Author
Ölmezoğulları, Erdi, Arı, İsmail, Çelebi, Ö. F., Ergüt, S.
Publication Date
2013
Publication Place
-
IEEE
Subject
Apriori, Data stream mining, FP-Growth, Association rule mining, Complex event processing
Type
Document
Language
English
Digital
Yes
Manuscript
No
Library
Özyeğin University
Library Asset ID
978-1-4673-5561-2
Record ID
3cbbd2c3-c5d7-4f20-99a1-e19457519380
Library Location
Computer Science
Date
2013
Notes
Due to copyright restrictions, the access to the full text of this article is only available via subscription.
Sample Text
Today, the informatics world is trying to cope with "big data" problems (mass, speed, diversity, inconsistency of data) on the way to reach useful information. In this article, the details and performance findings of the "online" implementation of Relational Rule Mining (RCM) on big data streams, which has not been done before in the literature, will be shared. For stream mining, Apriori and FP-Growth algorithms have been added to the event stream engine called Esper. These two algorithms were compared on the resulting system using sliding windows and LastFM social music site data. By choosing FPGrowth, which has high performance, a real-time and rule-based recommendation engine was created. Our most important findings are that online rule extraction has shown that: (1) many more rules can be calculated (2) much faster and more efficiently and (3) much earlier than offline rule extraction. Additionally, many interesting and realistic rules have been found to suit musical tastes, such as “George Harrison⇒The Beatles”. We hope that our results will shed light on the design and implementation of other big data analytics systems in the future.
DOI
10.1109/SIU.2013.6531483
