Motivations

By Matthew G. Snover, Bonnie J. Dorr, R. Schwartz, Linnea Micciulla, John I. Makhoul. 2000

The nightlife of college students marks the beginning highlights of their young adulthood. Where they choose to mingle under the dim lights are formative of their night out experiences. This aims to showcase the juxtapositions between two ‘nite out’ cultures: classic bar goers and alternat ive house show scene

Subfields: Philosophy

How researchers see this paper

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

In Philosophy

1. Jungo Kasai
2. M. V. Koroteev
3. Hermann Ney
4. Marco Turchi
5. Kyunghyun Cho
6. Matteo Negri
7. Luisa Bentivogli

Citations in the corpus

• Can Neural Machine Translation be Improved with User Feedback? (2018, methodology): “For significance tests we used approximate randomization (Clark et al., 2011), for BLEU score evaluation (lowercased) the multi-bleu script of the Moses decoder (Koehn et al., 2007), for TER computation the tercom tool (Snover et al., 2006).”
• Insertion-based Decoding with Automatically Inferred Generation Order (2019, background): “Semi-autoregressive modelling (Stern et al., 2018; Wang et al., 2018a) is a mixture of the two approaches, while largely adhering to left-to-right generation.”
• Automatic Machine Translation Evaluation in Many Languages via Zero-Shot Paraphrasing (2020, background): “Overlap can also be measured at the character level (Popović, 2015, 2017) or using edit distance (Snover et al., 2006).”
• Efforts on Machine Learning over Human-mediated Translation Edit Rate (2014, background): “One such metric, the Human Translation Edit Rate (HTER) (Snover et al., 2006), is the focus of the current submission.”
• VizSeq: a visual analysis toolkit for text generation tasks (2019, background): “VizSeq can present both text and image sources, and calculate the official BLEU, METEOR and TER metrics.”
• Assessing the Tolerance of Neural Machine Translation Systems Against Speech Recognition Errors (2017, methodology): “We use both BLEU [13] and Translation Edit Rate (TER) [14] as global evaluation metrics.”
• A Comparative Study on End-to-End Speech to Text Translation (2019, methodology): “The translation models are evaluated using the official scripts of WMT campaign, i.e. BLEU [34] computed by mteval-v13a5 and TER [35] computed by tercom6.”
• Not All Errors are Equal: Learning Text Generation Metrics using Stratified Error Synthesis (2022, background): “2002), ChrF (Popović, 2015) and TER (Snover et al., 2006)); (2) learned metrics requiring human rating data (COMET (Rei et al.”
• Referential Translation Machines for Predicting Translation Quality (2014, methodology): “2 is about predicting HTER (human-targeted translation edit rate) (Snover et al., 2006), where case insensitive translation edit rate (TER) scores obtained by TERp (Snover et al.”
• Automatic Text Evaluation through the Lens of Wasserstein Barycenters (2021, background): “, 2002; Lin, 2004; Banerjee and Lavie, 2005; Doddington, 2002; Popović, 2015), edit based (Leusch et al., 2006; Snover et al., 2006; Wang et al., 2016) and embedding based metrics (Chow et al.”
• Findings of the 2021 Conference on Machine Translation (WMT21) (2021, methodology): “Case-sensitive TER (Snover et al., 2006) and BLEU (Papineni et al.”
• Unsupervised Quality Estimation for Neural Machine Translation (2020, methodology): “QE Datasets The performance of QE systems has been typically assessed using the semi-automatic Human-mediated Translation Edit Rate (Snover et al., 2006) metric as gold standard.”
• Triple-to-Text: Converting RDF Triples into High-Quality Natural Languages via Optimizing an Inverse KL Divergence (2019, methodology): “We have adopted three widely used word-level metrics: BLEU [45], TER [47] and METEOR [5].”
• BERT: A Review of Applications in Natural Language Processing and Understanding (2021, methodology): “Methods such as TER [72] and ITER [61] take into account the semantic proximity of words and the normalization of grammatical forms.”
• Overview of the IWSLT 2017 Evaluation Campaign (2017, methodology): “manually corrected) by professional translators and then evaluated according to TER-based metrics [17].”
• Visually Grounded Word Embeddings and Richer Visual Features for Improving Multimodal Neural Machine Translation (2017, methodology): “Model Test Scores BLEU↑ METEOR↑ TER↓ Previous work [9] ResNet-50 + along (620) 36.50 55.0 43.7 [12] GoogLeNet v3 + along (620) 36.8 ± 0.8 55.8 ± 0.4 - [12] GoogLeNet v3 + along (620) + COCO + NC 37.8 ± 0.7 57.1 ± 0.2 - Baseline (B1) ResNet-50 + along (428) 36.27 53.9 43.6 (B2) DenseCap + along (428)”