Hash based Frequent Pattern Mining approach to Text Compression

Abstract

The paper explores the compression perspective of Data Mining. Huffman Encoding is enhanced through Frequent Pattern Mining, a non-trivial phase in Association Rule Mining(ARM) technique, in the field of Data Mining. The seminal Apriori algorithm has been modified in such a way that optimal number of patterns(sequence of characters) are obtained. These patterns are employed in the Encoding process of our algorithm, instead of single character based code assignment approach of Conventional Huffman Encoding. Our approach is built on an efficient hash based data structure, which minimizes the compression time by employing an efficient and novel technique for finding the frequency of the patterns. Simulation over benchmark corpus clearly shows the efficiency of our proposed work in relation to Conventional Huffman Encoding in terms of compression ratio and time.

Frequent Pattern Retrieval on Data Streams by using Sliding Window

“…However, frequent mining for uncertain data based on the extended support threshold needs more improvement. These methods are using some filtering constraints to find Frequent Patterns (FP) [35], [36], [37], [38]. Although this is a challenging task to find the use pattern, and different patterns carry different importance [23], [24], [25], [26].…”