSize() + i;
if (pos_in_translation < n - 2) continue; // need at least m_n - 1 words
// how many words needed from previous state?
int from_prev_state = n - (i+1);
skip = false;
if (from_prev_state > 0) {
if (prev_words.size() < from_prev_state) {
// context is too short, make new state from previous state and target phrase
vector<Word> new_prev_words;
for (size_t i = 0; i < prev_words.size(); ++i)
new_prev_words.push_back(prev_words[i]);
for (size_t i = 0; i < targetPhrase.GetSize(); ++i)
new_prev_words.push_back(targetPhrase.GetWord(i));
return new TargetNgramState(new_prev_words);
}
// add words from previous state
for (size_t j = prev_words.size()-from_prev_state; j < prev_words.size() && !skip; ++j)
appendNgram(prev_words[j], skip, curr_ngram);
}
// add words from current target phrase
int start = i - n + 1; // add m_n-1 previous words
if (start < 0) start = 0; // or less
for (size_t j = start; j < i && !skip; ++j)
appendNgram(targetPhrase.GetWord(j), skip, curr_ngram);
}
if (!skip) {
curr_ngram << curr_w;
accumulator->PlusEquals(this,curr_ngram.str(),1);
}
curr_ngram.str("");
}
}
if (cur_hypo.GetWordsBitmap().IsComplete()) {
for (size_t n = m_n; n >= smallest_n; --n) {
stringstream last_ngram;
skip = false;
for (size_t i = cur_hypo.GetSize() - n + 1; i < cur_hypo.GetSize() && !skip; ++i)
appendNgram(cur_hypo.GetWord(i), skip, last_ngram);
if (n > 1 && !skip) {
last_ngram << EOS_;
accumulator->PlusEquals(this, last_ngram.str(), 1);
}
}
return NULL;
}
// prepare new state
vector<Word> new_prev_words;
if (targetPhrase.GetSize() >= m_n-1) {
// take subset of target words
for (size_t i = targetPhrase.GetSize() - m_n + 1; i < targetPhrase.GetSize(); ++i)
new_prev_words.push_back(targetPhrase.GetWord(i));
}
else {
// take words from previous state and from target phrase
int from_prev_state = m_n - 1 - targetPhrase.GetSize();
for (size_t i = prev_words.size()-from_prev_state; i < prev_words.size(); ++i)
new_prev_words.push_back(prev_words[i]);
for (size_t i = 0; i < targetPhrase.GetSize(); ++i)
new_prev_words.push_back(targetPhrase.GetWord(i));
}
return new TargetNgramState(new_prev_words);
}
void TargetNgramFeature::appendNgram(const Word& word, bool& skip, stringstream &ngram) const {
// const string& w = word.GetFactor(m_factorType)->GetString();
const StringPiece& w = word.GetString(m_factorType);
if (m_vocab.size() && (FindStringPiece(m_vocab, w) == m_vocab.end())) skip = true;
else {
ngram << w;
ngram << ":";
}
}
FFState* TargetNgramFeature::EvaluateChart(const ChartHypothesis& cur_hypo, int featureId, ScoreComponentCollection* accumulator) const
{
vector<const Word*> contextFactor;
contextFactor.reserve(m_n);
// get index map for underlying hypotheses
const AlignmentInfo::NonTermIndexMap &nonTermIndexMap =
cur_hypo.GetCurrTargetPhrase().GetAlignNonTerm().GetNonTermIndexMap();
// loop over rule
bool makePrefix = false;
bool makeSuffix = false;
bool collectForPrefix = true;
size_t prefixTerminals = 0;
size_t suffixTerminals = 0;
bool onlyTerminals = true;
bool prev_is_NT = false;
size_t prev_subPhraseLength = 0;
for (size_t phrasePos = 0; phrasePos < cur_hypo.GetCurrTargetPhrase().GetSize(); phrasePos++)
{
// consult rule for either word or non-terminal
const Word &word = cur_hypo.GetCurrTargetPhrase().GetWord(phrasePos);
// cerr << "word: " << word << endl;
// regular word
if (!word.IsNonTerminal()) {
contextFactor.push_back(&word);
prev_is_NT = false;
if (phrasePos==0)
makePrefix = true;
if (phrasePos==cur_hypo.GetCurrTargetPhrase().GetSize()-1 || prev_is_NT)
makeSuffix = true;
// beginning/end of sentence symbol <s>,</s>?
StringPiece factorZero = word.GetString(0);
if (factorZero.compare("<s>") == 0)
prefixTerminals++;
// end of sentence symbol </s>?
else if (factorZero.compare("</s>") == 0)
suffixTerminals++;
// everything else
else {
stringstream ngram;
ngram << m_baseName;
if (m_factorType == 0)
ngram << factorZero;
else
ngram << word.GetString(m_factorType);
accumulator->SparsePlusEquals(ngram.str(), 1);
if (collectForPrefix)
prefixTerminals++;
else
suffixTerminals++;
}
}
// non-terminal, add phrase from underlying hypothesis
else if (m_n > 1)
{
// look up underlying hypothesis
size_t nonTermIndex = nonTermIndexMap[phrasePos];
const ChartHypothesis *prevHypo = cur_hypo.GetPrevHypo(nonTermIndex);
const TargetNgramChartState* prevState =
static_cast<const TargetNgramChartState*>(prevHypo->GetFFState(featureId));
size_t subPhraseLength = prevState->GetNumTargetTerminals();
// special case: rule starts with non-terminal
if (phrasePos == 0) {
if (subPhraseLength == 1) {
makePrefix = true;
++prefixTerminals;
const Word &word = prevState->GetSuffix().GetWord(0);
// cerr << "NT0 --> : " << word << endl;
contextFactor.push_back(&word);
}
else {
onlyTerminals = false;
collectForPrefix = false;
int suffixPos = prevState->GetSuffix().GetSize() - (m_n-1);
if (suffixPos < 0) suffixPos = 0; // push all words if less than order
for(;(size_t)suffixPos < prevState->GetSuffix().GetSize(); suffixPos++)
{
const Word &word = prevState->GetSuffix().GetWord(suffixPos);
// cerr << "NT0 --> : " << word << endl;
contextFactor.push_back(&word);
}
}
}
// internal non-terminal
else
{
// push its prefix
for(size_t prefixPos = 0; prefixPos < m_n-1
&& prefixPos < subPhraseLength; prefixPos++)
{
const Word &word = prevState->GetPrefix().GetWord(prefixPos);
// cerr << "NT --> " << word << endl;
contextFactor.push_back(&word);
}
if (subPhraseLength==1) {
if (collectForPrefix)
++prefixTerminals;
else
++suffixTerminals;
if (phrasePos == cur_hypo.GetCurrTargetPhrase().GetSize()-1)
makeSuffix = true;
}
else {
onlyTerminals = false;
collectForPrefix = true;
// check if something follows this NT
bool wordFollowing = (phrasePos < cur_hypo.GetCurrTargetPhrase().GetSize() - 1)? true : false;
// check if we are dealing with a large sub-phrase
if (wordFollowing && subPhraseLength > m_n - 1)
{
// clear up pending ngrams
MakePrefixNgrams(contextFactor, accumulator, prefixTerminals);
contextFactor.clear();
makePrefix = false;
makeSuffix = true;
collectForPrefix = false;
prefixTerminals = 0;
suffixTerminals = 0;
// push its suffix
size_t remainingWords = (remainingWords > m_n-1) ? m_n-1 : subPhraseLength - (m_n-1);
for(size_t suffixPos = 0; suffixPos < prevState->GetSuffix().GetSize(); suffixPos++) {
const Word &word = prevState->GetSuffix().GetWord(suffixPos);
// cerr << "NT --> : " << word << endl;
contextFactor.push_back(&word);
}
}
// subphrase can be used as suffix and as prefix for the next part
else if (wordFollowing && subPhraseLength == m_n - 1)
{
// clear up pending ngrams
MakePrefixNgrams(contextFactor, accumulator, prefixTerminals);
makePrefix = false;
makeSuffix = true;
collectForPrefix = false;
prefixTerminals = 0;
suffixTerminals = 0;
}
else if (prev_is_NT && prev_subPhraseLength > 1 && subPhraseLength > 1) {
// two NTs in a row: make transition
MakePrefixNgrams(contextFactor, accumulator, 1, m_n-2);
MakeSuffixNgrams(contextFactor, accumulator, 1, m_n-2);
makePrefix = false;
makeSuffix = false;
collectForPrefix = false;
prefixTerminals = 0;
suffixTerminals = 0;
// remove duplicates
stringstream curr_ngram;
curr_ngram << m_baseName;
curr_ngram << (*contextFactor[m_n-2]).GetString(m_factorType);
curr_ngram << ":";
curr_ngram << (*contextFactor[m_n-1]).GetString(m_factorType);
accumulator->SparseMinusEquals(curr_ngram.str(),1);
}
}
}
prev_is_NT = true;
prev_subPhraseLength = subPhraseLength;
}
}
if (m_n > 1) {
if (onlyTerminals) {
MakePrefixNgrams(contextFactor, accumulator, prefixTerminals-1);
}
else {
if (makePrefix)
MakePrefixNgrams(contextFactor, accumulator, prefixTerminals);
if (makeSuffix)
MakeSuffixNgrams(contextFactor, accumulator, suffixTerminals);
// remove duplicates
size_t size = contextFactor.size();
if (makePrefix && makeSuffix && (size <= m_n)) {
stringstream curr_ngram;
curr_ngram << m_baseName;
for (size_t i = 0; i < size; ++i) {
curr_ngram << (*contextFactor[i]).GetString(m_factorType);
if (i < size-1)
curr_ngram << ":";
}
accumulator->SparseMinusEquals(curr_ngram.str(), 1);
}
}
}
// cerr << endl;
return new TargetNgramChartState(cur_hypo, featureId, m_n);
}
void TargetNgramFeature::MakePrefixNgrams(std::vector<const Word*> &contextFactor, ScoreComponentCollection* accumulator, size_t numberOfStartPos, size_t offset) const {
stringstream ngram;
size_t size = contextFactor.size();
for (size_t k = 0; k < numberOfStartPos; ++k) {
size_t max_end = (size < m_n+k+offset)? size: m_n+k+offset;
for (size_t end_pos = 1+k+offset; end_pos < max_end; ++end_pos) {
ngram << m_baseName;
for (size_t i=k+offset; i <= end_pos; ++i) {
if (i > k+offset)
ngram << ":";
StringPiece factorZero = (*contextFactor[i]).GetString(0);
if (m_factorType == 0 || factorZero.compare("<s>") == 0 || factorZero.compare("</s>") == 0)
ngram << factorZero;
else
ngram << (*contextFactor[i]).GetString(m_factorType);
const Word w = *contextFactor[i];
}
// cerr << "p-ngram: " << ngram.str() << endl;
accumulator->SparsePlusEquals(ngram.str(), 1);
ngram.str("");
}
}
}
void TargetNgramFeature::MakeSuffixNgrams(std::vector<const Word*> &contextFactor, ScoreComponentCollection* accumulator, size_t numberOfEndPos, size_t offset) const {
stringstream ngram;
for (size_t k = 0; k < numberOfEndPos; ++k) {
size_t end_pos = contextFactor.size()-1-k-offset;
for (int start_pos=end_pos-1; (start_pos >= 0) && (end_pos-start_pos < m_n); --start_pos) {
ngram << m_baseName;
for (size_t j=start_pos; j <= end_pos; ++j){
StringPiece factorZero = (*contextFactor[j]).GetString(0);
if (m_factorType == 0 || factorZero.compare("<s>") == 0 || factorZero.compare("</s>") == 0)
ngram << factorZero;
else
ngram << (*contextFactor[j]).GetString(m_factorType);
if (j < end_pos)
ngram << ":";
}
// cerr << "s-ngram: " << ngram.str() << endl;
accumulator->SparsePlusEquals(ngram.str(), 1);
ngram.str("");
}
}
}
}