Och (2003) proposed using a log-linear model to incorporate multiple features for translation, and proposed a minimum error rate training (MERT) method to train the feature weights to optimize a desirable translation metric.

(2009) improved a syntactic SMT system by adding as many as ten thousand syntactic features, and used Margin Infused Relaxed Algorithm (MIRA) to train the feature weights .

The feature weights are trained on a tuning set with 2010 sentences using MIRA.

The development data utilized to tune the feature weights of our decoder is NIST’03 evaluation set, and test sets are NIST’05 and NIST’08 evaluation sets.

Therefore, we can alternatively update graph-based consensus features and feature weights in the log-linear model.

The decoder then adds the new features and retrains all the feature weights by Minimum Error Rate Training (MERT) (Och, 2003).

The decoder with new feature weights then provides new n-best candidates and their posteriors for constructing another consensus graph, which in turn gives rise to next round of

where 1/) is the feature vector, A is the feature weights , and H (f) is the set of translation hypotheses in the search space.

Before elaborating how the graph model of consensus is constructed for both a decoder and N-best output re-ranking in section 5, we will describe how the consensus features and their feature weights can be trained in a semi-supervised way, in section 4.

the training epoch (x-axis) and parsing feature weights (in legend).

We have some parameters to tune: parsing feature weight 0p, beam size, and training epoch.

Figure 2 shows the F1 scores of the proposed model (SegTagDep) on CTB-Sc-l with respect to the training epoch and different parsing feature weights , where “Seg”, “Tag”, and “Dep” respectively denote the F1 scores of word segmentation, POS tagging, and dependency parsing.