AbstractsComputer Science

Association rule mining visualization: learning association rule mining made easy

by Bindu Madhavi Khambam




Institution: California State University – Sacramento
Department: Computer Science
Degree: MS
Year: 2015
Keywords: Association rule mining; Apriori algorithm; FP growth algorithm; Data mining
Record ID: 2060502
Full text PDF: http://hdl.handle.net/10211.3/138872


Abstract

Association rule mining algorithms are the fundamentals for a data mining course. Association rule mining helps to extract useful information from the data for various applications such as market analysis. Association rules are used for finding frequent items set, associations, correlations, or causal structures among sets of items or object. Generally, the students find it difficult to understand these key concepts because it requires abstract thinking. In addition, conveying a clear explanation of how these processes work is a bit of a challenge for the instructors too. Since the best way to understand complex algorithms is to see them in action, it would be very helpful if a visualization tool of these algorithms were available to the students to play with. Hence, the drive to come up with a data mining visualization tool that can animate a few of the most widely used and complex data mining algorithms. The objective of the project is to provide an association rule-mining tutorial and make the students understand the basic underlying concepts. Additionally the tutorial provides visualization tool of these algorithms, which will help the students to understand the complex algorithms better and they will be able to test them. This project is intended to create an exploration environment, in which students can learn through experimentation. It is targeted at the students wanting to practice algorithms that are being covered in class, as well as instructors wishing to embellish their lectures with an animated interface to help the students. An understanding of the underlying mechanics of algorithms is of great importance to students who are taking data mining courses.