home edit page issue tracker

This page pertains to UD version 2.

Treebank Statistics: UD_Latvian-LVTB: POS Tags: PROPN

There are 4007 PROPN lemmas (18%), 5240 PROPN types (11%) and 11727 PROPN tokens (4%). Out of 17 observed tags, the rank of PROPN is: 3 in number of lemmas, 4 in number of types and 7 in number of tokens.

The 10 most frequent PROPN lemmas: Latvija, Rīga, Eiropa, Krievija, Sofija, ES, LETA, Saeima, Andris, Baltija

The 10 most frequent PROPN types: Latvijas, Latvijā, Eiropas, Rīgas, ES, Krievijas, LETA, Sofija, Baltijas, Rīgā

The 10 most frequent ambiguous lemmas: EK (PROPN 15, NOUN 1), g. (NOUN 11, PROPN 4), FM (PROPN 7, X 1), V (PROPN 6, ADJ 1, NUM 1), Positivus (PROPN 5, X 1), BAS (PROPN 4, X 1), Google (PROPN 4, X 1), SEB (PROPN 4, SYM 1), ABLV (PROPN 3, SYM 1), BKUS (PROPN 3, NOUN 1)

The 10 most frequent ambiguous types: Liepājas (PROPN 33, NOUN 2), Saules (PROPN 26, NOUN 1), Satversmes (PROPN 21, NOUN 1), EK (PROPN 15, NOUN 1), Jāņu (PROPN 15, NOUN 1), Mēness (PROPN 14, NOUN 2), airBaltic (PROPN 10, X 1), vilnis (NOUN 3, PROPN 1), Citadele (PROPN 7, NOUN 3), FM (PROPN 7, X 1)

Morphology

The form / lemma ratio of PROPN is 1.307712 (the average of all parts of speech is 2.233228).

The 1st highest number of forms (7) was observed with the lemma “Jānis”: Jāni, Jānim, Jānis, Jāņa, Jāņi, Jāņiem, Jāņu.

The 2nd highest number of forms (6) was observed with the lemma “Eiropa”: EIROPAS, Eiropa, Eiropai, Eiropas, Eiropu, Eiropā.

The 3rd highest number of forms (6) was observed with the lemma “Rīga”: RĪGAS, Rīga, Rīgai, Rīgas, Rīgu, Rīgā.

PROPN occurs with 5 features: Gender (10118; 86% instances), Case (10100; 86% instances), Number (10100; 86% instances), Abbr (1130; 10% instances), Typo (14; 0% instances)

PROPN occurs with 13 feature-value pairs: Abbr=Yes, Case=Acc, Case=Dat, Case=Gen, Case=Loc, Case=Nom, Case=Voc, Gender=Fem, Gender=Masc, Number=Plur, Number=Ptan, Number=Sing, Typo=Yes

PROPN occurs with 42 feature combinations. The most frequent feature combination is Case=Gen|Gender=Fem|Number=Sing (2607 tokens). Examples: Latvijas, Eiropas, Rīgas, Krievijas, Baltijas, Jelgavas, Bauskas, Saeimas, Ukrainas, Liepājas

Relations

PROPN nodes are attached to their parents using 21 different relations: nmod (3924; 33% instances), nsubj (2618; 22% instances), flat:name (1809; 15% instances), obl (1229; 10% instances), conj (884; 8% instances), iobj (352; 3% instances), obj (259; 2% instances), parataxis (182; 2% instances), root (147; 1% instances), appos (88; 1% instances), nsubj:pass (81; 1% instances), discourse (35; 0% instances), vocative (31; 0% instances), dep (24; 0% instances), orphan (23; 0% instances), xcomp (19; 0% instances), acl (18; 0% instances), advcl (1; 0% instances), ccomp (1; 0% instances), csubj (1; 0% instances), flat:foreign (1; 0% instances)

Parents of PROPN nodes belong to 15 different parts of speech: VERB (4348; 37% instances), NOUN (4242; 36% instances), PROPN (2699; 23% instances), (147; 1% instances), ADJ (99; 1% instances), ADV (55; 0% instances), X (49; 0% instances), PRON (35; 0% instances), NUM (34; 0% instances), AUX (7; 0% instances), INTJ (5; 0% instances), SYM (3; 0% instances), DET (2; 0% instances), CCONJ (1; 0% instances), PUNCT (1; 0% instances)

7318 (62%) PROPN nodes are leaves.

2086 (18%) PROPN nodes have one child.

1381 (12%) PROPN nodes have two children.

942 (8%) PROPN nodes have three or more children.

The highest child degree of a PROPN node is 21.

Children of PROPN nodes are attached using 23 different relations: punct (2017; 24% instances), flat:name (1841; 22% instances), nmod (1462; 17% instances), conj (927; 11% instances), case (672; 8% instances), cc (472; 6% instances), acl (193; 2% instances), parataxis (181; 2% instances), discourse (162; 2% instances), amod (125; 1% instances), appos (61; 1% instances), det (55; 1% instances), orphan (44; 1% instances), nsubj (32; 0% instances), advmod (31; 0% instances), cop (30; 0% instances), dep (23; 0% instances), advcl (16; 0% instances), obl (16; 0% instances), mark (8; 0% instances), nummod (6; 0% instances), aux (1; 0% instances), iobj (1; 0% instances)

Children of PROPN nodes belong to 16 different parts of speech: PROPN (2699; 32% instances), PUNCT (2017; 24% instances), NOUN (1677; 20% instances), ADP (638; 8% instances), CCONJ (465; 6% instances), VERB (240; 3% instances), PART (136; 2% instances), ADJ (125; 1% instances), X (76; 1% instances), ADV (72; 1% instances), NUM (60; 1% instances), DET (50; 1% instances), SYM (37; 0% instances), AUX (31; 0% instances), SCONJ (31; 0% instances), PRON (22; 0% instances)