| 摘要 |
A computer operable to encode a string of tokens is described. The string of tokens includes a group of tokens that immediately follow a token having a first value. The Computer operates to select a token value that occurs most often in the group. The computer can then operate to convert the first string of tokens into a second string of tokens by substituting each occurrence of a token-pair unit in the first string with a single token. The token-pair unit includes a first token having the first value and a second token having the selected value. The computer then operates to identify a group of tokens in the second string. The identified group of tokens are those tokens that follow a token having the first value. The computer then operates to Huffman encode the second string. In part, this is accomplished by Huffman encoding the identified group based upon the identified group's local frequencies.
|
