The Zero-Frequency Problem in Predictive Self-Organizing Lists Data Compression

スポンサーリンク

概要

• 論文の詳細を見る

• An adaptive or real-time data compression technique always faces with the problem of the occurrence of unexpected events called "The Zero-Frequency Problem".It is the problem of estimating the likelihood of a novel event occurring which is important in statistical adaptive data compression because the occurrence of novel events impair the coding efficiency. This problem has been described in [1] and [4], and also the cstimating of the probabilities of novel events has been proposed for arithmetic coding in PPM algorithm[3]. In predictive self-organizing lists data compression(PLDC),described in [5], it is necessary to reserve a space in the code table for a novel character/word. Because all characters/words in the code table are referred by their positions, a suitable allocation of space for novel event will help improve the compression efficiency. In this presentation, the zero-frequency problem in PLDC will be studied and investigated, We then present the allocation of space for novel event by varying with times and occurrence rates of novel events. These methods are evaluated and compared with the conventional method in terms of the compression rates efficiency.

• 一般社団法人情報処理学会の論文
• 1993-03-01

著者

• WONGSAWANG Damras

  Shinshu University

• Okamoto Masayuki

  Shinshu University

関連論文

• Predictive Text Compression with Self-Organizing List
• The Zero-Frequency Problem in Predictive Self-Organizing Lists Data Compression
• BSTW Data Compression with PPM

スポンサーリンク