Efficient Mining of FP-Growth Algorithm Structure And Apriori Algorithm using OFIM for Big Data

Mining big data is difficult.Working with massive datasets requires the use of computer software and an efficient algorithm to solve problems.Data mining specialists are familiar with two popular association rule algorithms: apriori and fp-growth.Businesses are able to make well-informed decisions based on customer trends and behavior thanks to these algorithms' assistance in finding patterns and correlations within large datasets.These data mining procedures are now even more accurate and efficient thanks to developments in machine learning techniques.However, there are some disadvantages to the association rule approach, including the requirement for a lot of memory, the necessity for extensive dataset searches to ascertain the item set's frequency, and sometimes less-than-ideal rules.The authors of this study examined the fp-growth, Apriori, and OFIM algorithms in order to analyze the rule results of the three methods.Significant performance differences were found in the results, with the fp-growth algorithm showing the best efficiency when working with big datasets.On the other hand, the Apriori approach had scalability issues and frequently resulted in longer processing times as the amount of the dataset expanded, despite being simpler to construct.Changes to the FP_Growth algorithm's operation are proposed in this study.The suggested method would reduce mining time and the quantity of regularly created items by using the suggested matrix OFIM to construct a highly compact FP-tree, which would significantly lessen the amount of decision-making required for huge datasets.Furthermore, by minimizing the number of items generated often, our method optimizes memory usage and greatly increases its speed while processing huge datasets.