Multi-layered data mining architecture in the context of Internet of Things

Abstract

In the context of Internet of Things (IoT), the amount of data collected from physical objects is ever increasing. However, in the whole infrastructure there are several bottlenecks related especially to hardware limitations (low end processors on devices and communication means), but also time, therefore analyzing this data requires knowledge about the available computational resources. In this article we propose a layered architecture for data mining, which relies on simple integer calculations at the bottom and goes up to complex (distributed) algorithms at the top layer. We present the characteristics of each layer and the advantages of using such an architecture.