Home // CLOUD COMPUTING 2018, The Ninth International Conference on Cloud Computing, GRIDs, and Virtualization // View article

Capturing Data Topology Using Graph-based Association Mining

Authors:
Khalid Kahloot
Peter Ekler

Keywords: Graph-based data representation; topology capturing; Apriori rule mining; Association Analysis

Abstract:
A dataset can underline a statistical plausibility and implausible characteristics. A graph can model the inter-relationship between the set variables in a dataset. On the other hand, the association mining produces causal structures for a transactional dataset in various kinds. Therefore, a better data representation can be attained by merging both of the two powerful tools together. Knowledge within a dataset is captured as a topology by combining an algorithm of association rule mining with a complex graph theory. In this paper, we present a modified graph-based version of Apriori algorithm for association mining, in which the probabilities of frequencies are represented using a graph data structure. A computational approach is reflected in the graph and all rules are composed of nodes, which are interconnected by in-degree and off-degree edges. The algorithm is using Apriori statistical rule mining to compose weighted nodes and weighted directed edges graph. The computational approach is necessary to be able to unravel complex relationships between co-occurred values due to multi-hop graph connectivity and navigability. The modified algorithm is tested based on heterogeneously composed traffic datasets.

Pages: 127 to 132

Copyright: Copyright (c) IARIA, 2018

Publication date: February 18, 2018

Published in: conference

ISSN: 2308-4294

ISBN: 978-1-61208-607-1

Location: Barcelona, Spain

Dates: from February 18, 2018 to February 22, 2018