Online detection of patterns in semantic trajectory data streams

Abstract

In this paper a system for online detection of patterns in symbolic trajectory data streams is presented. Indoor positioning and analysis of obtained location data using Data Stream Management System are combined into a very powerful system. It contains two separate components: a mobile application and a central server. These components are loosely coupled with an idea to demonstrate knowledge that could be extracted from mobile users from their everyday activities. The system has increased performance in comparison to other similar systems that use raw data from mobile application (GPS feeds) and analysis is done on central server. System process raw data within mobile application and enrich them with semantic meaning, and central server process that semantic data using the Data Stream Management System NEsper detecting patterns in semantic trajectory data streams.