Doorgaan naar hoofdnavigatie Doorgaan naar zoeken Ga verder naar hoofdinhoud

Modelling the phonotactic structure of natural language words with simple recurrent networks

  • [No Value] Stoianov
  • , J Nerbonne
  • , H Bouma

    OnderzoeksoutputAcademicpeer review

    Samenvatting

    Simple Recurrent Networks (SRN) are Neural Network (connectionist) models able to process natural language. Phonotactics concerns the order of symbols in words. We continued an earlier unsuccessful trial to model the phonotactics of Dutch words with SRNs. In order to overcome the previously reported obstacles, a new method for testing the network training was developed - optimal threshold evaluation. This method is based on minimising the erroneous character prediction of a trained SRN. The network training was improved as well. The training words were presented to the network according to their frequencies, which emphasizes the more frequent sequences. The achieved results are promising and provide a base for further study.

    Originele taal-2English
    TitelCOMPUTATIONAL LINGUISTICS IN THE NETHERLANDS 1997
    RedacteurenPA Coppen, H vanHalteren, L Teunissen
    Plaats van productieAMSTERDAM
    UitgeverijRodopi
    Pagina's77-95
    Aantal pagina's19
    ISBN van geprinte versie90-420-0494-0
    StatusPublished - 1998
    Evenement8th Computational Linguistics in the Netherlands Meeting (CLIN 1997) - , Netherlands
    Duur: 1-dec.-1997 → …

    Publicatie series

    NaamLANGUAGE AND COMPUTERS : STUDIES IN PRACTICAL LINGUISTICS
    UitgeverijEDITIONS RODOPI B V
    ISSN van geprinte versie0921-5034

    Other

    Other8th Computational Linguistics in the Netherlands Meeting (CLIN 1997)
    Land/RegioNetherlands
    Periode01/12/1997 → …

    Vingerafdruk

    Duik in de onderzoeksthema's van 'Modelling the phonotactic structure of natural language words with simple recurrent networks'. Samen vormen ze een unieke vingerafdruk.

    Citeer dit