Stationary and time-varying patterns in activity diary panel data - Explorative analysis with association rules

Abstract

The question of identifying temporal patterns in activity diary data has received only scant attention in the transportation literature, but interest is rapidly increasing. Most of the existing research uses well-known econometric methods to quantify change. Use of association rules to explore activity diary panel data, involving two waves, for possible stationary and time-varying patterns in activity-travel patterns is reported. The data for this analysis stem from the municipality of Voorhout in the Netherlands. Data were collected in 1997 and 1998 before and after opening of a new railway station. Results of the analysis indicate that specific household and individual attributes have a larger effect on daily activity patterns than others and that the effect of these attributes has significantly changed. Because changes in other sociodemographic attributes are almost nonexistent and activity patterns for communities are known to be stable, this study claims that the observed shifts in dependencies come from this new station.