SEARCH

SEARCH BY CITATION

Keywords:

  • natural language processing;
  • Hidden Markov models;
  • stochastic tagging;
  • rule-based tagging

Abstract

Presented here is a brief state-of-the-art account on part-of-speech (POS) tagging. POS tagging is an essential preprocessing task for many natural language processing goals and applications. Some POS tagging approaches make use of annotated corpora to train computational models to perform the task with minimal human intervention. Rule-based and stochastic methods have been successful, attaining accuracies of 96–97%. Representative approaches of these two methodologies are discussed. WIREs Comp Stat 2012, 4:107–113. doi: 10.1002/wics.195

For further resources related to this article, please visit the WIREs website.