Inter-space Machine Learning in Smart Environments

Abstract

Today, our built environment is not only producing large amounts of data, but -driven by the Internet of Things (IoT) paradigmit is also starting to talk back and communicate with its inhabitants and the surrounding systems and processes. In order to unleash the power of IoT enabled environments, they need to be trained and configured for space-specific properties and semantics. This paper investigates the potential of communication and transfer learning between smart environments for a seamless and automatic transfer of personalized services and machine learning models. To this end, we explore different knowledge types in context of smart built environments and propose a collaborative framework based on Knowledge Graph principles and IoT paradigm for supporting transfer learning between spaces.

A Transfer Learning Framework for IoT-enabled Environments

“…IoT-based systems are typically built following a three-layer model that consists of: (i) a sensing layer, which acquires the observation of interest from the environment; (ii) a cognitive layer, which is concerned with context acquisition, modeling, and reasoning; and (iii) an actuate layer that triggers an action or invokes a service according to some predefined logic [1]. The main effort required for automatic change management in a private IoT space is hidden in the cognitive layer.…”