A novel way to compute association rules

Association Rule mining is the prime booming field among researchers. Apriori algorithm is a prime algorithm to compute association rules. Apriori algorithm considers only frequent itemsets and it neglects the non-frequent itemsets. In real-time scenarios, Non-frequent itemsets also have the chance to give more utility. Utility mining is a newish form of data mining study topic that focuses solely on high utility itemsets computed from utility values. To overcome this problem, we proposed an approach that incorporates both frequent and utility values called the Novel Utility Frequent Apriori algorithm. This approach considered both frequent itemsets together with non-frequent itemsets. Utility computed for both frequent itemsets and rare itemsets. Finally, it categorized the itemsets based on utility value and frequent value like High-Profit High Frequency, High-Profit Rare Frequency, Low-Profit High Frequency, and Low-Profit Rare Frequency itemsets. Repeated transactions were handled efficiently by our proposed method. We experimented with different datasets by using python, The Novel Utility Frequent Apriori method surpasses the classic Apriori algorithm in terms of time i.e. average rate of time reduction was 63% with first experiment and 82% with second experiment. We found that our approach is effective in categories of itemsets and also this approach will be useful in E-Commerce to make more profit, Medical field to discover new diseases and Banking sector to discover fraud activities.