An efficient and optimised frequent pattern mining using novel multipath-graph structure
by Suhasini A. Itkar; Uday V. Kulkarni
International Journal of Data Mining, Modelling and Management (IJDMMM), Vol. 9, No. 1, 2017

Abstract: The frequent pattern mining is one of the most focused areas in the data mining domain. The frequent pattern growth (FP-growth) algorithm introduced compact prefix-based data structure, frequent pattern tree (FP-Tree), to store frequent itemsets in compressed format. The FP-growth attempts to overcome drawback of candidate generation approach of multiple database scan. This work aims to propose a novel optimised data structure multipath-Graph (MP-Graph) for improving memory utilisation and efficiency of mining algorithms. The MP-Graph is a compact graph structure to store frequent patterns in memory. It generates graph nodes equal to number of frequent 1-itemsets of transaction database. Further, it stores multiple occurrences of prefix subpaths along the edges of the graph in the form of transaction bitmap instead of storing frequency of individual item node. The proposed structure helps to mine frequent patterns without constructing conditional FP-Trees. The performance of the MP-Graph mining algorithm is compared with FP-growth, CT-PRO and IFP-growth. The experimental results show order of magnitude improvement in memory consumption to store frequent patterns, nodes generated and time complexity.

Online publication date: Wed, 01-Mar-2017

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Data Mining, Modelling and Management (IJDMMM):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com