Skip to main navigation Skip to search Skip to main content

Unsupervised translation disambiguation based on maximum web bilingual relatedness: Web as lexicon

  • Pengyuan Liu*
  • , Tiejun Zhao
  • *Corresponding author for this work
  • Peking University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper regards Web as a semantic lexicon and alleviates the problem of bilingual lexical knowledge acquiring. Based on mix-language web page counts, four Web Bilingual Relatedness (WBR) measurements are built. WBR measurements are evaluated by a modified Miller-Charles' dataset and it is found that the measurement based on pointwise mutual information achieves the best performance. Furthermore, this paper presents a fully unsupervised translation disambiguation method which selects the translation to maximize the sum of WBR between translation and all context words. By testing this disambiguation method on Multilingual Chinese English Lexical Sample Task in SemEval-2007, it is found that the WBR disambiguation model based on point-wise mutual information achieves the best performance, outperforms other previous work and gets the state-of-the-art results (Pmar=0.451)

Original languageEnglish
Title of host publication6th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2009
Pages607-611
Number of pages5
DOIs
StatePublished - 2009
Event6th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2009 - Tianjin, China
Duration: 14 Aug 200916 Aug 2009

Publication series

Name6th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2009
Volume7

Conference

Conference6th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2009
Country/TerritoryChina
CityTianjin
Period14/08/0916/08/09

Keywords

  • Bilingual relatedness
  • Semantic lexicon
  • Sense disambiguation
  • Web

Fingerprint

Dive into the research topics of 'Unsupervised translation disambiguation based on maximum web bilingual relatedness: Web as lexicon'. Together they form a unique fingerprint.

Cite this