Advertisement

The data driven approach applied to the OSTIA algorithm

  • José Oncina
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1433)

Abstract

The OSTIA (Onward Subsequential Transducer Inference Algorithm) is an algorithm for inferring mappings between languages from input-output pairs, wich identifies in the limit any total subsequential function. It has been applied over a wide number of machine translation problems with great success. Incorporating the suggestions made in De la Higuera, Vidal and Oncina [dOV96] for automata inference, the DD-OSTIA (Data Driven OSTIA) is presented here. The experiments show a great reduction of the size of the training set needed for obtaining good models.

Keywords

Machine Translation Input String Statistical Machine Translation International Colloquium Language Understanding 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [BPPM93]
    P. F. Brown, S.A. Della Pietra, V.J. Della Pietra, and R.L. Mercer. The mathematics of statistical machine translation: Parameter estimation. Computational Linguistics, 19:263–311, 1993.Google Scholar
  2. [CC97]
    M. A. Castaño and F. Casacuberta. A connectionist approach to machine translation. In Proceedings of the Euro Speech'97, volume 1, pages 91–94, Rodas, Greece, 1997. EuroSpeech'97.Google Scholar
  3. [CGV94]
    A. Castellanos, I. Galiano, and E. Vidal. Applications of ostia to machine translation tasks. In Grammatical Inference and Applications, Lecture Notes in Artificial Intelligence, pages 93–105, Campello, Alicante, Spain, September 1994. 2nd International Colloquium on Grammatical Inference, Springer-Verlag.Google Scholar
  4. [CVO93]
    A. Castellanos, E. Vidal, and J. Oncina. Language understanding and subsequential transducer learning. In 1nd International Colloquium on Grammatical Inference, pages 11/1–11/10, Clochester, England, 1993.Google Scholar
  5. [dOV96]
    C. de la Higuera, J. Oncina, and E. Vidal. Identification of dfa: Data-dependent versus data-independent algorithms. In Grammatical Inference: Learning Syntax from Sentences, Lecture Notes in Artificial Intelligence, pages 313–325, Montpellier, France, September 1996. 3rd International Colloquium on Grammatical Inference, Springer-Verlag.Google Scholar
  6. [FLSW90]
    J. A. Feldman, G. Lakoff, A. Stolke, and S. H. Weber. Miniature language acquisition: A touchstone for cognitive science. Technical Report TR-90-009, Internationa Computer Science Institute, Berkeley, CA, USA, 1990.Google Scholar
  7. [Loz96]
    M. Lozano. Asociación de gramaticas mediante modelos ocultos de markov. Master's thesis, Facultad de Informática, Universidad Politécnica de Valencia, Valencia, Spain, 1996.Google Scholar
  8. [OGV91]
    J. Oncina, P. Gracia, and E. Vidal. Learning subsequential transducers for pattern recognition interpretation tasks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 13(3):252–264, March 1991.CrossRefGoogle Scholar
  9. [OMV96]
    J. Oncina and 96 M.A. Varó. Using domain information during the learning of a subsequential transducer. In Grammatical Inference: Learning Syntax from Sentences, Lecture Notes in Artificial Intelligence, pages 301–312, Montpellier, France, September 1996. 3rd International Colloquium on Grammatical Inference, Springer-Verlag.Google Scholar
  10. [Onc9]
    J. Oncina. Aprendizaje de lenguajes regulares y transducciones subsecuenciales. PhD thesis, Universidad Politécnica de Valencia, Valencia, Spain, 1991.Google Scholar
  11. [PLV93]
    R. Pieraccini, E. Levin, and E. Vidal. Learning how to understand language. In Proceedings of the EuroSpeech'93, volume 2, pages 448–458, Berlin, Germany, September 1993. 3rd European Conference on Speech Communication and Technology.Google Scholar
  12. [Pra98]
    F. Prat. Traducción automática en dominios restringidos: Algunos modelos estocásticos susceptibles de ser aprendidos a partir de ejemplos. PhD thesis, Universidad Politécnica cd Valencia, Valencia, Spain, 1998.Google Scholar
  13. [VPL93]
    E. Vidal, R. Pieraccini, and E. Levin. Learning associations between grammars: A new approach to natural language understanding. In Proceedings of the EuroSpeech'93, volume 2, pages 1187–1190, Berlin, Germany, September 1993. 3rd European Conference on Speech Communication and Technology.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • José Oncina
    • 1
  1. 1.Departamento de Lenguajes y Sistemas informticosUniversidad de AlicanteAlicanteSpain

Personalised recommendations