Quality Estimation (QE) is the task of automatically predicting Machine Translation quality in the absence of reference translations, making it applicable in real-time settings, such as translating online social media conversations. Recent success in QE stems from the use of multilingual pre-trained representations, where very large models lead to impressive results. However, the inference time, disk and memory requirements of such models do not allow for wide usage in the real world. Models trained on distilled pre-trained representations remain prohibitively large for many usage scenarios. We instead propose to directly transfer knowledge from a strong QE teacher model to a much smaller model with a different, shallower architecture. We show that this approach, in combination with data augmentation, leads to light-weight QE models that perform competitively with distilled pre-trained representations with 8x fewer parameters.
Online political advertising is a central aspect of modern election campaigning for influencing public opinion. Computational analysis of political ads is of utmost importance in political science to understand the characteristics of digital campaigning. It is also important in computational linguistics to study features of political discourse and communication on a large scale. In this work, we present the first computational study on online political ads with the aim to (1) infer the political ideology of an ad sponsor; and (2) identify whether the sponsor is an official political party or a third-party organization. We develop two new large datasets for the two tasks consisting of ads from the U.S.. Evaluation results show that our approach that combines textual and visual information from pre-trained neural models outperforms a state-of-the-art method for generic commercial ad classification. Finally, we provide an in-depth analysis of the limitations of our best-performing models and linguistic analysis to study the characteristics of political ads discourse.
Natural language processing (NLP) methods for analyzing legal text offer legal scholars and practitioners a range of tools allowing to empirically analyze law on a large scale. However, researchers seem to struggle when it comes to identifying ethical limits to using NLP systems for acquiring genuine insights both about the law and the systems' predictive capacity. In this paper we set out a number of ways in which to think systematically about such issues. We place emphasis on three crucial normative parameters which have, to the best of our knowledge, been underestimated by current debates: (a) the importance of academic freedom, (b) the existence of a wide diversity of legal and ethical norms domestically but even more so internationally and (c) the threat of moralism in research related to computational law. For each of these three parameters we provide specific recommendations for the legal NLP community. Our discussion is structured around the study of a real-life scenario that has prompted recent debate in the legal NLP research community.
Neural network architectures in natural language processing often use attention mechanisms to produce probability distributions over input token representations. Attention has empirically been demonstrated to improve performance in various tasks, while its weights have been extensively used as explanations for model predictions. Recent studies (Jain and Wallace, 2019; Serrano and Smith, 2019; Wiegreffe and Pinter, 2019) have showed that it cannot generally be considered as a faithful explanation (Jacovi and Goldberg, 2020) across encoders and tasks. In this paper, we seek to improve the faithfulness of attention-based explanations for text classification. We achieve this by proposing a new family of Task-Scaling (TaSc) mechanisms that learn task-specific non-contextualised information to scale the original attention weights. Evaluation tests for explanation faithfulness, show that the three proposed variants of TaSc improve attention-based explanations across two attention mechanisms, five encoders and five text classification datasets without sacrificing predictive performance. Finally, we demonstrate that TaSc consistently provides more faithful attention-based explanations compared to three widely-used interpretability techniques.
Active Learning (AL) is a method to iteratively select data for annotation from a pool of unlabeled data, aiming to achieve better model performance than random selection. Previous AL approaches in Natural Language Processing (NLP) have been limited to either task-specific models that are trained from scratch at each iteration using only the labeled data at hand or using off-the-shelf pretrained language models (LMs) that are not adapted effectively to the downstream task. In this paper, we address these limitations by introducing BALM; Bayesian Active Learning with pretrained language Models. We first propose to adapt the pretrained LM to the downstream task by continuing training with all the available unlabeled data and then use it for AL. We also suggest a simple yet effective fine-tuning method to ensure that the adapted LM is properly trained in both low and high resource scenarios during AL. We finally apply Monte Carlo dropout to the downstream model to obtain well-calibrated confidence scores for data selection with uncertainty sampling. Our experiments in five standard natural language understanding tasks demonstrate that BALM provides substantial data efficiency improvements compared to various combinations of acquisition functions, models and fine-tuning methods proposed in recent AL literature.
Despite the high accuracy of pretrained transformer networks in text classification, a persisting issue is their significant complexity that makes them hard to interpret. Recent research has focused on developing feature scoring methods for identifying which parts of the input are most important for the model to make a particular prediction and use it as an explanation (i.e. rationale). A limitation of these approaches is that they assume that a particular feature scoring method should be used across all instances in a dataset using a predefined fixed length, which might not be optimal across all instances. To address this, we propose a method for extracting variable-length explanations using a set of different feature scoring methods at instance-level. Our method is inspired by word erasure approaches which assume that the most faithful rationale for a prediction should be the one with the highest divergence between the model's output distribution using the full text and the text after removing the rationale for a particular instance. Evaluation on four standard text classification datasets shows that our method consistently provides more faithful explanations compared to previous fixed-length and fixed-feature scoring methods for rationale extraction.