Download the paper here

Read it in ResearchGate

An awarded VAST challenge research at the IEEEVis 2016 Conference - Baltimore, Maryland, USA (2017)

Visual Anomaly Detection in Spatio-Temporal Data using Element-Specific References.

by Daniel Alcaide, Jansi Thiyagarajan, Houda Lamqaddam, Jaume Nualart Vilaplana, and Jan Aerts (Visual Data Analysis Lab. KULeuven. Belgium)

ABSTRACT
The analysis and exploration of dynamic spatio-temporal data presents particular challenges. The VAST 2016 contest provided the opportunity to explore solutions in this space, focusing on the identification of patterns and anomalies. In this paper, we present an approach based on element-level references that allows for the exploration of individual movement data as well as sensor readings. This method earned the VAST 2016 Award for Robust Support for Visual Anomaly Detection.

Index Terms
Visual data analysis, anomaly detection, pattern exploration, interactive user interfac