NTPC: N-fold templated piped correction

Dekai Wu, Grace Ngai, Marine Carpuat

Research output: Journal article publicationConference articleAcademic researchpeer-review

Abstract

We describe a broadly-applicable conservative error correcting model, N-fold Templated Piped Correction or NTPC ("nitpick"), that consistently improves the accuracy of existing high-accuracy base models. Under circumstances where most obvious approaches actually reduce accuracy more than they improve it, NTPC nevertheless comes with little risk of accidentally degrading performance. NTPC is particularly well suited for natural language applications involving high-dimensional feature spaces, such as bracketing and disambiguation tasks, since its easily customizable template-driven learner allows efficient search over the kind of complex feature combinations that have typically eluded the base models. We show empirically that NTPC yields small but consistent accuracy gains on top of even high-performing models like boosting. We also give evidence that the various extreme design parameters in NTPC are indeed necessary for the intended operating range, even though they diverge from usual practice.
Original languageEnglish
Pages (from-to)476-486
Number of pages11
JournalLecture Notes in Artificial Intelligence (Subseries of Lecture Notes in Computer Science)
Volume3248
Publication statusPublished - 17 Oct 2005
EventFirst International Joint Conference on Natural Language Processing - IJCNLP 2004 - Hainan Island, China
Duration: 22 Mar 200424 Mar 2004

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this