A new methodology for the regionalisation and aggregation of in-app location data.

Keywords: regionalisation, mobility, in-app data, aggregation, disclosure control
Abstract Type: Paper Abstract

Abstract

In most cases, sensitive location data collected by mobile phones is aggregated to pre-existing geographic units in order to minimise the disclosure of personal information. The use of grids as well as units created by official sources such as the ‘Output Areas’ developed for the British Census, is common in aggregating processes. We propose a new method for the aggregation of mobile phone generated location datasets that creates bespoke geometries. These geometries maximise the granularity of the data, whilst minimising the risks of disclosing personal information. The resulting small areas are built on Uber’s H3 hexagonal indexing system by attributing activity counts and land-use features to each cell, then merging cells into geographies containing a predetermined number of data points and respecting the underlying topography and land use. This methodology has applications to widely available data and enables bespoke geographical units to be created for different contexts. For example, the volume and spatial pattern of location data varies across the day with workplaces giving way to entertainment districts in the evening, it therefore follows that the geographical units should change too. We compare the generated units to established aggregates from the British Census amongst others. We demonstrate that our outputs are more representative of the original mobile phone dataset and minimise data omission caused by low counts. This speaks to the need for a data-driven and context-driven regionalisation methodology.