home edit page issue tracker

This page pertains to UD version 2.

Treebank Statistics: UD_English-EWT: POS Tags: PROPN

There are 4808 PROPN lemmas (26%), 4984 PROPN types (22%) and 16091 PROPN tokens (6%). Out of 17 observed tags, the rank of PROPN is: 2 in number of lemmas, 2 in number of types and 8 in number of tokens.

The 10 most frequent PROPN lemmas: Bush, US, al, Iraq, enron, State, Iran, China, September, Qaeda

The 10 most frequent PROPN types: bush, US, al, Iraq, enron, Iran, China, states, Qaeda, John

The 10 most frequent ambiguous lemmas: enron (PROPN 7, NOUN 5), president (NOUN 30, PROPN 1), American (ADJ 88, PROPN 54), mark (NOUN 13, VERB 12, PROPN 2, X 1), North (PROPN 34, ADJ 2), god (PROPN 5, NOUN 4), south (NOUN 10, ADV 6, ADJ 1, PROPN 1), air (NOUN 34, PROPN 1), bay (PROPN 2, NOUN 1), street (NOUN 29, PROPN 1)

The 10 most frequent ambiguous types: al (PROPN 67, X 1), states (NOUN 10, PROPN 6, VERB 5), John (PROPN 75, X 4), president (NOUN 24, PROPN 1), may (AUX 221, PROPN 1), google (PROPN 3, VERB 2), Vince (PROPN 45, X 1), mark (NOUN 10, VERB 6, PROPN 2), Paul (PROPN 35, X 1), north (NOUN 6, ADV 5, ADJ 2, PROPN 2)

Morphology

The form / lemma ratio of PROPN is 1.036606 (the average of all parts of speech is 1.236902).

The 1st highest number of forms (4) was observed with the lemma “Friday”: Fri, Fri., Fridays, friday.

The 2nd highest number of forms (4) was observed with the lemma “March”: MARCH, Mar, March, Marches.

The 3rd highest number of forms (4) was observed with the lemma “McDonald”: Mc.Donald, McDonal, mc, mcdonald.

PROPN occurs with 4 features: Number (16091; 100% instances), Abbr (121; 1% instances), Typo (27; 0% instances), Style (2; 0% instances)

PROPN occurs with 5 feature-value pairs: Abbr=Yes, Number=Plur, Number=Sing, Style=Expr, Typo=Yes

PROPN occurs with 7 feature combinations. The most frequent feature combination is Number=Sing (15218 tokens). Examples: bush, US, al, Iraq, enron, Iran, China, Qaeda, John, india

Relations

PROPN nodes are attached to their parents using 28 different relations: compound (3380; 21% instances), nsubj (1996; 12% instances), nmod (1981; 12% instances), flat (1842; 11% instances), obl (1667; 10% instances), root (1419; 9% instances), conj (1023; 6% instances), appos (638; 4% instances), obj (626; 4% instances), nmod:poss (483; 3% instances), list (201; 1% instances), obl:agent (141; 1% instances), vocative (133; 1% instances), nsubj:pass (125; 1% instances), iobj (76; 0% instances), xcomp (72; 0% instances), parataxis (64; 0% instances), obl:tmod (59; 0% instances), nmod:tmod (38; 0% instances), ccomp (32; 0% instances), advcl (28; 0% instances), nmod:npmod (25; 0% instances), obl:npmod (22; 0% instances), acl:relcl (8; 0% instances), acl (5; 0% instances), discourse (3; 0% instances), csubj (2; 0% instances), reparandum (2; 0% instances)

Parents of PROPN nodes belong to 14 different parts of speech: PROPN (5824; 36% instances), VERB (4227; 26% instances), NOUN (3954; 25% instances), (1419; 9% instances), ADJ (374; 2% instances), ADV (101; 1% instances), PRON (77; 0% instances), NUM (45; 0% instances), AUX (26; 0% instances), INTJ (18; 0% instances), SYM (14; 0% instances), DET (7; 0% instances), X (3; 0% instances), ADP (2; 0% instances)

6115 (38%) PROPN nodes are leaves.

4953 (31%) PROPN nodes have one child.

2712 (17%) PROPN nodes have two children.

2311 (14%) PROPN nodes have three or more children.

The highest child degree of a PROPN node is 18.

Children of PROPN nodes are attached using 37 different relations: case (4548; 23% instances), punct (2685; 14% instances), compound (2204; 11% instances), flat (1897; 10% instances), det (1465; 8% instances), conj (1173; 6% instances), amod (1131; 6% instances), cc (732; 4% instances), appos (630; 3% instances), nummod (610; 3% instances), list (574; 3% instances), nmod (517; 3% instances), cop (192; 1% instances), nsubj (181; 1% instances), advmod (154; 1% instances), acl:relcl (127; 1% instances), parataxis (126; 1% instances), nmod:poss (117; 1% instances), nmod:tmod (55; 0% instances), acl (47; 0% instances), mark (43; 0% instances), discourse (35; 0% instances), aux (30; 0% instances), cc:preconj (27; 0% instances), obl (24; 0% instances), nmod:npmod (20; 0% instances), advcl (11; 0% instances), expl (5; 0% instances), advcl:relcl (4; 0% instances), obl:tmod (4; 0% instances), nmod:desc (3; 0% instances), orphan (3; 0% instances), reparandum (3; 0% instances), goeswith (2; 0% instances), vocative (2; 0% instances), det:predet (1; 0% instances), obl:npmod (1; 0% instances)

Children of PROPN nodes belong to 17 different parts of speech: PROPN (5824; 30% instances), ADP (3943; 20% instances), PUNCT (2685; 14% instances), DET (1471; 8% instances), ADJ (1134; 6% instances), NOUN (1047; 5% instances), NUM (877; 5% instances), CCONJ (724; 4% instances), PART (586; 3% instances), VERB (322; 2% instances), AUX (224; 1% instances), PRON (176; 1% instances), ADV (146; 1% instances), X (103; 1% instances), SYM (61; 0% instances), SCONJ (34; 0% instances), INTJ (26; 0% instances)