In MT evaluation, pairwise comparisons are conducted to identify the better system. In conducting the comparison, the experimenter must allocate a budget to collect Direct Assessment (DA) judgments. We provide a cost effective way to spend the budget, but show that typical budget sizes often do not allow for solid comparison. Taking the perspective that the basis of solid comparison is in achieving statistical significance, we study the power (rate of achieving significance) on a large collection of pairwise DA comparisons. Due to the nature of statistical estimation, power is low for differentiating less than 1-2 DA points, and to achieve a notable increase in power requires at least 2-3x more samples. Applying variance reduction alone will not yield these gains, so we must face the reality of undetectable differences and spending increases. In this context, we propose interim testing, an “early stopping” collection procedure that yields more power per judgment collected, which adaptively focuses the budget on pairs that are borderline significant. Interim testing can achieve up to a 27% efficiency gain when spending 3x the current budget, or 18% savings at the current evaluation power.

Estimating the expected output quality of generation systems is central to NLG. This paper qualifies the notion that automatic metrics are not as good as humans in estimating system-level quality. Statistically, humans are unbiased, high variance estimators, while metrics are biased, low variance estimators. We compare these estimators by their error in pairwise prediction (which generation system is better?) using the bootstrap. Measuring this error is complicated: predictions are evaluated against noisy, human predicted labels instead of the ground truth, and metric predictions fluctuate based on the test sets they were calculated on. By applying a bias-variance-noise decomposition, we adjust this error to a noise-free, infinite test set setting. Our analysis compares the adjusted error of metrics to humans and a derived, perfect segment-level annotator, both of which are unbiased estimators dependent on the number of judgments collected. In MT, we identify two settings where metrics outperform humans due to a statistical advantage in variance: when the number of human judgments used is small, and when the quality difference between compared systems is small.

This paper presents the results of the WMT20 Metrics Shared Task. Participants were asked to score the outputs of the translation systems competing in the WMT20 News Translation Task with automatic metrics. Ten research groups submitted 27 metrics, four of which are reference-less “metrics”. In addition, we computed five baseline metrics, including sentBLEU, BLEU, TER and using the SacreBLEU scorer. All metrics were evaluated on how well they correlate at the system-, document- and segment-level with the WMT20 official human scores. We present an extensive analysis on influence of different reference translations on metric reliability, how well automatic metrics score human translations, and we also flag major discrepancies between metric and human scores when evaluating MT systems. Finally, we investigate whether we can use automatic metrics to flag incorrect human ratings.

Despite advances in open-domain dialogue systems, automatic evaluation of such systems is still a challenging problem. Traditional reference-based metrics such as BLEU are ineffective because there could be many valid responses for a given context that share no common words with reference responses. A recent work proposed Referenced metric and Unreferenced metric Blended Evaluation Routine (RUBER) to combine a learning-based metric, which predicts relatedness between a generated response and a given query, with reference-based metric; it showed high correlation with human judgments. In this paper, we explore using contextualized word embeddings to compute more accurate relatedness scores, thus better evaluation metrics. Experiments show that our evaluation metrics outperform RUBER, which is trained on static embeddings.

This paper presents the results of the WMT19 Metrics Shared Task. Participants were asked to score the outputs of the translations systems competing in the WMT19 News Translation Task with automatic metrics. 13 research groups submitted 24 metrics, 10 of which are reference-less “metrics” and constitute submissions to the joint task with WMT19 Quality Estimation Task, “QE as a Metric”. In addition, we computed 11 baseline metrics, with 8 commonly applied baselines (BLEU, SentBLEU, NIST, WER, PER, TER, CDER, and chrF) and 3 reimplementations (chrF+, sacreBLEU-BLEU, and sacreBLEU-chrF). Metrics were evaluated on the system level, how well a given metric correlates with the WMT19 official manual ranking, and segment level, how well the metric correlates with human judgements of segment quality. This year, we use direct assessment (DA) as our only form of manual evaluation.

Due to the presence of both Twitter-specific conventions and non-standard and dialectal language, Twitter presents a significant parsing challenge to current dependency parsing tools. We broaden English dependency parsing to handle social media English, particularly social media African-American English (AAE), by developing and annotating a new dataset of 500 tweets, 250 of which are in AAE, within the Universal Dependencies 2.0 framework. We describe our standards for handling Twitter- and AAE-specific features and evaluate a variety of cross-domain strategies for improving parsing with no, or very little, in-domain labeled data, including a new data synthesis approach. We analyze these methods’ impact on performance disparities between AAE and Mainstream American English tweets, and assess parsing accuracy for specific AAE lexical and syntactic features. Our annotated data and a parsing model are available at: http://slanglab.cs.umass.edu/TwitterAAE/.

Sequence to sequence (seq2seq) models are often employed in settings where the target output is natural language. However, the syntactic properties of the language generated from these models are not well understood. We explore whether such output belongs to a formal and realistic grammar, by employing the English Resource Grammar (ERG), a broad coverage, linguistically precise HPSG-based grammar of English. From a French to English parallel corpus, we analyze the parseability and grammatical constructions occurring in output from a seq2seq translation model. Over 93% of the model translations are parseable, suggesting that it learns to generate conforming to a grammar. The model has trouble learning the distribution of rarer syntactic rules, and we pinpoint several constructions that differentiate translations between the references and our model.

While language identification works well on standard texts, it performs much worse on social media language, in particular dialectal language—even for English. First, to support work on English language identification, we contribute a new dataset of tweets annotated for English versus non-English, with attention to ambiguity, code-switching, and automatic generation issues. It is randomly sampled from all public messages, avoiding biases towards pre-existing language classifiers. Second, we find that a demographic language model—which identifies messages with language similar to that used by several U.S. ethnic populations on Twitter—can be used to improve English language identification performance when combined with a traditional supervised language identifier. It increases recall with almost no loss of precision, including, surprisingly, for English messages written by non-U.S. authors. Our dataset and identifier ensemble are available online.