home edit page issue tracker

This page pertains to UD version 2.

Treebank Statistics: UD_Croatian-SET: POS Tags: PROPN

There are 4330 PROPN lemmas (22%), 5458 PROPN types (15%) and 12848 PROPN tokens (6%). Out of 17 observed tags, the rank of PROPN is: 3 in number of lemmas, 3 in number of types and 6 in number of tokens.

The 10 most frequent PROPN lemmas: EU, Hrvatska, Kosovo, Srbija, BiH, Makedonija, Turska, Europa, SETimes, NATO

The 10 most frequent PROPN types: EU, BiH, Kosova, Hrvatskoj, Srbije, SETimes, Hrvatska, Kosovo, Srbija, UN-a

The 10 most frequent ambiguous lemmas: BiH (PROPN 164, NOUN 1), Windows (PROPN 77, NOUN 1), bog (PROPN 14, NOUN 7), Velika (PROPN 23, ADJ 1), Hrvat (PROPN 17, NOUN 1), Ferrari (PROPN 15, NOUN 2), Facebook (PROPN 13, X 1), Internet (PROPN 13, ADJ 1), media (NOUN 1, PROPN 1), Explorer (PROPN 11, NOUN 1)

The 10 most frequent ambiguous types: BiH (PROPN 164, NOUN 1), Hrvatska (PROPN 93, ADJ 11), Hrvatske (PROPN 62, ADJ 23), Europe (PROPN 58, X 1), Turske (PROPN 45, ADJ 2), Bugarska (PROPN 41, ADJ 5), Turska (PROPN 40, ADJ 1), Crne (PROPN 38, ADJ 2), Bog (PROPN 32, NOUN 2), Bugarske (PROPN 28, ADJ 1)

Morphology

The form / lemma ratio of PROPN is 1.260508 (the average of all parts of speech is 1.848309).

The 1st highest number of forms (6) was observed with the lemma “Gora”: Gora, Gore, Gori, Gorin, Gorom, Goru.

The 2nd highest number of forms (6) was observed with the lemma “Hrvatska”: HRVATSKE, Hrvatska, Hrvatske, Hrvatskoj, Hrvatskom, Hrvatsku.

The 3rd highest number of forms (6) was observed with the lemma “Windows”: Windows, Windowsa, Windowse, Windowsi, Windowsima, Windowsu.

PROPN occurs with 4 features: Case (12825; 100% instances), Gender (12825; 100% instances), Number (12825; 100% instances), Animacy (644; 5% instances)

PROPN occurs with 14 feature-value pairs: Animacy=Anim, Animacy=Inan, Case=Acc, Case=Dat, Case=Gen, Case=Ins, Case=Loc, Case=Nom, Case=Voc, Gender=Fem, Gender=Masc, Gender=Neut, Number=Plur, Number=Sing

PROPN occurs with 33 feature combinations. The most frequent feature combination is Case=Nom|Gender=Masc|Number=Sing (4660 tokens). Examples: Windows, EU, Bog, Sanader, Zagreb, Nikola, Crvenkovski, Erdogan, HDZ, May

Relations

PROPN nodes are attached to their parents using 20 different relations: flat (4248; 33% instances), nmod (2980; 23% instances), nsubj (2393; 19% instances), conj (1143; 9% instances), obl (1038; 8% instances), obj (314; 2% instances), appos (217; 2% instances), amod (188; 1% instances), parataxis (182; 1% instances), iobj (48; 0% instances), root (47; 0% instances), acl (10; 0% instances), xcomp (10; 0% instances), orphan (7; 0% instances), flat:foreign (6; 0% instances), ccomp (5; 0% instances), advcl (4; 0% instances), vocative (4; 0% instances), discourse (2; 0% instances), list (2; 0% instances)

Parents of PROPN nodes belong to 13 different parts of speech: NOUN (5217; 41% instances), PROPN (3680; 29% instances), VERB (3161; 25% instances), ADJ (512; 4% instances), ADV (79; 1% instances), (47; 0% instances), AUX (46; 0% instances), X (46; 0% instances), DET (22; 0% instances), PRON (18; 0% instances), NUM (15; 0% instances), ADP (3; 0% instances), PART (2; 0% instances)

6253 (49%) PROPN nodes are leaves.

4426 (34%) PROPN nodes have one child.

1443 (11%) PROPN nodes have two children.

726 (6%) PROPN nodes have three or more children.

The highest child degree of a PROPN node is 13.

Children of PROPN nodes are attached using 30 different relations: flat (2626; 26% instances), case (2395; 24% instances), punct (1442; 14% instances), conj (1179; 12% instances), cc (679; 7% instances), amod (556; 5% instances), appos (266; 3% instances), nmod (247; 2% instances), acl (199; 2% instances), discourse (94; 1% instances), parataxis (89; 1% instances), advmod (72; 1% instances), nummod (68; 1% instances), cop (53; 1% instances), nsubj (49; 0% instances), orphan (26; 0% instances), det (20; 0% instances), aux (17; 0% instances), mark (15; 0% instances), flat:foreign (5; 0% instances), nummod:gov (5; 0% instances), obj (5; 0% instances), advcl (4; 0% instances), det:numgov (3; 0% instances), list (3; 0% instances), obl (3; 0% instances), xcomp (3; 0% instances), ccomp (2; 0% instances), dislocated (2; 0% instances), dep (1; 0% instances)

Children of PROPN nodes belong to 16 different parts of speech: PROPN (3680; 36% instances), ADP (2384; 24% instances), PUNCT (1442; 14% instances), CCONJ (771; 8% instances), NOUN (698; 7% instances), ADJ (498; 5% instances), VERB (147; 1% instances), NUM (120; 1% instances), ADV (100; 1% instances), X (79; 1% instances), AUX (72; 1% instances), PART (53; 1% instances), DET (41; 0% instances), SCONJ (31; 0% instances), PRON (11; 0% instances), SYM (1; 0% instances)