eTASC - Empirical Evidence for a Theoretical Approach to Semantic Components

Abstract

Our goal is the identification of meaning elements, i.e., semantic components, that contribute to rich meaning representations of verbs of change and transformation. Although the field of automatic processing of text and speech, natural language processing (NLP), has made considerable strides in the automated interpretation of typical newswire text, the flexibility and extensibility of language still cause great difficulty. Current NLP systems can only provide accurate meaning representations for phrases that are very similar to text previously seen in training data. As a result, performance of such systems on new genres and new domains drops dramatically. Breaking through this barrier to more robust and generalizable meaning representations will only be possible through major advances in our understanding of how individual words contribute to the overall construction of a sentence s interpretation in context. This requires a deeper level of meaning representation based on shared semantic components, such as that posited by the Generative Lexicon. A desired outcome is a richer broad-coverage lexical resource such as SemNet, based on VerbNet, enhanced by concepts from the Generative Lexicon, that can support the dynamic adaptation of these deeper semantic representations to new grammatical and lexical contexts. Therefore we propose theoretical research into the connections between the Generative Lexicon and VerbNet. These will provide the theoretical foundation on which we can build novel, rich semantic representations of complex changes of state, such as material transformations, object modifications, creation events and disruption events. Based on preliminary Generative Lexicon work on intrinsic change of state verbs, we will focus on the interactions between the Generative Lexicon qualia structure of the objects being affected by the change of state verbs, and the logical representations of the verbs themselves, bringing to bear elements from the Generative Lexicon and from VerbNet that create a whole that is greater than the sum of the parts.

Document Details

Document Type
DoD Grant Award
Publication Date
May 26, 2016
Source ID
HDTRA11610002

Entities

People

  • Martha Palmer

Organizations

  • Defense Threat Reduction Agency
  • University of Colorado

Tags

Readers

  • Computational Linguistics
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Machine Translation