Using Syntax to Resolve NPE in English

19752

Authors

nan

Abstract

AbstractThis paper describes a novel, syntax-based system for automatic detection and resolution of Noun Phrase Ellipsis (NPE) in English. The system takes in free input English text, detects the site of nominal elision, and if present, selects potential antecedent candidates. The rules are built using the syntactic information on ellipsis and its antecedent discussed in previous theoretical linguistics literature on NPE. Additionally, we prepare a curated dataset of 337 sentences from well-known, reliable sources, containing positive and negative samples of NPE. We split this dataset into two parts, and use one part to refine our rules and the other to test the performance of our final system. We get an F1-score of 76.47% for detection and 70.27% for NPE resolution on the testset. To the best of our knowledge, ours is the first system that detects and resolves NPE in English. The curated dataset used for this task, albeit small, covers a wide variety of NPE cases and will be made public for future work.

Keywords

nan

Full Content

Proceedings of Recent Advances in Natural Language Processing, pages 534–540, Varna, Bulgaria, Sep 2–4, 2019. https://doi.org/10.26615/978-954-452-056-4_063 534 Using Syntax to Resolve NPE in English Payal Khullar Allen Antony Manish Shrivastava Language Technologies Research Centre International Institute of Information Technology, Hyderabad Gachibowli, Hyderabad, Telangana-500032 {payal.khullar@research., allen.antony@research., m.shrivastava@}iiit.ac.in Abstract This paper describes a novel, syntax-based system for automatic detection and resolu- tion of Noun Phrase Ellipsis (NPE) in En- glish. The system takes in free input English text, detects the site of nominal elision, and if present, selects potential antecedent candi- dates. The rules are built using the syntactic information on ellipsis and its antecedent dis- cussed in previous theoretical linguistics lit- erature on NPE. Additionally, we prepare a curated dataset of 337 sentences from well- known, reliable sources, containing positive and negative samples of NPE. We split this dataset into two parts, and use one part to re- fine our rules and the other to test the perfor- mance of our final system. We get an F1-score of 76.47% for detection and 70.27% for NPE resolution on the testset. To the best of our knowledge, ours is the first system that de- tects and resolves NPE in English. The curated dataset used for this task, albeit small, covers a wide variety of NPE cases and will be made public for future work. 1 Introduction Nominal Ellipsis or Noun Phrase Ellipsis (NPE, henceforth) is a type of ellipsis in linguists wherein the sub-parts of a nominal projection are elided, with the remaining projection pronounced in the overt syntax. For example in the sentence presented in (1), the noun chairs is elided at the position marked by [e]. 1. There are three chairs in the living room and two [e] in the hall. The full meaning of such a sentence can only be understood when we reconstruct the meaning of the elided part from the antecedent, which can be present in the linguistic context as in (1) or has to be retrieved from real world knowledge as in (2), where Mary’s actually means Mary’s place. 2. We are all partying at Mary’s [e]. All world languages use some or the other mecha- nism to elide redundant information and, hence, ellipses is fairly pervasive in natural language, more so in conversational settings (Langacker, 1999). While human interlocutors effectively re- solve and disambiguate any elided information in a sentence based on context and cognitive com- monsense extension (Chen, 2016), the realization of the complexity of processing of ellipsis be- comes evident when it poses a serious challenge for computational systems involved in natural lan- guage understanding. The approaches that handle ellipsis in both the- oretical linguistics and NLP are largely classified as syntactic, semantic and pragmatic (Merchant, 2010). For the current paper, we present a system that automatically detects and resolves NPE in En- glish using a syntax-driven approach. 2 Previous Work The syntax and semantics of the ellipsis phe- nomenon has been thoroughly studied in theoret- ical linguistics (Halliday and Hasan, 1976; Dal- rymple …

Content truncated to 500 words

Labeled Text

AbstractThis paper describes a novel, syntax-based system for automatic detection and resolution of Noun Phrase Ellipsis (NPE) in English. The system takes in free input English text, detects the site of nominal elision, and if present, selects potential antecedent candidates. The rules are built using the syntactic information on ellipsis and its antecedent discussed in previous theoretical linguistics literature on NPE. Additionally, we prepare a curated dataset of 337 sentences from well-known, reliable sources, containing positive and negative samples of NPE. We split this dataset into two parts, and use one part to refine our rules and the other to test the performance of our final system. We get an F1-score of 76.47% for detection and 70.27% for NPE resolution on the testset. To the best of our knowledge, ours is the first system that detects and resolves NPE in English. The curated dataset used for this task, albeit small, covers a wide variety of NPE cases and will be made public for future work.

Publication Details

Paper ID:
19752
Year:
2019
Publication:
TIPSTER
Publication Type:
conf
DOI:
nan