home edit page issue tracker

This page pertains to UD version 2.

Treebank Statistics: UD_English-GUM: POS Tags: PROPN

There are 4489 PROPN lemmas (26%), 4649 PROPN types (22%) and 13280 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: State, University, President, America, York, New, American, south, Warhol, Figure

The 10 most frequent PROPN types: States, University, President, York, America, New, south, Warhol, State, figure

The 10 most frequent ambiguous lemmas: New (PROPN 56, ADJ 50), American (ADJ 81, PROPN 50), south (NOUN 17, ADV 12, ADJ 2, PROPN 1), Nation (PROPN 46, NOUN 2), north (ADV 15, NOUN 9, PROPN 2), US (PROPN 43, ADJ 3, NOUN 1), Union (PROPN 33, NOUN 1), English (PROPN 31, ADJ 20), de (PROPN 25, X 8), east (ADV 14, NOUN 8, ADJ 3, PROPN 3)

The 10 most frequent ambiguous types: States (PROPN 95, NOUN 1), University (PROPN 91, NOUN 2), New (PROPN 56, ADJ 54), south (NOUN 17, ADV 11, ADJ 2, PROPN 1), State (PROPN 48, NOUN 3, VERB 1), figure (NOUN 14, VERB 9, PROPN 2), north (ADV 15, NOUN 8, PROPN 2), US (PROPN 43, ADJ 3, NOUN 1), party (NOUN 30, PROPN 3, VERB 1), world (NOUN 153, PROPN 1)

Morphology

The form / lemma ratio of PROPN is 1.035643 (the average of all parts of speech is 1.243866).

The 1st highest number of forms (4) was observed with the lemma “Saturday”: Sat, Sat., Saturday, Saturdays.

The 2nd highest number of forms (3) was observed with the lemma “American”: AMERICANS, American, Americans.

The 3rd highest number of forms (3) was observed with the lemma “Chatham”: Chatham, Chathams, Chatnam.

PROPN occurs with 6 features: Number (13279; 100% instances), Abbr (573; 4% instances), Foreign (71; 1% instances), Typo (35; 0% instances), NumForm (2; 0% instances), NumType (2; 0% instances)

PROPN occurs with 8 feature-value pairs: Abbr=Yes, Foreign=Yes, NumForm=Word, NumType=Card, Number=Plur, Number=Ptan, Number=Sing, Typo=Yes

PROPN occurs with 12 feature combinations. The most frequent feature combination is Number=Sing (11912 tokens). Examples: University, President, York, America, New, south, Warhol, State, figure, Scientology

Relations

PROPN nodes are attached to their parents using 31 different relations: compound (2500; 19% instances), nmod (1983; 15% instances), flat (1707; 13% instances), nsubj (1477; 11% instances), obl (1225; 9% instances), conj (938; 7% instances), root (630; 5% instances), appos (510; 4% instances), obj (463; 3% instances), nmod:poss (425; 3% instances), nmod:desc (257; 2% instances), nmod:unmarked (201; 2% instances), dep (170; 1% instances), nsubj:pass (147; 1% instances), list (110; 1% instances), obl:agent (100; 1% instances), vocative (98; 1% instances), xcomp (85; 1% instances), parataxis (80; 1% instances), iobj (38; 0% instances), obl:unmarked (29; 0% instances), advcl (21; 0% instances), orphan (20; 0% instances), ccomp (18; 0% instances), acl:relcl (15; 0% instances), dislocated (12; 0% instances), reparandum (7; 0% instances), nsubj:outer (5; 0% instances), acl (4; 0% instances), discourse (4; 0% instances), csubj (1; 0% instances)

Parents of PROPN nodes belong to 14 different parts of speech: PROPN (5510; 41% instances), VERB (3284; 25% instances), NOUN (3212; 24% instances), (630; 5% instances), NUM (237; 2% instances), ADJ (208; 2% instances), ADV (81; 1% instances), PRON (43; 0% instances), AUX (30; 0% instances), X (22; 0% instances), SYM (11; 0% instances), INTJ (7; 0% instances), DET (4; 0% instances), ADP (1; 0% instances)

4834 (36%) PROPN nodes are leaves.

3682 (28%) PROPN nodes have one child.

2259 (17%) PROPN nodes have two children.

2505 (19%) PROPN nodes have three or more children.

The highest child degree of a PROPN node is 14.

Children of PROPN nodes are attached using 40 different relations: case (3918; 22% instances), punct (2711; 15% instances), flat (1769; 10% instances), det (1718; 10% instances), compound (1644; 9% instances), amod (1071; 6% instances), conj (1030; 6% instances), nmod (837; 5% instances), cc (606; 3% instances), appos (408; 2% instances), nmod:unmarked (323; 2% instances), nmod:desc (259; 1% instances), cop (193; 1% instances), nsubj (189; 1% instances), nmod:poss (179; 1% instances), advmod (156; 1% instances), acl:relcl (134; 1% instances), list (130; 1% instances), parataxis (128; 1% instances), dep (97; 1% instances), discourse (83; 0% instances), acl (79; 0% instances), orphan (34; 0% instances), obl (33; 0% instances), nummod (30; 0% instances), mark (29; 0% instances), reparandum (19; 0% instances), cc:preconj (18; 0% instances), aux (17; 0% instances), obj (11; 0% instances), obl:unmarked (11; 0% instances), advcl (7; 0% instances), advcl:relcl (4; 0% instances), expl (4; 0% instances), vocative (4; 0% instances), csubj (2; 0% instances), xcomp (2; 0% instances), compound:prt (1; 0% instances), det:predet (1; 0% instances), goeswith (1; 0% instances)

Children of PROPN nodes belong to 17 different parts of speech: PROPN (5510; 31% instances), ADP (3418; 19% instances), PUNCT (2711; 15% instances), DET (1729; 10% instances), ADJ (1007; 6% instances), NUM (621; 3% instances), NOUN (610; 3% instances), CCONJ (598; 3% instances), PART (465; 3% instances), VERB (434; 2% instances), AUX (212; 1% instances), PRON (209; 1% instances), ADV (173; 1% instances), INTJ (68; 0% instances), X (66; 0% instances), SYM (35; 0% instances), SCONJ (24; 0% instances)