Discriminative Latent Variable Based Classifier for Translation Error Detection

Abstract

This paper presents a discriminative latent variable model (DPLVM) based classifier for improving the translation error detection performance for statistical machine translation (SMT). It uses latent variables to carry additional information which may not be expressed by those original labels and capture more complicated dependencies between translation errors and their corresponding features to improve the classification performance. Specifically, we firstly detail the mathematical representation of the proposed DPLVM method, and then introduce features, namely word posterior probabilities (WPP), linguistic features, syntactic features. Finally, we compare the proposed method with MaxEnt and SVM classifiers to verify its effectiveness. Experimental results show that the proposed DPLVM-based classifier reduce classification error rate (CER) by relative 1.75%, 1.69%, 2.61% compared to the MaxEnt classifier, and relative 0.17%, 0.91%, 2.12% compared to the SVM classifier over three different feature combinations.