Wikidata World Maps June 2015

From International Center for Computational Logic
Toggle side column

Wikidata World Maps June 2015

Wikidata contains a significant amount of geographic data: over 2 million items have a location on Earth. This can be used to generate maps that show the distribution of these items. This has first been done by Denny Vrandecic in the context of the EU Render project. However, the software used for this purpose is no longer functional, and the original maps are quite dated. Therefore, we have rebuild this tool using Wikidata Toolkit.

The resulting maps are shown below. Following Denny's original concept, we create a map for all data and further maps that are restricted to items that have a Wikipedia article in a particular language. The following images are scaled down: click on each map for the full resolution image.

These maps are based on the data dump of 22 June 2015. The languages picked here are ordered by the number of geolocated articles (shown in parentheses). The colours have the same meaning on all images. The scale at the bottom of each image gives an idea of the absolute magnitude: it is a logarithmic scale that shows a vertical line between numbers 1-9, 20-90, 100-900, and finally 1000-9000.

Finally, there is also a monochrome image of all Wikidata items in super-high resolution (17280x8640 pixels, about 2MB).

Discussion

It's important to understand what the colours mean, at least roughly. The scale is logarithmic. At this resolution, the first few colours mark single steps: blue denotes a single item, green denotes two items, ochre denotes three, red until rather bright orange are still representing numbers below 10, everything above 30 is essentially white. Nevertheless, all maps contain pixels with more than 1000 items. At this high resolution, however, the patterns of nearby bright and dark spots also give an impression of density in a certain area. If one would generate even much larger maps, a single item would be plain white, but the patterns would still give an impression of brightness for certain areas. On the other hand, if one picks a smaller resolution, the distinguishable colours extend into the 100s or even 1000s of items, but many more items come together in each point.

We can immediately make some basic observations:

  • Wikipedia (and Wikidata) has a general bias towards areas in Europe and Northern America.
  • All Wikipedias contain much fewer items than Wikidata, with even English having less than half of the overall entries (a complete list with numbers of geoitems for every Wikimedia project is at the bottom of this page).
  • Each Wikipedia has its distinct strengths in coverage, where it clearly surpasses other Wikipedias (even the English).
  • Items of interest follow real-world structures, e.g., the river Nile, the train lines in Japan, or the shore of Lake Erie (zoom in to see this more clearly)
  • Administrative regions may receive an above or below average coverage. One can see countries and states within countries having very different concentrations from nearby regions in many maps. This can create a rather patchy look in some maps (e.g., Russian Wikipedia).
  • Wikimedia Commons has relatively few items so far (possibly owing to the fact that links to Commons are a much more recent feature than links to Wikipedias). It seems to have a focus on points of (touristic) interest, rather than on administrative regions.


Some of the data shown will have been imported into Wikipedias from other projects, which can explain the patchiness of some item concentrations (e.g., importing Italian towns would be distinct from importing Austrian towns). Indeed, strong boundaries may suggest an artificial, non-organic growth of articles that does not so much reflect community interests but rather the priorities of individual special-interest groups and bot authors.

Some other random observations are:

  • Most Wikipedias with many geoitems have a good coverage of Europe, especially of France and Italy, while most Wikipedias other than English have very little coverage of the US and the UK. There are some exceptions:
    • Spanish Wikipedia has a decent coverage of North America.
    • Russian Wikipedia has many articles about places in Minnesota, but not about any other U.S. state.
    • Polish Wikipedia has a strong interest in England, but not in the rest of the UK.
  • English and Russian Wikipedia seem to divide Europe: there is a sharp line through Europe where interest seems to end for English Wikipedia and seems to start for Russian Wikipedia.
  • Japanese Wikipedia is mainly about Japan, and about Italy. Italian Wikipedia is mainly about Europe, and about Japan.

Possible Errors and Sources of Misinterpretation

The maps translate latitudes and longitudes of coordinates directly into coordinates in the image without compensating for the spherical shape of Earth. Therefore, pixels closer to the equator represent a bigger area than pixels closer to the poles. This means that the same level of brightness signifies a higher concentration of items when closer to the poles.

Common projections of maps use different ways of mapping coordinates to locations on a map. The most typical approach is Web Mercator, which is used in major mapping services online. It makes shapes closer to the poles appear stretched vertically compared to our naive projection. Images of maps of, say, Europe (which is rather far north) will thus give a slightly different impression of the shapes of some coastlines and borders.

The logarithmic colour scale may lead to false impressions. In addition, human colour perception is of course individually different and generally unreliable. The total amount of items can also be seen on ViziData, and the subjective impression of density may already be different there.

The absolute numbers of items do not reflect the quality of the underlying Wikipedia articles. Some Wikipedias have automatically created large amounts of very short articles from external data, while others have organically gathered detailed descriptions of many places. The maps do not distinguish these cases. It would be interesting to do, e.g., a "words per geographic location" analysis, but this would require further datasets to be combined with the Wikidata dumps.

Create Your Own Maps

The images have been created from the Wikidata JSON dumps of 2015-06-22 using Wikidata Toolkit (see the WorldMapProcessor example in the examples package). Developers can easily download this software and run it with different settings. It is possible to:

  • Create maps for arbitrary Wikimedia projects or for Wikidata as a whole
  • Change the size of the resulting image
  • Change the color scale (brightness)
  • Run the code on arbitrary past or future data dumps to get updated maps

License

The images have been created by Markus Krötzsch. The source data is published under Creative Commons CC-0, and Wikidata Toolkit is free software (License: Apache 2.0). Running the software does not constitute an act of creativity, and indeed the images can be recreated by anybody using the same software. The images are therefore released under Creative Commons CC-0 as well. However, if you find them useful, links to this page and to Wikidata Toolkit are still appreciated.

Links

Number of geoitems for all Wikimedia projects

Below is a table that gives the total number of all pages with associated geographic coordinates for each Wikimedia project. See List of Wikipedias for the meaning of the language codes.

Wikimedia project Number of geoitems
wikidata total 2165028
enwiki 950277
dewiki 356904
frwiki 347419
plwiki 328891
nlwiki 328162
ruwiki 298215
eswiki 261495
itwiki 249666
zhwiki 218602
ptwiki 185133
ukwiki 181372
svwiki 176639
viwiki 170164
srwiki 153675
cawiki 151013
shwiki 149511
mswiki 143254
rowiki 143136
fawiki 131844
jawiki 112344
vowiki 108071
warwiki 102505
commonswiki 89122
arwiki 87017
eowiki 79272
huwiki 79260
kkwiki 76763
nowiki 76344
skwiki 76147
cswiki 72436
hywiki 72134
uzwiki 71840
euwiki 66157
ocwiki 56541
idwiki 53995
lawiki 53382
nnwiki 50396
pmswiki 44136
bgwiki 44108
mgwiki 42112
urwiki 41601
kowiki 40467
cebwiki 39948
dawiki 39628
fiwiki 39565
ltwiki 33306
trwiki 32319
hrwiki 32275
slwiki 31945
simplewiki 30660
cewiki 30441
etwiki 26361
kawiki 24916
lmowiki 23246
htwiki 22239
enwikivoyage 22237
bewiki 20540
cywiki 20305
hewiki 19901
mkwiki 18074
sqwiki 17292
glwiki 16601
tlwiki 16391
elwiki 16053
scowiki 15391
pnbwiki 15280
bpywiki 15268
tgwiki 15004
brwiki 14916
bugwiki 13949
anwiki 13764
ttwiki 13282
newwiki 13281
lvwiki 12567
azwiki 12560
lbwiki 12332
be_x_oldwiki 12081
bswiki 11613
dewikivoyage 11544
gawiki 11460
bawiki 11356
thwiki 10203
scnwiki 10131
cvwiki 10077
iowiki 9452
iawiki 9279
napwiki 9031
afwiki 9011
hiwiki 8406
swwiki 8204
alswiki 8123
tawiki 7746
ndswiki 6849
fywiki 6633
bnwiki 6595
fawikivoyage 6569
jvwiki 6561
mrwiki 6240
quwiki 5467
pamwiki 5459
gdwiki 5051
newiki 5014
zh_yuewiki 4944
iswiki 4869
frwikivoyage 4820
roa_tarawiki 4756
mznwiki 4455
zh_min_nanwiki 4163
mlwiki 3906
ckbwiki 3847
barwiki 3748
bclwiki 3659
vecwiki 3606
astwiki 3408
itwikivoyage 3350
ilowiki 3329
liwiki 3267
nlwikivoyage 3124
kuwiki 3108
oswiki 3090
plwikivoyage 3076
sewiki 3039
tewiki 3010
pawiki 2988
ptwikivoyage 2959
mnwiki 2911
mrjwiki 2805
zeawiki 2732
nahwiki 2535
fowiki 2414
rmwiki 2413
minwiki 2314
kwwiki 2294
cbk_zamwiki 2235
hsbwiki 2229
xmfwiki 2220
hakwiki 2043
guwiki 2024
yiwiki 2006
ruwikivoyage 1998
iewiki 1995
nds_nlwiki 1933
bat_smgwiki 1923
pagwiki 1838
arzwiki 1794
scwiki 1788
hawwiki 1741
aywiki 1741
gvwiki 1738
yowiki 1737
vepwiki 1735
hewikivoyage 1735
csbwiki 1696
knwiki 1680
eswikivoyage 1668
sahwiki 1651
pflwiki 1628
crhwiki 1620
pcdwiki 1607
acewiki 1588
sawiki 1586
cowiki 1542
viwikivoyage 1513
stqwiki 1487
szlwiki 1471
ruewiki 1470
wawiki 1443
ladwiki 1432
mywiki 1401
svwikivoyage 1338
amwiki 1317
suwiki 1262
dsbwiki 1247
bhwiki 1228
diqwiki 1218
mhrwiki 1207
angwiki 1187
vlswiki 1181
orwiki 1178
ugwiki 1169
frpwiki 1147
wuuwiki 1140
lijwiki 1138
frrwiki 1131
kywiki 1084
lezwiki 1044
emlwiki 1038
zhwikivoyage 1034
fiu_vrowiki 1027
hifwiki 1014
bowiki 961
kabwiki 961
furwiki 937
klwiki 912
rwwiki 910
ganwiki 907
siwiki 904
gnwiki 900
extwiki 880
miwiki 874
cdowiki 868
tywiki 837
udmwiki 809
xalwiki 766
novwiki 765
pswiki 763
gagwiki 760
nawiki 758
sowiki 753
tkwiki 729
mwlwiki 724
dewikisource 724
tpiwiki 721
zh_classicalwiki 710
tetwiki 707
papwiki 671
kgwiki 650
rowikivoyage 636
pdcwiki 635
dvwiki 627
ruwikinews 620
snwiki 609
kvwiki 601
kaawiki 599
kmwiki 596
kshwiki 576
plwikiquote 570
piwiki 547
nrmwiki 540
mtwiki 518
bxrwiki 516
nvwiki 498
itwikiquote 495
elwikivoyage 494
twwiki 479
wowiki 477
koiwiki 455
lnwiki 448
map_bmswiki 433
kbdwiki 422
ukwikivoyage 422
arcwiki 401
avwiki 398
rmywiki 360
aswiki 356
myvwiki 340
tnwiki 331
biwiki 327
ltgwiki 308
lowiki 304
enwikiquote 302
pihwiki 299
krcwiki 298
ruwikiquote 296
lbewiki 296
eowikinews 293
omwiki 292
zuwiki 291
cuwiki 286
srnwiki 259
bjnwiki 252
roa_rupwiki 251
lrcwiki 249
sswiki 237
svwikinews 234
abwiki 231
enwikinews 221
jbowiki 220
chywiki 219
eewiki 215
mdfwiki 214
maiwiki 204
kiwiki 196
cawikiquote 196
rnwiki 195
igwiki 190
akwiki 188
smwiki 188
stwiki 185
hawiki 184
dewikinews 182
frwikinews 181
zawiki 173
sgwiki 172
mowiki 159
glkwiki 157
ikwiki 155
bmwiki 154
frwikiquote 152
srwikinews 139
tswiki 132
ffwiki 132
cswikiquote 131
dzwiki 129
nsowiki 126
pntwiki 125
itwikinews 123
dewikiquote 116
fjwiki 112
gotwiki 111
eswikiquote 109
chwiki 106
iuwiki 105
eswikinews 104
towiki 103
vewiki 96
ruwikisource 95
nywiki 93
xhwiki 90
plwikinews 89
kswiki 86
sdwiki 83
tyvwiki 82
chrwiki 81
fawikinews 76
ptwikinews 69
cawikinews 69
lgwiki 66
jawikinews 64
cswikinews 64
ukwikinews 61
tiwiki 61
jawikiquote 60
trwikinews 56
enwikisource 56
tumwiki 56
arwikinews 53
ukwikiquote 52
elwikinews 51
huwikinews 50
bswikinews 49
sqwikinews 45
fiwikinews 42
nowikinews 40
itwikisource 39
rowikinews 37
kowikinews 33
tawikinews 31
skwikiquote 31
zhwikinews 31
crwiki 27
ptwikiquote 26
hewikiquote 25
huwikiquote 25
ptwikibooks 25
nnwikiquote 24
fawikiquote 23
hewikinews 22
bswikiquote 17
frwikisource 17
iswikibooks 16
bgwikiquote 16
elwikiquote 16
dawikiquote 14
fiwikiquote 12
itwikibooks 11
ltwikiquote 11
eswikisource 11
azwikiquote 10
trwikiquote 10
plwikisource 9
eowikiquote 9
ngwiki 9
svwikiquote 8
rowikibooks 7
srwikiquote 7
slwikiquote 7
dewikibooks 7
tewikiquote 7
enwikibooks 6
bgwikinews 6
azwikibooks 6
nowikibooks 6
hrwikiquote 6
jawikisource 5
nlwikiquote 5
kowikisource 5
nowikiquote 4
zhwikiquote 4
arwikiquote 4
simplewikiquote 4
nlwikibooks 4
sqwikiquote 4
frwikibooks 4
hewikisource 4
zhwikisource 4
glwikiquote 3
jawikibooks 3
kowikiquote 2
trwikisource 2
guwikiquote 2
etwikiquote 2
nlwikinews 2
slwikisource 2
iiwiki 2
hywikiquote 2
trwikibooks 2
idwikibooks 2
zhwikibooks 2
fiwikibooks 2
iswikiquote 1
svwikibooks 1
cswikisource 1
mlwikisource 1
fawikibooks 1
elwikibooks 1
kuwikiquote 1
mlwikiquote 1
huwikibooks 1
hrwikisource 1
thwikisource 1
ruwikibooks 1
euwikiquote 1
ukwikisource 1
srwikibooks 1
kawikibooks 1
afwikiquote 1
cywikiquote 1
lawikiquote 1
mhwiki 1
plwikibooks 1
tlwikibooks 1
mkwikibooks 1
ttwikibooks 1
hrwikibooks 1
vecwikisource 1
hewikibooks 1
lawikisource 1
skwikisource 1
thwikiquote 1
srwikisource 1
glwikisource 1
cswikibooks 1
urwikiquote 1
sqwikibooks 1
eowikibooks 1
bgwikisource 1
thwikibooks 1
guwikisource 1
tawikiquote 1
eswikibooks 1
tawikisource 1