• DocumentCode
    1721957
  • Title

    The research and application of the Chinese machinery word segmentation algorithm based on improved PATRICIA tree dictionary

  • Author

    Abudoulikemu, Yimamuaishan

  • Author_Institution
    Inf. & Constr. Coll., Urumqi Vocational Univ., Urumqi, China
  • Volume
    3
  • fYear
    2010
  • Abstract
    Chinese mechanical word segmentation is divided into word segmentation and dictionary segmentation in accordance with segmentation approach. Word segmentation is simple but large redundant degree. However, dictionary segmentation is accurate but complex structure. In this paper, the Chinese word segmentation mechanical dictionary algorithm and the dictionary´s mechanism has been conducted the thorough research, proposed the improvement PATRICIA tree dictionary mechanical algorithm, which is combined of the word segmentation and word segmentation, and design the participle system is realized and applied in the emergency management platform full text retrieval system. Full text retrieval in emergency management platform requires a higher speed of word segmentation. Therefore, the establishment of rapid and efficient word segmentation dictionary and use a good word segmentation method has significant practical significance.
  • Keywords
    information retrieval; natural language processing; text analysis; Chinese machinery word segmentation algorithm; PATRICIA tree dictionary; emergency management platform; full text retrieval system; Algorithm design and analysis; Complexity theory; Dictionaries; Information processing; Information science; Machine learning algorithms; Signal processing algorithms; Chinese mechanical word segmentation algorithm; PATRICIA tree improvement; full text retrieval; word segmentation dictionary;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Signal Processing Systems (ICSPS), 2010 2nd International Conference on
  • Conference_Location
    Dalian
  • Print_ISBN
    978-1-4244-6892-8
  • Electronic_ISBN
    978-1-4244-6893-5
  • Type

    conf

  • DOI
    10.1109/ICSPS.2010.5555788
  • Filename
    5555788