Abstract

Accurate lexical entailment (LE) and natural language inference (NLI) often require large quantities of costly annotations. To alleviate the need for labeled data, we introduce WikiNLI: a resource for improving model performance on NLI and LE tasks. It contains 428,899 pairs of phrases constructed from naturally annotated category hierarchies in Wikipedia. We show that we can improve strong baselines such as BERT and RoBERTa by pretraining them on WikiNLI and transferring the models on downstream tasks. We conduct systematic comparisons with phrases extracted from other knowledge bases such as WordNet and Wikidata to find that pretraining on WikiNLI gives the best performance. In addition, we construct WikiNLI in other languages, and show that pretraining on them improves performance on NLI tasks of corresponding languages.

Highlights

  • Natural language inference (NLI) is the task of classifying the relationship, such as entailment or contradiction, between sentences

  • We are interested in automatically generating a large-scale dataset from Wikipedia categories that can improve performance on both NLI and lexical entailment (LE) tasks

  • We describe how the WIKINLI dataset is constructed from Wikipedia and its principal characteristics

Read more

Summary

Introduction

Natural language inference (NLI) is the task of classifying the relationship, such as entailment or contradiction, between sentences. NLI involves rich natural language understanding capabilities, many of which relate to world knowledge. To acquire such knowledge, researchers have found benefit from external knowledge bases like WordNet (Fellbaum, 1998), FrameNet (Baker, 2014), Wikidata (Vrandecicand Krotzsch, 2014), and large-scale human-annotated datasets (Bowman et al, 2015; Williams et al, 2018; Nie et al, 2020). Researchers have found benefit from external knowledge bases like WordNet (Fellbaum, 1998), FrameNet (Baker, 2014), Wikidata (Vrandecicand Krotzsch, 2014), and large-scale human-annotated datasets (Bowman et al, 2015; Williams et al, 2018; Nie et al, 2020) Creating these resources generally requires expensive human annotation. We are interested in automatically generating a large-scale dataset from Wikipedia categories that can improve performance on both NLI and lexical entailment (LE) tasks

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call