Matches in SemOpenAlex for { <https://semopenalex.org/work/W3200930617> ?p ?o ?g. }
- W3200930617 abstract "Several neural-based metrics have been recently proposed to evaluate machine translation quality. However, all of them resort to point estimates, which provide limited information at segment level. This is made worse as they are trained on noisy, biased and scarce human judgements, often resulting in unreliable quality predictions. In this paper, we introduce uncertainty-aware MT evaluation and analyze the trustworthiness of the predicted quality. We combine the COMET framework with two uncertainty estimation methods, Monte Carlo dropout and deep ensembles, to obtain quality scores along with confidence intervals. We compare the performance of our uncertainty-aware MT evaluation methods across multiple language pairs from the QT21 dataset and the WMT20 metrics task, augmented with MQM annotations. We experiment with varying numbers of references and further discuss the usefulness of uncertainty-aware quality estimation (without references) to flag possibly critical translation mistakes." @default.
- W3200930617 created "2021-09-27" @default.
- W3200930617 creator A5004372721 @default.
- W3200930617 creator A5017626161 @default.
- W3200930617 creator A5039347839 @default.
- W3200930617 creator A5051693368 @default.
- W3200930617 date "2021-09-13" @default.
- W3200930617 modified "2023-09-27" @default.
- W3200930617 title "Uncertainty-Aware Machine Translation Evaluation" @default.
- W3200930617 cites W1534477342 @default.
- W3200930617 cites W1618905105 @default.
- W3200930617 cites W1654441844 @default.
- W3200930617 cites W1995945562 @default.
- W3200930617 cites W2069266605 @default.
- W3200930617 cites W2095705004 @default.
- W3200930617 cites W2098824882 @default.
- W3200930617 cites W2101105183 @default.
- W3200930617 cites W2108677974 @default.
- W3200930617 cites W2117670920 @default.
- W3200930617 cites W2117897510 @default.
- W3200930617 cites W2133459682 @default.
- W3200930617 cites W2137556846 @default.
- W3200930617 cites W2144746247 @default.
- W3200930617 cites W2149327368 @default.
- W3200930617 cites W2158840489 @default.
- W3200930617 cites W2167433878 @default.
- W3200930617 cites W222053410 @default.
- W3200930617 cites W2250342921 @default.
- W3200930617 cites W2252166243 @default.
- W3200930617 cites W2254249950 @default.
- W3200930617 cites W2575629043 @default.
- W3200930617 cites W2792388013 @default.
- W3200930617 cites W2806471870 @default.
- W3200930617 cites W2894218541 @default.
- W3200930617 cites W2900013662 @default.
- W3200930617 cites W2903275493 @default.
- W3200930617 cites W2907176385 @default.
- W3200930617 cites W2953072129 @default.
- W3200930617 cites W2963081736 @default.
- W3200930617 cites W2963215553 @default.
- W3200930617 cites W2963238274 @default.
- W3200930617 cites W2963366552 @default.
- W3200930617 cites W2963832833 @default.
- W3200930617 cites W2963913268 @default.
- W3200930617 cites W2964059111 @default.
- W3200930617 cites W2964144363 @default.
- W3200930617 cites W2964212410 @default.
- W3200930617 cites W2970971315 @default.
- W3200930617 cites W2971302374 @default.
- W3200930617 cites W2995464762 @default.
- W3200930617 cites W2996403597 @default.
- W3200930617 cites W2997129641 @default.
- W3200930617 cites W3034776473 @default.
- W3200930617 cites W3035252911 @default.
- W3200930617 cites W3037939899 @default.
- W3200930617 cites W3047281852 @default.
- W3200930617 cites W3081828086 @default.
- W3200930617 cites W3084095723 @default.
- W3200930617 cites W3099942180 @default.
- W3200930617 cites W3103450644 @default.
- W3200930617 cites W3104278118 @default.
- W3200930617 cites W3104939451 @default.
- W3200930617 cites W3105214104 @default.
- W3200930617 cites W3118824058 @default.
- W3200930617 cites W3119878165 @default.
- W3200930617 cites W3123791752 @default.
- W3200930617 cites W3134774296 @default.
- W3200930617 cites W3154806625 @default.
- W3200930617 cites W3159892921 @default.
- W3200930617 cites W3118587867 @default.
- W3200930617 hasPublicationYear "2021" @default.
- W3200930617 type Work @default.
- W3200930617 sameAs 3200930617 @default.
- W3200930617 citedByCount "0" @default.
- W3200930617 crossrefType "posted-content" @default.
- W3200930617 hasAuthorship W3200930617A5004372721 @default.
- W3200930617 hasAuthorship W3200930617A5017626161 @default.
- W3200930617 hasAuthorship W3200930617A5039347839 @default.
- W3200930617 hasAuthorship W3200930617A5051693368 @default.
- W3200930617 hasConcept C104317684 @default.
- W3200930617 hasConcept C105580179 @default.
- W3200930617 hasConcept C105795698 @default.
- W3200930617 hasConcept C111472728 @default.
- W3200930617 hasConcept C119857082 @default.
- W3200930617 hasConcept C124101348 @default.
- W3200930617 hasConcept C138885662 @default.
- W3200930617 hasConcept C149364088 @default.
- W3200930617 hasConcept C153701036 @default.
- W3200930617 hasConcept C154945302 @default.
- W3200930617 hasConcept C162324750 @default.
- W3200930617 hasConcept C185592680 @default.
- W3200930617 hasConcept C187736073 @default.
- W3200930617 hasConcept C19499675 @default.
- W3200930617 hasConcept C203005215 @default.
- W3200930617 hasConcept C2524010 @default.
- W3200930617 hasConcept C2776145597 @default.
- W3200930617 hasConcept C2779530757 @default.
- W3200930617 hasConcept C2780451532 @default.
- W3200930617 hasConcept C28719098 @default.
- W3200930617 hasConcept C33923547 @default.