Rule Base with Frequent Bit Pattern and Enhanced k-Medoid Algorithm for the Evaluation of Lossless Data Compression.

Nishad P.M., Dr. N. Nalayini


This paper presents a study of various lossless compression algorithms; to test the performance and the ability of compression of each algorithm based on ten different parameters. For evaluation the compression ratios of each algorithm on different parameters are processed. To classify the algorithms based on the compression ratio, rule base is constructed to mine with frequent bit pattern to analyze the variations in various compression algorithms. Also, enhanced K- Medoid clustering is used to cluster the various data compression algorithms based on various parameters. The cluster falls dissentingly high to low after the enhancement. The framed rule base consists of 1,048,576 rules, which is used to evaluate the compression algorithm. Two hundred and eleven Compression algorithms are used for this study. The experimental result shows only few algorithm satisfies the range “High” for more number of parameters.


Keywords: Lossless compression, parameters, compression ratio, rule mining, frequent bit pattern, K–Medoid, clustering.

Full Text:




  • There are currently no refbacks.

Copyright (c) 2016 International Journal of Advanced Research in Computer Science