Efficient inference in large conditional random fields

  • Authors:
  • Trevor Cohn

  • Affiliations:
  • School of Informatics, University of Edinburgh, United Kingdom

  • Venue:
  • ECML'06 Proceedings of the 17th European conference on Machine Learning
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Conditional Random Fields (CRFs) are widely known to scale poorly, particularly for tasks with large numbers of states or with richly connected graphical structures. This is a consequence of inference having a time complexity which is at best quadratic in the number of states. This paper describes a novel parameterisation of the CRF which ties the majority of clique potentials, while allowing individual potentials for a subset of the labellings. This has two beneficial effects: the parameter space of the model (and thus the propensity to over-fit) is reduced, and the time complexity of training and decoding becomes sub-quadratic. On a standard natural language task, we reduce CRF training time four-fold, with no loss in accuracy. We also show how inference can be performed efficiently in richly connected graphs, in which current methods are intractable.