PERFORMANCE EVALUATION AND ANALYSIS OF SQL BASED APPROACHES FOR ASSOCIATION RULE MINING
Data mining aims at discovering important and previously unknown patterns from the datasets. Database mining performs mining directly on data stored in Data Base Management Systems. Several SQL based approaches for mining have been studied in the literature.
The main focus in this thesis is on the performance evaluation of these approaches. We study several additional optimizations for the K-way join approach and SQL-OR based approaches and evaluate them using IBM DB2/UDB and Oracle RDBMSs. We experimentally evaluate these approaches and their optimizations and compare their performance on large data sets. We also present analytical evaluation for the K-way join approach and its optimizations. Finally, we summarize the results and indicate the conditions for which the individual optimizations are useful.
The larger goal of this work is to feed these results into a mining optimizer that chooses the specific strategy for mining the input dataset based on its characteristics.