• DocumentCode
    1732834
  • Title

    Training database preparation for LTS rules training process with Wagon

  • Author

    Vasek, Matúa ; Rozinaj, Gregor ; Rybárová, Renata

  • Author_Institution
    Inst. of Telecommun., Slovak Univ. of Technol., Bratislava, Slovakia
  • fYear
    2011
  • Firstpage
    221
  • Lastpage
    224
  • Abstract
    This article describes a training database preparation for letter-to-sound (LTS) rules training process with Wagon. In some languages LTS rules can be simply handwritten. Slovak language is not that case. Slovak language includes a large number of special lingual phenomena. That is the reason, why automatic rule training is required and an automatic training system, which strongly depends on database is used. Training data have to be transformed and formatted into correct format. The whole process can be described in several steps. By following these steps the final product, which is suitable for training procedure, is reached.
  • Keywords
    database management systems; natural language processing; speech synthesis; training; LTS rules training process; Slovak language; Wagon; letter-to-sound rules training process; special lingual phenomena; training database preparation; Databases; Dictionaries; Regression tree analysis; Speech; Speech synthesis; Training; Training data; LTS rules; Speech synthesis; Wagon; training set;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    ELMAR, 2011 Proceedings
  • Conference_Location
    Zadar
  • ISSN
    1334-2630
  • Print_ISBN
    978-1-61284-949-2
  • Electronic_ISBN
    1334-2630
  • Type

    conf

  • Filename
    6044290