Mobility behaviour based on mobile cellular data: a fuzzy-clustering approach to identify mobility behavioural segments and geographical patterns

Published: 20-04-2020| Version 1 | DOI: 10.17632/86h7vd7grh.1
Contributors:
Tünde Szabó,
Ákos Jakobi,
Márton Prorok,
József Tabányi ,
Richárd Tomka

Description

Multidimensional fuzzy clustering based on individual mobile cellular data provides information for mobility pattern recognition, to define typical groups of mobility behaviour. Grouping of individual mobility patterns is a multivariable, empirical way for behavioural profiling of individuals. Geographical component of mobile cellular data is registered, collected and processed all over the world, using fairly similar technologies for data retrieval, which makes the analysis replicable. Academics, business professionals, civil planners aiming at being informed about general and specific mobility behaviour of a geographical unit can benefit from multidimensional fuzzy clustering method. The data and the method we share here provides a data procession framework, computational sequence, along with some domain-specific parameter setting for multidimensional fuzzy clustering analysis. Fuzzy clustering can be the core of further segmentation analysis, aimed at identifying general and specific behavioural profiles regarding individual mobility. The data we published here is an aggregated, analysed dataset derived from long-term system operation data, that meets the criteria of big data in volume, variety, velocity and exhaustiveness. It provides aggregated data about individual mobility trajectories of about 5.6 million equipment, about 45% of the Hungarian telecommunication devices, which makes the dataset unique in terms of spatial and temporal resolution as well as coverage.

Files