... Thayer, and J Uszkoreit 2008 Lattice-based Minimum Error Rate Training for Statistical Machine Translation In EMNLP, Honolulu, Hawaii, USA H Mi, L Huang, and Q Liu 2008 Forest-Based Translation ... two hypotheses E and E , P (E|F ) is the probability model, and G is the space of translations (N -best list, lattice, or a hypergraph) MBR decoding for translation can be performed by reranking ... and candidate length If E and E are the reference and the candidate translations respectively, this linear function is given by: Figure 2: Transformation of a hypergraph into a factor graph and...