PLOMaR: An ontology framework for context modeling and reasoning on crowd-sensing platform

Abstract

Crowd-sensing is a popular way to sense and collect data using smartphones that reveals user behaviors and their correlations with device performance. PhoneLab is one of the largest crowd-sensing platform based on the Android system. Through experimental instrumentations and system modifications, researchers can tap into a sea of insightful information that can be further processed to reveal valuable context information about the device, user and the environment. However, the PhoneLab data is in JSON format. The process of inferring reasons from data in this format is not straightforward. In this paper, we introduce PLOMaR -an ontology framework that uses SPARQL rules to help researchers access information and derive new information without complex data processing. The goals are to (i) make the measurement data more accessible, (ii) increase interoperability and reusability of data gathered from different sources, (iii) develop extensible data representation to support future development of the PhoneLab platform. We describe the models, the JSON to RDF mapping processes, and the SPARQL rules used for deriving new information. We evaluate our framework with three application examples based on the sample dataset provided.