MEC - Monitoring Clusters' Transitions

Abstract

In this paper we address the problem of monitoring the evolution of clusters, which has become an important research issue in recent years. We adopt two main strategies for cluster characterization - representation by enumeration and representation by comprehension -, and propose the MEC (Monitoring the Evolution of Clusters) framework, which was developed along the lines of the change mining paradigm. MEC includes a taxonomy of various types of cluster transitions, a tracking mechanism that depends on cluster representation, and a transition detection algorithm. Our tracking mechanism can be subdivided in two methods, devised to monitor clusters' transitions: one based on graph transitions, and another based on clusters' overlap. These are the most relevant contributions of this work. To demonstrate the feasibility and applicability of MEC we present illustrative examples, using datasets from different knowledge areas, such as Economy and Education. Our results are encouraging and demonstrate the ability of MEC framework to provide an efficient diagnosis of clusters' transitions.