• DocumentCode
    3508003
  • Title

    Developing the first balanced corpus for Bangla language

  • Author

    Salam, Khan Md Anwarus ; Yamada, Setsuo ; Nishino, Tetsuro

  • Author_Institution
    Univ. of Electro-Commun., Tokyo, Japan
  • fYear
    2012
  • fDate
    18-19 May 2012
  • Firstpage
    1081
  • Lastpage
    1084
  • Abstract
    The objective of this paper is to propose the development process of the first Bangladeshi National Corpus. The purpose of the study is to specify the domains to create a balanced Bangla corpus based on some selection criteria. This study focuses on three independent selection criteria: domain, time and medium. This paper also explains domain classifications and weight percentage for each domain. We also identify the prospective source of information for preparing the corpus.
  • Keywords
    natural language processing; pattern classification; Bangla language; domain classifications; domain selection criteria; first Bangladeshi national corpus; medium selection criteria; time selection criteria; weight percentage; Blogs; Business; Electronic mail; Encyclopedias; History; Patents; Writing; Bangla Language Processing; corpus development;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Informatics, Electronics & Vision (ICIEV), 2012 International Conference on
  • Conference_Location
    Dhaka
  • Print_ISBN
    978-1-4673-1153-3
  • Type

    conf

  • DOI
    10.1109/ICIEV.2012.6317356
  • Filename
    6317356