Association Rule Mining on Spambase Dataset using Tanagra

• Published in: International journal of innovative technology and exploring engineering 2020-04, Vol.9 (6), p.890-894
• Main Authors: Devi, Pushpa, Bhardwaj, Vikrant Singh, Bansal, Dr. Kishori Lal
• Format: Article
• Language: English
• ISSN: 2278-3075; 2278-3075
• DOI: 10.35940/ijitee.C8022.049620

There is huge amount of data being generated every minute on internet. This data is of no use until we cannot extract useful information from it. Data mining is the process of extracting useful information or knowledge from this huge amount of data that can be further used for various purposes. Discovering Association rules is one of the most important tasks among all other data mining tasks. Association rules contain the rules in the form of IF then THAN form. The leftmost part of the rule i.e. IF is called as the Antecedent which defines the condition and the rightmost part i.e. ELSE is called as the Consequent which defines the result. In this paper, we present the overview and comparison of Apriori, Apriori PT and Frequent Itemsets algorithm of association component in Tanagra Tool. We analyzed the performance based on the execution time and memory used for different number of instances, support and Rule Length in Spambase Dataset. The results show that when we increase the support value the Apriori PT takes the less execution time and Apriori takes less memory space. When numbers of instances are reduced Frequent Itemsets outperforms well both in case of memory and execution time. When rule length is increased the Apriori algorithm performs better than Apriori PT and Frequent Itemsets.