Wikidata:WordGraph

On November 7, 2024, Google released the WordGraph dataset as a belated present for Wikidata’s 12th birthday. According to its self-description, “[t]he WordGraph dataset contains multilingual lexicon entries linked to wikipedia entities, focusing on human-denoting nouns and demonym adjectives. Each lexicon entries contain inflected word-form and morphological information [of] all locales.” WordGraph was made available on GitHub.

The dataset is published under Creative Commons CC0 License (Q6938433), which makes it compatible for integration with Wikidata. We want to thank Bruno Cartoni (Q111290954), Saran Lertpradit, Seungmin Back, Daniel Calvelo Aros (Q100784694), Kuang-Yu Samuel Chang (Q86087144) and Abdelrahman Nabil at Google for this beautiful gift.

Here we look at some statistics about WordGraph, particularly compared to Wikidata.

All together, the WordGraph release contains almost a million word forms (968,153) in 39 languages: #Arabic, #Bulgarian, #Bangla, #Catalan, #Czech, #Danish, #German, #Greek, #English, #Spanish, #Estonian, #Persian, #Finnish, #French, #Hebrew, #Hindi, #Croatian, #Hungarian, #Indonesian, #Italian, #Kannada, #Lithuanian, #Latvian, #Marathi, #Dutch, #Norwegian, #Polish, #Portuguese, #Romanian, #Russian, #Slovak, #Slovenian, #Serbian, #Swedish, #Tamil, #Telugu, #Turkish, #Ukrainian, and #Urdu.

Norwegian, Hindi, and Urdu are not properly mapped, due to their different representation between WordGraph and Wikidata. I am asking for help to fix this mapping.

Explanations of the following statistics: Forms are the different forms a lexeme can have, counted for each combination of grammatical features. Unique forms only count the forms that actually differ. Unique lemmas count the different headwords of lexemes. Unique topics are the different QIDs mentioned in the claims respectively the mappings of the entry. Senses are the mappings of lemmas to QIDs. In Wikidata, this might overcount, as it counts all QIDs on claims of the senses, not just the sense-related ones. Shared forms and lemmas is counting how many forms and lemmas are in both WordGraph and Wikidata. Shared senses count only matches of lemmas to QIDs for their meaning. Missing senses are QIDs given on WordGraph that are missing on a shared lemma in Wikidata.

Arabic

32526 forms in WordGraph
12379 unique forms in WordGraph
919 unique lemmas in WordGraph
944 unique topics in WordGraph
957 senses in WordGraph
2231 lexemes in Wikidata
2221 unique lemmas in Wikidata
600 unique topics in Wikidata
656 senses in Wikidata (possibe overcount)
1842 forms in Wikidata
1779 unique forms in Wikidata
5 shared forms
2 shared lemmas
0 shared senses
2 missing senses on existing Lexemes

Bulgarian

14232 forms in WordGraph
10471 unique forms in WordGraph
1291 unique lemmas in WordGraph
1340 unique topics in WordGraph
1345 senses in WordGraph
173 lexemes in Wikidata
173 unique lemmas in Wikidata
67 unique topics in Wikidata
67 senses in Wikidata (possibe overcount)
245 forms in Wikidata
233 unique forms in Wikidata
7 shared forms
2 shared lemmas
2 shared senses
1 missing senses on existing Lexemes

Bangla

27282 forms in WordGraph
8058 unique forms in WordGraph
1398 unique lemmas in WordGraph
1347 unique topics in WordGraph
1464 senses in WordGraph
10780 lexemes in Wikidata
10243 unique lemmas in Wikidata
1823 unique topics in Wikidata
3818 senses in Wikidata (possibe overcount)
69902 forms in Wikidata
57227 unique forms in Wikidata
136 shared forms
34 shared lemmas
22 shared senses
19 missing senses on existing Lexemes

Catalan

3797 forms in WordGraph
2935 unique forms in WordGraph
924 unique lemmas in WordGraph
957 unique topics in WordGraph
957 senses in WordGraph
175 lexemes in Wikidata
173 unique lemmas in Wikidata
104 unique topics in Wikidata
107 senses in Wikidata (possibe overcount)
208 forms in Wikidata
200 unique forms in Wikidata
4 shared forms
2 shared lemmas
1 shared senses
1 missing senses on existing Lexemes

Czech

51247 forms in WordGraph
22823 unique forms in WordGraph
1430 unique lemmas in WordGraph
1466 unique topics in WordGraph
1479 senses in WordGraph
29644 lexemes in Wikidata
29260 unique lemmas in Wikidata
20902 unique topics in Wikidata
22451 senses in Wikidata (possibe overcount)
861902 forms in Wikidata
211267 unique forms in Wikidata
2333 shared forms
83 shared lemmas
67 shared senses
21 missing senses on existing Lexemes

Danish

13736 forms in WordGraph
12194 unique forms in WordGraph
1584 unique lemmas in WordGraph
1633 unique topics in WordGraph
1665 senses in WordGraph
91505 lexemes in Wikidata
90417 unique lemmas in Wikidata
7820 unique topics in Wikidata
9037 senses in Wikidata (possibe overcount)
611986 forms in Wikidata
564466 unique forms in Wikidata
6646 shared forms
292 shared lemmas
261 shared senses
67 missing senses on existing Lexemes

German

79418 forms in WordGraph
13787 unique forms in WordGraph
2969 unique lemmas in WordGraph
2838 unique topics in WordGraph
3148 senses in WordGraph
214090 lexemes in Wikidata
213399 unique lemmas in Wikidata
12739 unique topics in Wikidata
16586 senses in Wikidata (possibe overcount)
567158 forms in Wikidata
206858 unique forms in Wikidata
3747 shared forms
556 shared lemmas
474 shared senses
145 missing senses on existing Lexemes

Greek

10736 forms in WordGraph
4369 unique forms in WordGraph
635 unique lemmas in WordGraph
681 unique topics in WordGraph
681 senses in WordGraph
43326 lexemes in Wikidata
43263 unique lemmas in Wikidata
133 unique topics in Wikidata
144 senses in Wikidata (possibe overcount)
70987 forms in Wikidata
38905 unique forms in Wikidata
701 shared forms
4 shared lemmas
4 shared senses
0 missing senses on existing Lexemes

English

6904 forms in WordGraph
5923 unique forms in WordGraph
3567 unique lemmas in WordGraph
3355 unique topics in WordGraph
3784 senses in WordGraph
75194 lexemes in Wikidata
67495 unique lemmas in Wikidata
25335 unique topics in Wikidata
28160 senses in Wikidata (possibe overcount)
137911 forms in Wikidata
115078 unique forms in Wikidata
1885 shared forms
960 shared lemmas
811 shared senses
297 missing senses on existing Lexemes

Spanish

14620 forms in WordGraph
8882 unique forms in WordGraph
2832 unique lemmas in WordGraph
3058 unique topics in WordGraph
3270 senses in WordGraph
60088 lexemes in Wikidata
56089 unique lemmas in Wikidata
5179 unique topics in Wikidata
7617 senses in Wikidata (possibe overcount)
679598 forms in Wikidata
555915 unique forms in Wikidata
3491 shared forms
710 shared lemmas
897 shared senses
127 missing senses on existing Lexemes

Estonian

2744 forms in WordGraph
2351 unique forms in WordGraph
87 unique lemmas in WordGraph
97 unique topics in WordGraph
97 senses in WordGraph
83199 lexemes in Wikidata
80642 unique lemmas in Wikidata
1078 unique topics in Wikidata
1136 senses in Wikidata (possibe overcount)
2801896 forms in Wikidata
2637583 unique forms in Wikidata
512 shared forms
0 shared lemmas
0 shared senses
0 missing senses on existing Lexemes

Persian

2142 forms in WordGraph
1677 unique forms in WordGraph
81 unique lemmas in WordGraph
80 unique topics in WordGraph
81 senses in WordGraph
18293 lexemes in Wikidata
17594 unique lemmas in Wikidata
1715 unique topics in Wikidata
34642 senses in Wikidata (possibe overcount)
40389 forms in Wikidata
37092 unique forms in Wikidata
3 shared forms
1 shared lemmas
0 shared senses
1 missing senses on existing Lexemes

Finnish

7020 forms in WordGraph
6073 unique forms in WordGraph
225 unique lemmas in WordGraph
232 unique topics in WordGraph
232 senses in WordGraph
1054 lexemes in Wikidata
1053 unique lemmas in Wikidata
459 unique topics in Wikidata
558 senses in Wikidata (possibe overcount)
9337 forms in Wikidata
9205 unique forms in Wikidata
22 shared forms
2 shared lemmas
1 shared senses
1 missing senses on existing Lexemes

French

14372 forms in WordGraph
8905 unique forms in WordGraph
2906 unique lemmas in WordGraph
2853 unique topics in WordGraph
3119 senses in WordGraph
19835 lexemes in Wikidata
19505 unique lemmas in Wikidata
8414 unique topics in Wikidata
10401 senses in Wikidata (possibe overcount)
331819 forms in Wikidata
253717 unique forms in Wikidata
1627 shared forms
569 shared lemmas
419 shared senses
252 missing senses on existing Lexemes

Hebrew

14353 forms in WordGraph
9107 unique forms in WordGraph
1245 unique lemmas in WordGraph
1257 unique topics in WordGraph
1301 senses in WordGraph
29640 lexemes in Wikidata
27893 unique lemmas in Wikidata
5838 unique topics in Wikidata
6024 senses in Wikidata (possibe overcount)
441485 forms in Wikidata
322656 unique forms in Wikidata
2056 shared forms
216 shared lemmas
182 shared senses
59 missing senses on existing Lexemes

Hindi

Note: Hindi was not properly mapped. Besides the first five stats, all others are not trustworthy.

15526 forms in WordGraph
3199 unique forms in WordGraph
1868 unique lemmas in WordGraph
1892 unique topics in WordGraph
1974 senses in WordGraph
1 lexemes in Wikidata
1 unique lemmas in Wikidata
0 unique topics in Wikidata
0 senses in Wikidata (possibe overcount)
0 forms in Wikidata
0 unique forms in Wikidata
0 shared forms
0 shared lemmas
0 shared senses
0 missing senses on existing Lexemes

Croatian

52171 forms in WordGraph
17990 unique forms in WordGraph
1413 unique lemmas in WordGraph
1451 unique topics in WordGraph
1482 senses in WordGraph
773 lexemes in Wikidata
747 unique lemmas in Wikidata
828 unique topics in Wikidata
1159 senses in Wikidata (possibe overcount)
10180 forms in Wikidata
5025 unique forms in Wikidata
345 shared forms
28 shared lemmas
28 shared senses
3 missing senses on existing Lexemes

Hungarian

10574 forms in WordGraph
10160 unique forms in WordGraph
299 unique lemmas in WordGraph
310 unique topics in WordGraph
310 senses in WordGraph
121 lexemes in Wikidata
118 unique lemmas in Wikidata
62 unique topics in Wikidata
64 senses in Wikidata (possibe overcount)
159 forms in Wikidata
154 unique forms in Wikidata
0 shared forms
0 shared lemmas
0 shared senses
0 missing senses on existing Lexemes

Indonesian

13451 forms in WordGraph
5132 unique forms in WordGraph
1647 unique lemmas in WordGraph
1725 unique topics in WordGraph
1751 senses in WordGraph
20212 lexemes in Wikidata
20191 unique lemmas in Wikidata
385 unique topics in Wikidata
402 senses in Wikidata (possibe overcount)
412640 forms in Wikidata
393154 unique forms in Wikidata
365 shared forms
15 shared lemmas
13 shared senses
5 missing senses on existing Lexemes

Italian

9932 forms in WordGraph
6786 unique forms in WordGraph
2287 unique lemmas in WordGraph
2301 unique topics in WordGraph
2453 senses in WordGraph
64254 lexemes in Wikidata
62040 unique lemmas in Wikidata
18765 unique topics in Wikidata
20732 senses in Wikidata (possibe overcount)
521626 forms in Wikidata
412361 unique forms in Wikidata
2129 shared forms
546 shared lemmas
453 shared senses
177 missing senses on existing Lexemes

Kannada

299 forms in WordGraph
275 unique forms in WordGraph
63 unique lemmas in WordGraph
63 unique topics in WordGraph
63 senses in WordGraph
14 lexemes in Wikidata
14 unique lemmas in Wikidata
14 unique topics in Wikidata
14 senses in Wikidata (possibe overcount)
9 forms in Wikidata
9 unique forms in Wikidata
0 shared forms
0 shared lemmas
0 shared senses
0 missing senses on existing Lexemes

Lithuanian

17424 forms in WordGraph
11798 unique forms in WordGraph
584 unique lemmas in WordGraph
617 unique topics in WordGraph
618 senses in WordGraph
22 lexemes in Wikidata
22 unique lemmas in Wikidata
16 unique topics in Wikidata
16 senses in Wikidata (possibe overcount)
95 forms in Wikidata
85 unique forms in Wikidata
0 shared forms
0 shared lemmas
0 shared senses
0 missing senses on existing Lexemes

Latvian

15348 forms in WordGraph
9296 unique forms in WordGraph
611 unique lemmas in WordGraph
638 unique topics in WordGraph
641 senses in WordGraph
273 lexemes in Wikidata
272 unique lemmas in Wikidata
122 unique topics in Wikidata
128 senses in Wikidata (possibe overcount)
2637 forms in Wikidata
1950 unique forms in Wikidata
54 shared forms
3 shared lemmas
3 shared senses
0 missing senses on existing Lexemes

Marathi

28748 forms in WordGraph
10371 unique forms in WordGraph
1340 unique lemmas in WordGraph
1280 unique topics in WordGraph
1397 senses in WordGraph
54 lexemes in Wikidata
50 unique lemmas in Wikidata
59 unique topics in Wikidata
116 senses in Wikidata (possibe overcount)
111 forms in Wikidata
94 unique forms in Wikidata
2 shared forms
2 shared lemmas
0 shared senses
2 missing senses on existing Lexemes

Dutch

7912 forms in WordGraph
5024 unique forms in WordGraph
1891 unique lemmas in WordGraph
1641 unique topics in WordGraph
1941 senses in WordGraph
560 lexemes in Wikidata
534 unique lemmas in Wikidata
396 unique topics in Wikidata
415 senses in Wikidata (possibe overcount)
1172 forms in Wikidata
1104 unique forms in Wikidata
13 shared forms
5 shared lemmas
4 shared senses
1 missing senses on existing Lexemes

Norwegian

Note: Norwegian was not properly mapped. Besides the first four stats, all others are not trustworthy.

18044 forms in WordGraph
8118 unique forms in WordGraph
1083 unique lemmas in WordGraph
1138 unique topics in WordGraph
1139 senses in WordGraph
0 lexemes in Wikidata
0 unique lemmas in Wikidata
0 unique topics in Wikidata
0 senses in Wikidata (possibe overcount)
0 forms in Wikidata
0 unique forms in Wikidata
0 shared forms
0 shared lemmas
0 shared senses
0 missing senses on existing Lexemes

Polish

82047 forms in WordGraph
28445 unique forms in WordGraph
1912 unique lemmas in WordGraph
1915 unique topics in WordGraph
2034 senses in WordGraph
3533 lexemes in Wikidata
3376 unique lemmas in Wikidata
2167 unique topics in Wikidata
3193 senses in Wikidata (possibe overcount)
31744 forms in Wikidata
19313 unique forms in Wikidata
204 shared forms
21 shared lemmas
20 shared senses
6 missing senses on existing Lexemes

Portuguese

15735 forms in WordGraph
9403 unique forms in WordGraph
3056 unique lemmas in WordGraph
2879 unique topics in WordGraph
3271 senses in WordGraph
4544 lexemes in Wikidata
4464 unique lemmas in Wikidata
2694 unique topics in Wikidata
3115 senses in Wikidata (possibe overcount)
46865 forms in Wikidata
34634 unique forms in Wikidata
561 shared forms
151 shared lemmas
116 shared senses
68 missing senses on existing Lexemes

Romanian

22990 forms in WordGraph
13208 unique forms in WordGraph
1245 unique lemmas in WordGraph
1303 unique topics in WordGraph
1310 senses in WordGraph
99 lexemes in Wikidata
98 unique lemmas in Wikidata
96 unique topics in Wikidata
100 senses in Wikidata (possibe overcount)
119 forms in Wikidata
101 unique forms in Wikidata
6 shared forms
2 shared lemmas
1 shared senses
1 missing senses on existing Lexemes

Russian

106335 forms in WordGraph
26400 unique forms in WordGraph
2182 unique lemmas in WordGraph
2174 unique topics in WordGraph
2312 senses in WordGraph
101971 lexemes in Wikidata
99742 unique lemmas in Wikidata
12077 unique topics in Wikidata
16131 senses in Wikidata (possibe overcount)
1243365 forms in Wikidata
913065 unique forms in Wikidata
10412 shared forms
380 shared lemmas
351 shared senses
69 missing senses on existing Lexemes

Slovak

45758 forms in WordGraph
20794 unique forms in WordGraph
1354 unique lemmas in WordGraph
1390 unique topics in WordGraph
1401 senses in WordGraph
16477 lexemes in Wikidata
16216 unique lemmas in Wikidata
574 unique topics in Wikidata
643 senses in Wikidata (possibe overcount)
235262 forms in Wikidata
128234 unique forms in Wikidata
3141 shared forms
43 shared lemmas
36 shared senses
9 missing senses on existing Lexemes

Slovenian

53478 forms in WordGraph
21836 unique forms in WordGraph
1545 unique lemmas in WordGraph
1609 unique topics in WordGraph
1616 senses in WordGraph
32 lexemes in Wikidata
32 unique lemmas in Wikidata
30 unique topics in Wikidata
32 senses in Wikidata (possibe overcount)
170 forms in Wikidata
103 unique forms in Wikidata
11 shared forms
2 shared lemmas
1 shared senses
1 missing senses on existing Lexemes

Serbian

45843 forms in WordGraph
15997 unique forms in WordGraph
1305 unique lemmas in WordGraph
1338 unique topics in WordGraph
1350 senses in WordGraph
25 lexemes in Wikidata
25 unique lemmas in Wikidata
18 unique topics in Wikidata
18 senses in Wikidata (possibe overcount)
39 forms in Wikidata
32 unique forms in Wikidata
0 shared forms
0 shared lemmas
0 shared senses
0 missing senses on existing Lexemes

Swedish

13664 forms in WordGraph
11132 unique forms in WordGraph
1589 unique lemmas in WordGraph
1621 unique topics in WordGraph
1660 senses in WordGraph
42779 lexemes in Wikidata
42385 unique lemmas in Wikidata
8256 unique topics in Wikidata
9251 senses in Wikidata (possibe overcount)
294289 forms in Wikidata
262349 unique forms in Wikidata
4351 shared forms
405 shared lemmas
361 shared senses
83 missing senses on existing Lexemes

Tamil

23618 forms in WordGraph
11544 unique forms in WordGraph
1249 unique lemmas in WordGraph
1129 unique topics in WordGraph
1296 senses in WordGraph
883 lexemes in Wikidata
871 unique lemmas in Wikidata
706 unique topics in Wikidata
1057 senses in Wikidata (possibe overcount)
6670 forms in Wikidata
6572 unique forms in Wikidata
140 shared forms
22 shared lemmas
19 shared senses
8 missing senses on existing Lexemes

Telugu

14751 forms in WordGraph
9672 unique forms in WordGraph
1129 unique lemmas in WordGraph
1133 unique topics in WordGraph
1173 senses in WordGraph
75 lexemes in Wikidata
72 unique lemmas in Wikidata
38 unique topics in Wikidata
50 senses in Wikidata (possibe overcount)
52 forms in Wikidata
52 unique forms in Wikidata
3 shared forms
2 shared lemmas
2 shared senses
0 missing senses on existing Lexemes

Turkish

12078 forms in WordGraph
10058 unique forms in WordGraph
851 unique lemmas in WordGraph
905 unique topics in WordGraph
906 senses in WordGraph
1815 lexemes in Wikidata
1766 unique lemmas in Wikidata
1081 unique topics in Wikidata
1425 senses in Wikidata (possibe overcount)
2517 forms in Wikidata
2298 unique forms in Wikidata
24 shared forms
15 shared lemmas
12 shared senses
10 missing senses on existing Lexemes

Ukrainian

47172 forms in WordGraph
19775 unique forms in WordGraph
1175 unique lemmas in WordGraph
1219 unique topics in WordGraph
1221 senses in WordGraph
16295 lexemes in Wikidata
16293 unique lemmas in Wikidata
310 unique topics in Wikidata
319 senses in Wikidata (possibe overcount)
508091 forms in Wikidata
238647 unique forms in Wikidata
73 shared forms
4 shared lemmas
2 shared senses
2 missing senses on existing Lexemes

Urdu

Note: Urdu was not properly mapped. Besides the first five stats, all others are not trustworthy.

126 forms in WordGraph
77 unique forms in WordGraph
69 unique lemmas in WordGraph
68 unique topics in WordGraph
70 senses in WordGraph
0 lexemes in Wikidata
0 unique lemmas in Wikidata
0 unique topics in Wikidata
0 senses in Wikidata (possibe overcount)
0 forms in Wikidata
0 unique forms in Wikidata
0 shared forms
0 shared lemmas
0 shared senses
0 missing senses on existing Lexemes