SHEF-NN: Translation Quality Estimation with Neural Networks

By Kashif Shah, Varvara Logacheva, Gustavo Paetzold, Frédéric Blain, Daniel Beck, Fethi Bougares, Lucia Specia. 2015

25 citations recorded by the source corpus.

We describe our systems for Tasks 1 and 2 of the WMT15 Shared Task on Quality Estimation. Our submissions use (i) a continuous space language model to extract additional features for Task 1 (SHEFGP, SHEF-SVM), (ii) a continuous bagof-words model to produce word embeddings as features for Task 2 (SHEF-W2V) and (iii) a combination of features produced by QuEst++ and a feature produced with word embedding models (SHEFQuEst++). Our systems outperform the baseline as well as many other submissions. The results are especially encouraging for Task 2, where our best performing system (SHEF-W2V) only uses features learned in an unsupervised fashion.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Phrase-level Quality Estimation for Machine Translation (2015, methodology): “– Shef-word2vec [22] — system that used word vector representations as features and performed labelling with a CRF model.”
MARMOT: A Toolkit for Translation Quality Estimation at the Word Level (2016, methodology): “The most widely used training models are Conditional Random Fields (CRF) (Luong et al., 2014; Shah et al., 2015) or neural networks: feed-forward (Kreutzer et al., 2015) as well as recurrent (Camargo de Souza et al., 2014).”
Recurrent Neural Network based Translation Quality Estimation (2016, methodology): “Shah et al. (2015) use continuous space language model features for sentence-level QE and word embedding features for word-level QE, in combination with other features produced by QuEst++ (Specia et al., 2015).”