Application of pattern recognition to text compression

The proposed algorithm is based on the rule that when a pattern is recognised in a textual material that was previously considered as disordered, the compression ratio achieved is optimal. The recognised pattern becomes the basis for the codebook upon which an algorithm can operate. The rule based pattern recognition algorithm (RB) operates in three stages. In the first stage, the text is encoded to developed a codebook by applying Welch's algorithm, commonly known as the LZW algorithm, a variation of LZ method. In the second stage, the RB algorithm is used to compress the text, by using Zadeh's rule based logics. In the third stage, the RB algorithm decompresses the compressed text.<<ETX>>