Fig 1.
UI showing regional variants for ‘non-professional soccer playing’ (left) and ‘(second) breakfast that people have at their workplace’ (right).
Table 1.
Variables chosen for the prediction of the users’ regional background.
Fig 2.
Density distribution of users by age and gender.
Fig 3.
Distribution of localities in the historic WDU survey (A) and contemporary data (B).
Counts of localities in the contemporary data are aggregated for each WDU locality.
Fig 4.
Sketch for computing the degree of change for each WDU locality.
Fig 5.
Variant and degree of change maps for ‘non-professional soccer playing’.
The left-hand panel shows contemporary data as hexagons with historical data superimposed as dots. The right-hand panel indicates change (black: substantial change, bright grey: very little change).
Fig 6.
Variant and degree of change maps for ‘slippers’.
Fig 7.
Variant and degree of change maps for ‘breadman’.
Fig 8.
Variant and degree of change maps for ‘(second) breakfast that people have at their workplace’.
Fig 9.
Variant and degree of change maps for ‘pencil case’.
Fig 10.
Variant and degree of change maps for ‘pancake’.
Fig 11.
Variant and degree of change maps for ‘10:15’.
Fig 12.
Variant and degree of change maps for ‘hiccups’.
Fig 13.
Variant and degree of change maps for ‘beef patty’.
Fig 14.
Median degree of change aggregated across 14 variables.
Fig 15.
Degree of change maps for ‘10:15’.
Users <25 years old are shown on the left, users >60 are shown on the right.
Fig 16.
Aggregated change for users <25 years old (left) and users >60 years old (right).
Fig 17.
Regional distribution of variants for ‘(second) breakfast that people have at their workplace’.
Right is our contemporary data (2015) and left is AdA (2003ff).