QuEst - A translation quality estimation framework

By Lucia Specia, Kashif Shah, José G. C. de Souza, Trevor Cohn. 2013

238 citations recorded by the source corpus.

Subfields: Natural Language Processing, Evaluation & Benchmarks

How researchers see this paper

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

In Natural Language Processing

In Evaluation & Benchmarks

Citations in the corpus

Findings of the 2015 Workshop on Statistical Machine Translation (2015, positive): “Sentence-level baseline system: For Task 1, QUEST7 (Specia et al., 2013) was used to extract 17 MT system-independent features from the source and translation (target) files and parallel corpora:”
Findings of the 2014 Workshop on Statistical Machine Translation (2014, methodology): “The SVM regression algorithm within Q U E ST (Specia et al., 2013) 10 was applied for that with RBF kernel and grid search for parameter optimisation.”
Findings of the 2017 Conference on Machine Translation (WMT17) (2017, methodology): “The output of these three systems is integrated into an n-best list of translations hypotheses, which are scored and ranked by means of a sentence-level QE approach (Specia et al., 2013) and a statistical language model (Stolcke, 2002).”
SemEval-2016 Task 1: Semantic Textual Similarity, Monolingual and Cross-Lingual Evaluation (2016, methodology): “, 2016) proposes a model combining cross-lingual word embeddings with features from QuEst (Specia et al., 2013), a tool for machine translation quality estimation.”
FBK-UPV-UEdin participation in the WMT14 Quality Estimation shared-task (2014, methodology): “These features are extracted using the implementation provided by the QuEst framework (Specia et al., 2013).”
Multi-level Translation Quality Prediction with QuEst++ (2015, methodology): “Two open source feature extraction toolkits are available for that: A SIYA and Q U E ST 2 (Specia et al., 2013).”
An Investigation on the Effectiveness of Features for Translation Quality Estimation (2013, methodology): “In this paper we propose to use Gaussian Processes for feature selection, a technique that has proved effective in ranking features according to their discriminative power (Specia et al., 2013).”
Pushing the Limits of Translation Quality Estimation (2017, background): “The goal of quality estimation (QE) is to evaluate a translation system’s quality without access to reference translations (Blatz et al., 2004; Specia et al., 2013).”
Quality estimation for translation selection (2014, methodology): “Features We use the QuEst toolkit [Specia et al., 2013, Shah et al., 2013] to extract two feature sets for each dataset: • BL : 17 features used as baseline in the WMT shared tasks on QE.”
Learning What’s Easy: Fully Differentiable Neural Easy-First Taggers (2017, background): “The goal is to evaluate a translation system’s quality without access to reference translations (Blatz et al., 2004; Specia et al., 2013).”
Target-Centric Features for Translation Quality Estimation (2014, methodology): “To build our quality estimation system, we use and extend the QuEst framework for translation quality estimation 2 (Shah et al., 2013; Specia et al., 2013).”
Exploring Consensus in Machine Translation for Quality Estimation (2014, methodology): “The scikit-learn algorithm available in the QuEst Framework 4 (Specia et al., 2013) was used for that.”
USHEF and USAAR-USHEF participation in the WMT15 QE shared task (2015, methodology): “The ﬁrst set contains document-aware features, based on QuEst features for sentence-level QE (Specia et al., 2013; Specia et al., 2015).”
DFKI’s experimental hybrid MT system for WMT 2015 (2015, methodology): “• Basic feature set + 17 QuEst baseline features: this feature set combines the basic syntax-based feature set described above with the baseline feature set of the QuEst toolkit (Specia et al., 2013) as per WMT -13 (Bojar et al., 2013).”
SHEF-Lite: When Less is More for Translation Quality Estimation (2013, methodology): “They were extracted using the open source toolkit QuEst 1 (Specia et al., 2013).”
Bilingual Embeddings and Word Alignments for Translation Quality Estimation (2016, methodology): “A set of 17 system-independent features was developed by Specia et al. (2013) to set the base-line system for QE tasks.”
UGENT-LT3 SCATE System for Machine Translation Quality Estimation (2015, background): “Some of the additional features are based on ideas that were explored in previous work on QE, such as; context features for the target word and of POS tags, (Xiong et al., 2010), alignment context features (Bach et al., 2011) and adequacy and fluency indicators (Specia et al., 2013).”
Using on-line available sources of bilingual information for word-level machine translation quality estimation (2015, background): “4 Among the recent works on MTQE, it is worth mentioning the QuEst project (Specia et al., 2013), which sets a framework for MTQE, both at the sentence level and at the word level.”
The FBK Participation in the WMT 2016 Automatic Post-editing Shared Task (2016, methodology): “These features, obtained with the QuEst feature extractor implementation (Specia et al., 2013) are used to train a regression model that predicts the actual post-editing effort as measured by the TER between the MT-generated translation or the factored APE output and a human post-edited version.”