Hierarchical sub-sentential alignment with anymalign

We present a sub-sentential alignment algorithm that relies on association scores between words or phrases. This algorithm is inspired by previous work on alignment by recursive binary segmentation and on document clustering. We evaluate the resulting alignments on machine translation tasks and show that we can obtain state-of-the-art results, with gains up to more than 4 BLEU points compared to previous work, with a method that is simple, independent of the size of the corpus to be aligned, and directly computes symmetric alignments. This work also provides new insights regarding the use of "heuristic" alignment scores in statistical machine translation.