Title: Oriented spatial box plot, a new pattern for points clusters
Authors: Laurent Etienne; Thomas Devogele; Gavin McArdle
Addresses: Laboratory of Computer Science, Tours University, Tours, France ' Laboratory of Computer Science, Tours University, Tours, France ' National Centre for Geocomputation, Maynooth University, National University of Ireland Maynooth, Maynooth, Co. Kildare, Ireland
Abstract: Nowadays, an abundance of sensors are used to collect very large datasets containing spatial points which can be mined and analysed to extract meaningful patterns and information. This article examines patterns which describe the dispersion of 2D data around a central tendency. Several state of the art patterns for point cluster analysis are presented and critiqued before a new pattern, the oriented spatial box plot, is defined. The oriented spatial box plot extends the classical one-dimensional box plot for summarising and visualising 2D point clusters. The pattern is suitable for detecting outliers and understanding the spatial density of point clusters.
Keywords: oriented spatio-temporal box plot; bagplot; quelplot; outlier detection; point clusters; spatio-temporal patterns; cluster analysis; outliers; spatial density.
DOI: 10.1504/IJBIDM.2014.068367
International Journal of Business Intelligence and Data Mining, 2014 Vol.9 No.3, pp.233 - 253
Published online: 10 Apr 2015 *
Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article