We present DART, an open domain structured DAta Record to Text generation dataset with over 82k instances (DARTs). Data-to-text annotations can be a costly process, especially when dealing with tables which are the major source of structured data and contain nontrivial structures. To this end, we propose a procedure of extracting semantic triples from tables that encodes their structures by exploiting the semantic dependencies among table headers and the table title. Our dataset construction framework effectively merged heterogeneous sources from open domain semantic parsing and spoken dialogue systems by utilizing techniques including tree ontology annotation, question-answer pair to declarative sentence conversion, and predicate unification, all with minimum post-editing. We present systematic evaluation on DART as well as new state-of-the-art results on WebNLG 2017 to show that DART (1) poses new challenges to existing data-to-text datasets and (2) facilitates out-of-domain generalization. Our data and code can be found at https://github.com/Yale-LILY/dart.

Despite impressive performance on standard benchmarks, natural language processing (NLP) models are often brittle when deployed in real-world systems. In this work, we identify challenges with evaluating NLP systems and propose a solution in the form of Robustness Gym (RG), a simple and extensible evaluation toolkit that unifies 4 standard evaluation paradigms: subpopulations, transformations, evaluation sets, and adversarial attacks. By providing a common platform for evaluation, RG enables practitioners to compare results from disparate evaluation paradigms with a single click, and to easily develop and share novel evaluation methods using a built-in set of abstractions. RG is under active development and we welcome feedback & contributions from the community.

Named entity linking (NEL) or mapping “strings” to “things” in a knowledge base is a fundamental preprocessing step in systems that require knowledge of entities such as information extraction and question answering. In this work, we lay out and investigate two challenges faced by individuals or organizations building NEL systems. Can they directly use an off-the-shelf system? If not, how easily can such a system be repurposed for their use case? First, we conduct a study of off-the-shelf commercial and academic NEL systems. We find that most systems struggle to link rare entities, with commercial solutions lagging their academic counterparts by 10%+. Second, for a use case where the NEL model is used in a sports question-answering (QA) system, we investigate how to close the loop in our analysis by repurposing the best off-the-shelf model (Bootleg) to correct sport-related errors. We show how tailoring a simple technique for patching models using weak labeling can provide a 25% absolute improvement in accuracy of sport-related errors.

State-of-the-art models in NLP are now predominantly based on deep neural networks that are opaque in terms of how they come to make predictions. This limitation has increased interest in designing more interpretable deep models for NLP that reveal the ‘reasoning’ behind model outputs. But work in this direction has been conducted on different datasets and tasks with correspondingly unique aims and metrics; this makes it difficult to track progress. We propose the Evaluating Rationales And Simple English Reasoning (ERASER a benchmark to advance research on interpretable models in NLP. This benchmark comprises multiple datasets and tasks for which human annotations of “rationales” (supporting evidence) have been collected. We propose several metrics that aim to capture how well the rationales provided by models align with human rationales, and also how faithful these rationales are (i.e., the degree to which provided rationales influenced the corresponding predictions). Our hope is that releasing this benchmark facilitates progress on designing more interpretable NLP systems. The benchmark, code, and documentation are available at https://www.eraserbenchmark.com/

Word embeddings derived from human-generated corpora inherit strong gender bias which can be further amplified by downstream models. Some commonly adopted debiasing approaches, including the seminal Hard Debias algorithm, apply post-processing procedures that project pre-trained word embeddings into a subspace orthogonal to an inferred gender subspace. We discover that semantic-agnostic corpus regularities such as word frequency captured by the word embeddings negatively impact the performance of these algorithms. We propose a simple but effective technique, Double Hard Debias, which purifies the word embeddings against such corpus regularities prior to inferring and removing the gender subspace. Experiments on three bias mitigation benchmarks show that our approach preserves the distributional semantics of the pre-trained word embeddings while reducing gender bias to a significantly larger degree than prior approaches.

Neural networks lack the ability to reason about qualitative physics and so cannot generalize to scenarios and tasks unseen during training. We propose ESPRIT, a framework for commonsense reasoning about qualitative physics in natural language that generates interpretable descriptions of physical events. We use a two-step approach of first identifying the pivotal physical events in an environment and then generating natural language descriptions of those events using a data-to-text approach. Our framework learns to generate explanations of how the physical simulation will causally evolve so that an agent or a human can easily reason about a solution using those interpretable descriptions. Human evaluations indicate that ESPRIT produces crucial fine-grained details and has high coverage of physical concepts compared to even human annotations. Dataset, code and documentation are available at https://github.com/salesforce/esprit.

A standard way to address different NLP problems is by first constructing a problem-specific dataset, then building a model to fit this dataset. To build the ultimate artificial intelligence, we desire a single machine that can handle diverse new problems, for which task-specific annotations are limited. We bring up textual entailment as a unified solver for such NLP problems. However, current research of textual entailment has not spilled much ink on the following questions: (i) How well does a pretrained textual entailment system generalize across domains with only a handful of domain-specific examples? and (ii) When is it worth transforming an NLP task into textual entailment? We argue that the transforming is unnecessary if we can obtain rich annotations for this task. Textual entailment really matters particularly when the target NLP task has insufficient annotations. Universal NLP can be probably achieved through different routines. In this work, we introduce Universal Few-shot textual Entailment (UFO-Entail). We demonstrate that this framework enables a pretrained entailment model to work well on new entailment domains in a few-shot setting, and show its effectiveness as a unified solver for several downstream NLP tasks such as question answering and coreference resolution when the end-task annotations are limited.

To assist human review process, we build a novel ReviewRobot to automatically assign a review score and write comments for multiple categories such as novelty and meaningful comparison. A good review needs to be knowledgeable, namely that the comments should be constructive and informative to help improve the paper; and explainable by providing detailed evidence. ReviewRobot achieves these goals via three steps: (1) We perform domain-specific Information Extraction to construct a knowledge graph (KG) from the target paper under review, a related work KG from the papers cited by the target paper, and a background KG from a large collection of previous papers in the domain. (2) By comparing these three KGs, we predict a review score and detailed structured knowledge as evidence for each review category. (3) We carefully select and generalize human review sentences into templates, and apply these templates to transform the review scores and evidence into natural language comments. Experimental results show that our review score predictor reaches 71.4%-100% accuracy. Human assessment by domain experts shows that 41.7%-70.5% of the comments generated by ReviewRobot are valid and constructive, and better than human-written ones for 20% of the time. Thus, ReviewRobot can serve as an assistant for paper reviewers, program chairs and authors.

Deep learning models perform poorly on tasks that require commonsense reasoning, which often necessitates some form of world-knowledge or reasoning over information not immediately present in the input. We collect human explanations for commonsense reasoning in the form of natural language sequences and highlighted annotations in a new dataset called Common Sense Explanations (CoS-E). We use CoS-E to train language models to automatically generate explanations that can be used during training and inference in a novel Commonsense Auto-Generated Explanation (CAGE) framework. CAGE improves the state-of-the-art by 10% on the challenging CommonsenseQA task. We further study commonsense reasoning in DNNs using both human and auto-generated explanations including transfer to out-of-domain tasks. Empirical results indicate that we can effectively leverage language models for commonsense reasoning.

Visual Question Answering (VQA) is a well-known and challenging task that requires systems to jointly reason about natural language and vision. Deep learning models in various forms have been the standard for solving VQA. However, some of these VQA models are better at certain types of image-question pairs than other models. Ensembling VQA models intelligently to leverage their diverse expertise is, therefore, advantageous. Stacking With Auxiliary Features (SWAF) is an intelligent ensembling technique which learns to combine the results of multiple models using features of the current problem as context. We propose four categories of auxiliary features for ensembling for VQA. Three out of the four categories of features can be inferred from an image-question pair and do not require querying the component models. The fourth category of auxiliary features uses model-specific explanations. In this paper, we describe how we use these various categories of auxiliary features to improve performance for VQA. Using SWAF to effectively ensemble three recent systems, we obtain a new state-of-the-art. Our work also highlights the advantages of explainable AI models.

Linking spans of natural language text to concepts in a structured source is an important task for many problems. It allows intelligent systems to leverage rich knowledge available in those sources (such as concept properties and relations) to enhance the semantics of the mentions of these concepts in text. In the medical domain, it is common to link text spans to medical concepts in large, curated knowledge repositories such as the Unified Medical Language System. Different approaches have different strengths: some are precision-oriented, some recall-oriented; some better at considering context but more prone to hallucination. The variety of techniques suggests that ensembling could outperform component technologies at this task. In this paper, we describe our process for building a Stacking ensemble using additional, auxiliary features for Entity Linking in the medical domain. We report experiments that show that naive ensembling does not always outperform component Entity Linking systems, that stacking usually outperforms naive ensembling, and that auxiliary features added to the stacker further improve its performance on three distinct datasets. Our best model produces state-of-the-art results on several medical datasets.