Complementing Semantic Roles with Temporally Anchored Spatial Knowledge: Crowdsourced Annotations and Experiments

Authors

  • Alakananda Vempala University of North Texas
  • Eduardo Blanco University of North Texas

DOI:

https://doi.org/10.1609/aaai.v30i1.10331

Keywords:

spatial inference, temporally-anchored knowledge, semantic roles

Abstract

This paper presents a framework to infer spatial knowledge from semantic role representations. We infer whether entities are or are not located somewhere, and temporally anchor this spatial information. A large crowdsourcing effort on top of OntoNotes shows that these temporally-anchored spatial inferences are ubiquitous and intuitive to humans. Experimental results show that inferences can be performed automatically and semantic features bring significant improvement.

Downloads

Published

2016-03-05

How to Cite

Vempala, A., & Blanco, E. (2016). Complementing Semantic Roles with Temporally Anchored Spatial Knowledge: Crowdsourced Annotations and Experiments. Proceedings of the AAAI Conference on Artificial Intelligence, 30(1). https://doi.org/10.1609/aaai.v30i1.10331

Issue

Section

Technical Papers: NLP and Knowledge Representation