An empirical analysis and comparison of apriori and FP- growth algorithm for frequent pattern mining

Abstract

In this paper, we determine the empirical comparison of Apriori and FP-growth algorithm for frequent item set sequences for Web Usage data. We define the data structure, its implementation and algorithmic features mainly focusing on those that also arise in frequent item set mining. Web usage mining itself can be defined further depending on the type of usage data is considered like web server data, application server data and application level data. User logs that are collected at web server are also known as web server data. Some of the characteristic data collected at a web server include IP addresses of users, page references, and access time of the users and these are the main input to the present research. The comparison of algorithm concentrates on web usage mining and particularly focuses on determining the web usage patterns of websites from the server log files. In our analysis we take into empirical comparison for properties like memory size, input data, pre-fetching, scalability and processing efficiency etc, in order to better understand the results of the evaluation.