Multi-view discriminative sequential learning

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Authors

Discriminative learning techniques for sequential data have proven to be more effective than generative models for named entity recognition, information extraction, and other tasks of discrimination. However, semi-supervised learning mechanisms that utilize inexpensive unlabeled sequences in addition to few labeled sequences - such as the Baum-Welch algorithm - are available only for generative models. The multi-view approach is based on the principle of maximizing the consensus among multiple independent hypotheses; we develop this principle into a semi-supervised hidden Markov perceptron, and a semi-supervised hidden Markov support vector learning algorithm. Experiments reveal that the resulting procedures utilize unlabeled data effectively and discriminate more accurately than their purely supervised counterparts.

Original languageEnglish
Title of host publicationMachine Learning: ECML 2005 : 16th European Conference on Machine Learning
Number of pages12
PublisherSpringer
Publication date01.01.2005
Pages60-71
ISBN (print)978-3-540-29243-2
ISBN (electronic)978-3-540-31692-3
DOIs
Publication statusPublished - 01.01.2005
Externally publishedYes
Event16th European Conference on Machine Learning - Porto, Portugal
Duration: 03.10.200507.10.2005
Conference number: 16
http://www.inescporto.pt/~jgama/ecmlpkdd05/

Bibliographical note

Funding Information:
Thanks to NICT for their support, Takayuki Kuribayashi for providing native judgments, and Marcus Dickinson for comments on an early draft.

    Research areas

  • Informatics - Unlabeled Data, Neural Information Processing System, Name Entity Recognition, Entity Recognition, Word Sense Disambiguation
  • Business informatics

DOI