home edit page issue tracker

This page pertains to UD version 2.

Treebank Statistics: UD_French-Sequoia: POS Tags: PROPN

There are 1164 PROPN lemmas (16%), 1172 PROPN types (12%) and 3537 PROPN tokens (5%). Out of 16 observed tags, the rank of PROPN is: 2 in number of lemmas, 4 in number of types and 7 in number of tokens.

The 10 most frequent PROPN lemmas: Aclasta, Angiox, Paris, commission, Jacques, France, Parlement, Union, Chirac, RPR

The 10 most frequent PROPN types: Aclasta, Angiox, paris, commission, Jacques, France, Parlement, Union, Chirac, RPR

The 10 most frequent ambiguous lemmas: commission (NOUN 64, PROPN 53), RPR (PROPN 37, NOUN 2), conseil (PROPN 34, NOUN 26), Reuters (PROPN 18, X 1), UE (PROPN 8, NOUN 1), société (NOUN 24, PROPN 1), république (NOUN 15, PROPN 2), banque (NOUN 9, PROPN 6), Koweit (PROPN 4, NOUN 1), Beirut (PROPN 3, X 1)

The 10 most frequent ambiguous types: commission (NOUN 35, PROPN 2), RPR (PROPN 37, NOUN 2), conseil (NOUN 19, PROPN 1), Reuters (PROPN 18, X 1), UE (PROPN 8, NOUN 1), société (NOUN 17, PROPN 1), Parti (PROPN 7, NOUN 3, VERB 1), république (NOUN 1, PROPN 1), Guerre (NOUN 10, PROPN 5), Ligue (PROPN 5, NOUN 1)

Morphology

The form / lemma ratio of PROPN is 1.006873 (the average of all parts of speech is 1.399011).

The 1st highest number of forms (2) was observed with the lemma “Aclasta”: ACLASTA, Aclasta.

The 2nd highest number of forms (2) was observed with the lemma “Angiox”: ANGIOX, Angiox.

The 3rd highest number of forms (2) was observed with the lemma “Courcelles-lès-Montbéliard”: COURCELLES-LES-MONTBELIARD, Courcelles-lès-Montbéliard.

PROPN occurs with 2 features: Number (1750; 49% instances), Gender (1589; 45% instances)

PROPN occurs with 4 feature-value pairs: Gender=Fem, Gender=Masc, Number=Plur, Number=Sing

PROPN occurs with 8 feature combinations. The most frequent feature combination is _ (1786 tokens). Examples: Aclasta, Angiox, RPR, Halphen, Jean-Claude, Méry, Schuller, Thomson, Parlement, Reuters

Relations

PROPN nodes are attached to their parents using 18 different relations: nmod (1338; 38% instances), flat:name (898; 25% instances), nsubj (301; 9% instances), obl (299; 8% instances), conj (205; 6% instances), root (130; 4% instances), appos (96; 3% instances), obj (87; 2% instances), nsubj:pass (81; 2% instances), obl:agent (70; 2% instances), dep (9; 0% instances), dislocated (7; 0% instances), xcomp (7; 0% instances), orphan (4; 0% instances), acl:relcl (2; 0% instances), ccomp (1; 0% instances), det (1; 0% instances), nsubj:caus (1; 0% instances)

Parents of PROPN nodes belong to 8 different parts of speech: NOUN (1529; 43% instances), PROPN (1019; 29% instances), VERB (790; 22% instances), (130; 4% instances), ADJ (36; 1% instances), PRON (20; 1% instances), X (10; 0% instances), NUM (3; 0% instances)

1249 (35%) PROPN nodes are leaves.

1108 (31%) PROPN nodes have one child.

646 (18%) PROPN nodes have two children.

534 (15%) PROPN nodes have three or more children.

The highest child degree of a PROPN node is 19.

Children of PROPN nodes are attached using 20 different relations: case (1376; 29% instances), punct (745; 16% instances), det (709; 15% instances), flat:name (704; 15% instances), nmod (332; 7% instances), conj (227; 5% instances), amod (159; 3% instances), appos (136; 3% instances), cc (125; 3% instances), dep (52; 1% instances), acl (45; 1% instances), nummod (42; 1% instances), acl:relcl (34; 1% instances), advmod (14; 0% instances), mark (12; 0% instances), orphan (7; 0% instances), cop (5; 0% instances), nsubj (5; 0% instances), parataxis (3; 0% instances), ccomp (2; 0% instances)

Children of PROPN nodes belong to 14 different parts of speech: ADP (1374; 29% instances), PROPN (1019; 22% instances), PUNCT (745; 16% instances), DET (709; 15% instances), NOUN (381; 8% instances), ADJ (160; 3% instances), CCONJ (125; 3% instances), VERB (84; 2% instances), X (51; 1% instances), NUM (43; 1% instances), ADV (16; 0% instances), PRON (11; 0% instances), SCONJ (11; 0% instances), AUX (5; 0% instances)