Affiliations: [a] Data Analysis and Modeling, Hasselt University, Diepenbeek, Belgium | [b] School of Information Technology and Engineering (SITE), University of Ottawa, Ottawa, ON, Canada | [x] LIAAD, University of Porto, Porto, Portugal | [y] Fraunhofer IAIS, Sankt Augustin, Germany
Correspondence:
[*]
Corresponding author: Benoît Depaire, Data Analysis and Modeling, Hasselt University, 3590 Diepenbeek, Belgium. E-mail: benoit.depaire@uhasselt.be.
Note: [1] The authors wish it to be known that, in their opinion, both first two authors should be regarded as joint First Authors.
Abstract: The goal of this article is to introduce a collaborative clustering approach to the domain of ubiquitous knowledge discovery. This clustering approach is suitable in peer-to-peer networks where different data sites want to cluster their local data as if they consolidated their data sets, but which is prevented by privacy restrictions. Two variants exist, i.e. one for data sites with the same observations but different features and one for data sites with the same features but different observations. The technique contains two parts, i.e. a collaborative fuzzy clustering technique and a particle swarm optimization to optimize the collaboration between data sites. Empirical analysis show how and when this PSO-CFC approach outperforms local fuzzy clustering.
