An official website of the United States government
To build robust question answering systems, we need the ability to verify whether answers to questions are truly correct, not just “good enough” in the context of imperfect QA datasets. We explore the use of natural language inference (NLI) as a way to achieve this goal, as NLI inherently requires the premise (document context) to contain all necessary information to support the hypothesis (proposed answer to the question). We leverage large pre-trained models and recent prior datasets to construct powerful question conversion and decontextualization modules, which can reformulate QA instances as premise-hypothesis pairs with very high reliability. Then, by combining standard NLI datasets with NLI examples automatically derived from QA training data, we can train NLI models to evaluate QA models’ proposed answers. We show that our approach improves the confidence estimation of a QA model across different domains, evaluated in a selective QA setting. Careful manual analysis over the predictions of our NLI model shows that it can further identify cases where the QA model produces the right answer for the wrong reason, i.e., when the answer sentence cannot address all aspects of the question.
more »
« less
Collecting Diverse Natural Language Inference Problems for Sentence Representation Evaluation
We present a large-scale collection of diverse natural language inference (NLI) datasets that help provide insight into how well a sentence representation captures distinct types of reasoning. The collection results from recasting 13 existing datasets from 7 semantic phenomena into a common NLI structure, resulting in over half a million labeled context-hypothesis pairs in total. We refer to our collection as the DNC: Diverse Natural Language Inference Collection. The DNC is available online at https://www.decomp.net, and will grow over time as additional resources are recast and added from novel sources.
more »
« less
- Award ID(s):
- 1749025
- PAR ID:
- 10111906
- Date Published:
- Journal Name:
- Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing
- Page Range / eLocation ID:
- 67 to 81
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Natural language inference (NLI) datasets (e.g., MultiNLI) were collected by soliciting hypotheses for a given premise from annotators. Such data collection led to annotation artifacts: systems can identify the premise-hypothesis relationship without observing the premise (e.g., negation in hypothesis being indicative of contradiction). We address this problem by recasting the CommitmentBank for NLI, which contains items involving reasoning over the extent to which a speaker is committed to complements of clause-embedding verbs under entailment-canceling environments (conditional, negation, modal and question). Instead of being constructed to stand in certain relationships with the premise, hypotheses in the recast CommitmentBank are the complements of the clause-embedding verb in each premise, leading to no annotation artifacts in the hypothesis. A state-of-the-art BERT-based model performs well on the CommitmentBank with 85% F1. However analysis of model behavior shows that the BERT models still do not capture the full complexity of pragmatic reasoning, nor encode some of the linguistic generalizations, highlighting room for improvement.more » « less
-
null (Ed.)Natural language inference (NLI) is an increasingly important task for natural language understanding, which requires one to infer whether a sentence entails another. However, the ability of NLI models to make pragmatic inferences remains understudied. We create an IMPlicature and PRESupposition diagnostic dataset (IMPPRES), consisting of 32K semi-automatically generated sentence pairs illustrating well-studied pragmatic inference types. We use IMPPRES to evaluate whether BERT, InferSent, and BOW NLI models trained on MultiNLI (Williams et al., 2018) learn to make pragmatic inferences. Although MultiNLI appears to contain very few pairs illustrating these inference types, we find that BERT learns to draw pragmatic inferences. It reliably treats scalar implicatures triggered by “some” as entailments. For some presupposition triggers like “only”, BERT reliably recognizes the presupposition as an entailment, even when the trigger is embedded under an entailment canceling operator like negation. BOW and InferSent show weaker evidence of pragmatic reasoning. We conclude that NLI training encourages models to learn some, but not all, pragmatic inferences.more » « less
-
Recent advances in data and information technologies have enabled extensive digital datasets to be available to decision makers throughout the life cycle of a transportation project. However, most of these data are not yet fully reused due to the challenging and time-consuming process of extracting the desired data for a specific purpose. Digital datasets are presented only in computer-readable formats and they are mostly complicated. Extracting data from complex and large data sources is significantly time-consuming and requires considerable expertise. Thus, there is a need for a user-friendly data exploration framework that allows users to present their data interests in human language. To fulfill that demand, this study employs natural language processing (NLP) techniques to develop a natural language interface (NLI) which can understand users’ intent and automatically convert their inputs in the human language into formal queries. This paper presents the results of an important task of the development of such a NLI that is to establish a method for classifying the tokens of an ad-hoc query in accordance with their semantic contribution to the corresponding formal query. The method was validated on a small test set of 30 plain English questions manually annotated by an expert. The result shows an impressive accuracy of over 95%. The token classification presented in this paper is expected to provide a fundamental means for developing an effective NLI to transportation asset databases.more » « less
-
This work introduces a natural language inference (NLI) dataset that focuses on the validity of statements in legal wills. This dataset is unique because: (a) each entailment decision requires three inputs: the statement from the will, the law, and the conditions that hold at the time of the testator’s death; and (b) the included texts are longer than the ones in current NLI datasets. We trained eight neural NLI models in this dataset. All the models achieve more than 80% macro F1 and accuracy, which indicates that neural approaches can handle this task reasonably well. However, group accuracy, a stricter evaluation measure that is calculated with a group of positive and negative examples generated from the same statement as a unit, is in mid 80s at best, which suggests that the models’ understanding of the task remains superficial. Further ablative analyses and explanation experiments indicate that all three text segments are used for prediction, but some decisions rely on semantically irrelevant tokens. This indicates that overfitting on these longer texts likely happens, and that additional research is required for this task to be solved.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.18653/v1/D18-1007
