Performance Analysis of T-PASCAL on Sparse and Dense Datasets

Usually popular temporal association rule mining methods are having performance bottleneck for database with different characteristics. Methods like temporal-apriori suffer from problem of candidate generation and database scans for temporal association rule mining. To overcome some of these problems of temporal-apriori TPASCAL has been discussed recently. The TPASCAL uses counting inference approach that minimizes as much as possible the number of pattern support counts performed when extracting frequent patterns. TPASCAL is calendric temporal association rule mining, which is working on precise-match association rules that require the association rule hold during every Interval. TPASCAL is based on the level wise extraction of frequent patterns. Here an attempt has been made to evaluate and compare the performance of TPASCAL with temporal-apriori on datasets with different characteristics. The relationship of execution time with characteristics like denseness, sparseness and volume of data extra has been obtained by implementing the algorithm on synthetic dataset available online. The parameter which is affecting the efficiency of two algorithm have been explored and evaluated.