About this document
Hierarchical Permutation Complexity in MT by miloshstanojevic is a document available to read on EtoBox.
This document proposes new metrics for evaluating word order in machine translation based on hierarchical permutation complexity. It introduces permutation trees (PETs) that reveal the atomic building blocks (primal permutations) of permutations. Tighter metrics are defined that assign lower complexity to permutations that factorize into shorter primal permutations, reflecting simpler word order. Experiments on WMT13 data show the new tight metrics often outperform baselines in ranking machine translation s
- Author
- miloshstanojevic
- Language
- EN