A Joint Inference Architecture for Global Coreference Clustering with Anaphoricity

  • Thomas Bögel
  • Anette Frank
Conference paper

DOI: 10.1007/978-3-642-40722-2_4

Part of the Lecture Notes in Computer Science book series (LNCS, volume 8105)
Cite this paper as:
Bögel T., Frank A. (2013) A Joint Inference Architecture for Global Coreference Clustering with Anaphoricity. In: Gurevych I., Biemann C., Zesch T. (eds) Language Processing and Knowledge in the Web. Lecture Notes in Computer Science, vol 8105. Springer, Berlin, Heidelberg

Abstract

We present an architecture for coreference resolution based on joint inference over anaphoricity and coreference, using Markov Logic Networks. Mentions are discriminatively clustered with discourse entities established by an anaphoricity classifier. Our entity-based coreference architecture is realized in a joint inference setting to compensate for erroneous anaphoricity classifications and avoids local coreference misclassifications through global consistency constraints. Defining pairwise coreference features in a global setting achieves an efficient entity-based perspective. With a small feature set we obtain a performance of 63.56% (gold mentions) on the official CoNLL 2012 data set.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Thomas Bögel
    • 1
  • Anette Frank
    • 1
  1. 1.Department of Computational LinguisticsHeidelberg UniversityGermany

Personalised recommendations