Abstract

Several corpora have been annotated with negation scope—the set of words whose meaning is negated by a cue like the word “not”—leading to the development of classifiers that detect negation scope with high accuracy. We show that for nearly all of these corpora, this high accuracy can be attributed to a single fact: they frequently annotate negation scope as a single span of text delimited by punctuation. For negation scopes not of this form, detection accuracy is low and under-sampling the easy training examples does not substantially improve accuracy. We demonstrate that this is partly an artifact of annotation guidelines, and we argue that future negation scope annotation efforts should focus on these more difficult cases.

Highlights

  • Textual negation scope is the largest span affected by a negation cue in a negative sentence (Morante and Daelemans, 2012).1 For example, given the marker not in (1), its scope is use the 56k conextant modem.2(1) I do not [use the 56k conextant modem] since I have cable access for the internetFancellu et al (2016) recently presented a model that detects negation scope with state-of-the-art accuracy on the Sherlock Holmes corpus, which has been annotated for this task (SHERLOCK; Morante and Daelemans, 2012)

  • We experiment with the Chinese Negation and Speculation (CNeSp) corpus (Zhou, 2015), which consisting of three subcorpora: product reviews (PRODUCT), financial articles (FINANCIAL) and computer-related articles (SCIENTIFIC)

  • We compute the percentage of correct scopes (PCS), the proportion of negation scopes that we fully and exactly match in the test corpus

Read more

Summary

Introduction

Fancellu et al (2016) recently presented a model that detects negation scope with state-of-the-art accuracy on the Sherlock Holmes corpus, which has been annotated for this task (SHERLOCK; Morante and Daelemans, 2012). We confirm that it is state-of-the-art, we show that it can be improved by making joint predictions for all words, incorporating an insight from Morante et al (2008) that classifiers tend to leave gaps in what should otherwise be a continuous prediction. We accomplish this with a sequence model over the predictions

Methods
Results
Conclusion

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

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.