. . "However, when document data, such as electronicized Japanese, Chinese, etc. are compressed in units of words, first, it is necessary to judge at high speed whether or not a character string inputted from a document is a word registered in a dictionary in advance." . .