mert software now works with different reference length policies: shortest, average, closest (default) and with case information (default is preserving case). Pay attention that both defaults are different from the previous version (which were shortest reflen and case-insensitive).

git-svn-id: https://mosesdecoder.svn.sourceforge.net/svnroot/mosesdecoder/trunk@2459 1f5c12ca-751b-0410-a591-d2e778427230

4 files changed, 42 insertions, 10 deletions

diff --git a/mert/BleuScorer.cpp b/mert/BleuScorer.cpp
index f2e83dc5c..5f6349b55 100644
--- a/mert/BleuScorer.cpp
+++ b/mert/BleuScorer.cpp
@@ -53,7 +53,7 @@ void BleuScorer::setReferenceFiles(const vector<string>& referenceFiles) {
 		string line;
 		size_t sid = 0; //sentence counter
 		while (getline(refin,line)) {
-			//§cerr << line << endl;
+			//cerr << line << endl;
 			if (i == 0) {
 				counts_t* counts = new counts_t(); //these get leaked
 				_refcounts.push_back(counts);
@@ -116,13 +116,19 @@ void BleuScorer::prepareStats(size_t sid, const string& text, ScoreStats& entry)
 			stats.push_back(mean);
 		} else if (_refLengthStrategy == BLEU_CLOSEST)  {
 			int min_diff = INT_MAX;
+			int min_idx = 0;
 			for (size_t i = 0; i < _reflengths[sid].size(); ++i) {
 				int reflength = _reflengths[sid][i];
-				if (abs(reflength-(int)length) < abs(min_diff)) {
+				if (abs(reflength-(int)length) < abs(min_diff)) { //look for the closest reference
 					min_diff = reflength-length;
-				}
+					min_idx = i;
+				}else if (abs(reflength-(int)length) == abs(min_diff)) { // if two references has the same closest length, take the shortest
+                                        if (reflength < (int)_reflengths[sid][min_idx]){
+                                        	min_idx = i;
+					}
+                                }
 			}
-			stats.push_back(length + min_diff);
+			stats.push_back(_reflengths[sid][min_idx]);
 		} else {
 			throw runtime_error("Unsupported reflength strategy");
 		}
diff --git a/mert/BleuScorer.h b/mert/BleuScorer.h
index fc5d5f332..cd2471b55 100644
--- a/mert/BleuScorer.h
+++ b/mert/BleuScorer.h
@@ -25,7 +25,26 @@ enum BleuReferenceLengthStrategy { BLEU_AVERAGE, BLEU_SHORTEST, BLEU_CLOSEST };
  **/
 class BleuScorer: public StatisticsBasedScorer {
 	public:
-		BleuScorer(const string& config = "") : StatisticsBasedScorer("BLEU",config),_refLengthStrategy(BLEU_SHORTEST) {}
+		BleuScorer(const string& config = "") : StatisticsBasedScorer("BLEU",config),_refLengthStrategy(BLEU_CLOSEST) {
+    //configure regularisation
+    static string KEY_REFLEN = "reflen";
+    static string REFLEN_AVERAGE = "average";
+    static string REFLEN_SHORTEST = "shortest";
+    static string REFLEN_CLOSEST = "closest";
+    
+    
+    string reflen = getConfig(KEY_REFLEN,REFLEN_CLOSEST);
+    if (reflen == REFLEN_AVERAGE) {
+        _refLengthStrategy = BLEU_AVERAGE;
+    } else if (reflen == REFLEN_SHORTEST) {
+        _refLengthStrategy = BLEU_SHORTEST;
+    } else if (reflen == REFLEN_CLOSEST) {
+        _refLengthStrategy = BLEU_CLOSEST;
+    } else {
+        throw runtime_error("Unknown reference length strategy: " + reflen);
+    }
+    cerr << "Using reference length strategy: " << reflen << endl;
+}
 		virtual void setReferenceFiles(const vector<string>& referenceFiles);
 		virtual void prepareStats(size_t sid, const string& text, ScoreStats& entry);
 		static const int LENGTH;	
diff --git a/mert/Scorer.h b/mert/Scorer.h
index 9be1047a2..ebe80c3de 100644
--- a/mert/Scorer.h
+++ b/mert/Scorer.h
@@ -31,7 +31,7 @@ class Scorer {
     
   public:
         
-    Scorer(const string& name, const string& config): _name(name), _scoreData(0),_preserveCase(false){ 
+    Scorer(const string& name, const string& config): _name(name), _scoreData(0), _preserveCase(true){ 
         cerr << "Scorer config string: " << config << endl;
         size_t start = 0;
         while (start < config.size()) {
@@ -196,9 +196,12 @@ class StatisticsBasedScorer : public Scorer {
     //configure regularisation
     static string KEY_TYPE = "regtype";
     static string KEY_WINDOW = "regwin";
+    static string KEY_CASE = "case";
     static string TYPE_NONE = "none";
     static string TYPE_AVERAGE = "average";
     static string TYPE_MINIMUM = "min";
+    static string TRUE = "true";
+    static string FALSE = "false";
     
     
     string type = getConfig(KEY_TYPE,TYPE_NONE);
@@ -217,8 +220,15 @@ class StatisticsBasedScorer : public Scorer {
     _regularisationWindow = atoi(window.c_str());
     cerr << "Using scorer regularisation window: " << _regularisationWindow << endl;
     
+    string preservecase = getConfig(KEY_CASE,TRUE);
+    if (preservecase == TRUE) {
+        _preserveCase = true;
+    }else if (preservecase == FALSE) {
+        _preserveCase = false;
+    }
+    cerr << "Using case preservation: " << _preserveCase << endl;
+
 
-                
   }
     ~StatisticsBasedScorer(){};
     virtual void score(const candidates_t& candidates, const diffs_t& diffs,
diff --git a/mert/mert.cpp b/mert/mert.cpp
index 203906b1b..8ddb8c5b3 100755
--- a/mert/mert.cpp
+++ b/mert/mert.cpp
@@ -242,7 +242,4 @@ int main (int argc, char **argv) {
  res<<bestP<<endl;
  
  PrintUserTime("Stopping...");
- /*
-	timer.stop("Stopping...");
-		*/
 }