Meteor scoring for Moses MERT

1 files changed, 81 insertions, 0 deletions

diff --git a/mert/MeteorScorer.h b/mert/MeteorScorer.h
new file mode 100644
index 000000000..8e55c49a3
--- /dev/null
+++ b/mert/MeteorScorer.h
@@ -0,0 +1,81 @@
+#ifndef MERT_METEOR_SCORER_H_
+#define MERT_METEOR_SCORER_H_
+
+#include <set>
+#include <string>
+#include <vector>
+
+#include <boost/thread/mutex.hpp>
+
+#include "Types.h"
+#include "StatisticsBasedScorer.h"
+
+namespace MosesTuning
+{
+
+class ofdstream;
+class ifdstream;
+class ScoreStats;
+
+/**
+ * Meteor scoring 
+ *
+ * https://github.com/mjdenkowski/meteor
+ * http://statmt.org/wmt11/pdf/WMT07.pdf
+ *
+ * Config:
+ * jar - location of meteor-*.jar (meteor-1.4.jar at time of writing)
+ * lang - optional language code (default: en)
+ * task - optional task (default: tune)
+ * m - optional quoted, space delimited module string "exact stem synonym paraphrase" (default varies by language)
+ * p - optional quoted, space delimited parameter string "alpha beta gamma delta" (default for tune: "0.5 1.0 0.5 0.5")
+ * w - optional quoted, space delimited weight string "w_exact w_stem w_synonym w_paraphrase" (default for tune: "1.0 0.5 0.5 0.5")
+ *
+ * Usage with mert-moses.pl:
+ * --mertargs="--sctype METEOR --scconfig jar:/path/to/meteor-1.4.jar"
+ */
+class MeteorScorer: public StatisticsBasedScorer
+{
+public:
+  explicit MeteorScorer(const std::string& config = "");
+  ~MeteorScorer();
+
+  virtual void setReferenceFiles(const std::vector<std::string>& referenceFiles);
+  virtual void prepareStats(std::size_t sid, const std::string& text, ScoreStats& entry);
+
+  virtual std::size_t NumberOfScores() const {
+    // From edu.cmu.meteor.scorer.MeteorStats
+    // tstLen refLen tstFuncWords refFuncWords stage1tstMatchesContent
+    // stage1refMatchesContent stage1tstMatchesFunction stage1refMatchesFunction
+    // s2tc s2rc s2tf s2rf s3tc s3rc s3tf s3rf s4tc s4rc s4tf s4rf chunks
+    // tstwordMatches refWordMatches
+    return 23;
+  }
+
+  virtual float calculateScore(const std::vector<int>& comps) const;
+
+private:
+  // Meteor and process IO
+  std::string meteor_jar;
+  std::string meteor_lang;
+  std::string meteor_task;
+  std::string meteor_m;
+  std::string meteor_p;
+  std::string meteor_w;
+  ofdstream* m_to_meteor;
+  ifdstream* m_from_meteor;
+  boost::mutex* mtx;
+
+  // data extracted from reference files
+  std::vector<std::string> m_references;
+  std::vector<std::vector<std::string> > m_multi_references;
+
+  // no copying allowed
+  MeteorScorer(const MeteorScorer&);
+  MeteorScorer& operator=(const MeteorScorer&);
+
+};
+
+}
+
+#endif // MERT_METEOR_SCORER_H_