home edit page issue tracker

This page pertains to UD version 2.

Treebank Statistics: UD_English-EWT: POS Tags: ADJ

There are 2272 ADJ lemmas (12%), 2449 ADJ types (11%) and 16807 ADJ tokens (7%). Out of 17 observed tags, the rank of ADJ is: 3 in number of lemmas, 4 in number of types and 7 in number of tokens.

The 10 most frequent ADJ lemmas: good, great, new, other, many, more, last, same, few, little

The 10 most frequent ADJ types: good, great, new, other, best, many, more, last, same, few

The 10 most frequent ambiguous lemmas: good (ADJ 845, NOUN 8, ADV 2, INTJ 1), great (ADJ 378, ADV 2, INTJ 1), other (ADJ 328, NOUN 51, ADV 3), many (ADJ 219, DET 2), more (ADJ 212, ADV 163), last (ADJ 160, VERB 19, ADV 1), little (ADJ 140, ADV 7), sure (ADJ 139, ADV 6, INTJ 4), first (ADJ 130, ADV 65, NOUN 3), bad (ADJ 130, ADV 7, NOUN 2)

The 10 most frequent ambiguous types: good (ADJ 379, NOUN 8, ADV 2), great (ADJ 252, ADV 2, INTJ 1), new (ADJ 252, VERB 3, NOUN 1), other (ADJ 320, ADV 2), best (ADJ 211, ADV 22), many (ADJ 193, DET 2), more (ADJ 200, ADV 156), last (ADJ 145, VERB 7, ADV 1), little (ADJ 135, ADV 6), sure (ADJ 138, ADV 5)

Morphology

The form / lemma ratio of ADJ is 1.077905 (the average of all parts of speech is 1.236902).

The 1st highest number of forms (5) was observed with the lemma “easy”: eaiser, easier, easiest, east, easy.

The 2nd highest number of forms (5) was observed with the lemma “good”: bast, best, better, good, gud.

The 3rd highest number of forms (4) was observed with the lemma “bad”: bad, badder, worse, worst.

ADJ occurs with 8 features: Degree (16770; 100% instances), NumForm (257; 2% instances), NumType (257; 2% instances), ExtPos (92; 1% instances), Typo (92; 1% instances), Number (7; 0% instances), Style (7; 0% instances), Abbr (3; 0% instances)

ADJ occurs with 14 feature-value pairs: Abbr=Yes, Degree=Cmp, Degree=Pos, Degree=Sup, ExtPos=ADP, ExtPos=ADV, ExtPos=SCONJ, NumForm=Combi, NumForm=Word, NumType=Frac, NumType=Ord, Number=Sing, Style=Expr, Typo=Yes

ADJ occurs with 20 feature combinations. The most frequent feature combination is Degree=Pos (15151 tokens). Examples: good, great, new, other, many, last, same, few, little, sure

Relations

ADJ nodes are attached to their parents using 31 different relations: amod (11391; 68% instances), root (1710; 10% instances), conj (1101; 7% instances), xcomp (682; 4% instances), ccomp (316; 2% instances), advcl (307; 2% instances), parataxis (246; 1% instances), obj (162; 1% instances), obl (148; 1% instances), acl:relcl (138; 1% instances), nmod (112; 1% instances), nsubj (105; 1% instances), advmod (65; 0% instances), obl:npmod (48; 0% instances), fixed (46; 0% instances), case (39; 0% instances), appos (38; 0% instances), compound (33; 0% instances), acl (31; 0% instances), list (22; 0% instances), advcl:relcl (15; 0% instances), nsubj:pass (12; 0% instances), csubj (11; 0% instances), mark (10; 0% instances), nmod:npmod (4; 0% instances), obl:tmod (4; 0% instances), orphan (4; 0% instances), discourse (3; 0% instances), nsubj:outer (2; 0% instances), csubj:outer (1; 0% instances), obl:agent (1; 0% instances)

Parents of ADJ nodes belong to 14 different parts of speech: NOUN (10529; 63% instances), VERB (2020; 12% instances), (1710; 10% instances), PROPN (1134; 7% instances), ADJ (988; 6% instances), PRON (160; 1% instances), NUM (100; 1% instances), ADV (63; 0% instances), DET (36; 0% instances), AUX (25; 0% instances), ADP (22; 0% instances), SYM (18; 0% instances), INTJ (1; 0% instances), X (1; 0% instances)

10594 (63%) ADJ nodes are leaves.

2313 (14%) ADJ nodes have one child.

799 (5%) ADJ nodes have two children.

3101 (18%) ADJ nodes have three or more children.

The highest child degree of a ADJ node is 13.

Children of ADJ nodes are attached using 41 different relations: punct (3019; 16% instances), cop (2885; 16% instances), advmod (2521; 14% instances), nsubj (2512; 14% instances), obl (1162; 6% instances), conj (1137; 6% instances), cc (998; 5% instances), advcl (661; 4% instances), mark (497; 3% instances), aux (403; 2% instances), det (312; 2% instances), case (286; 2% instances), parataxis (263; 1% instances), xcomp (253; 1% instances), ccomp (233; 1% instances), obl:npmod (181; 1% instances), csubj (169; 1% instances), expl (140; 1% instances), amod (109; 1% instances), fixed (92; 1% instances), discourse (85; 0% instances), nmod (78; 0% instances), obl:tmod (47; 0% instances), compound (33; 0% instances), nmod:poss (29; 0% instances), obj (29; 0% instances), nsubj:outer (26; 0% instances), list (22; 0% instances), acl:relcl (21; 0% instances), appos (19; 0% instances), nummod (19; 0% instances), goeswith (16; 0% instances), vocative (13; 0% instances), acl (10; 0% instances), cc:preconj (10; 0% instances), advcl:relcl (7; 0% instances), reparandum (4; 0% instances), det:predet (3; 0% instances), orphan (3; 0% instances), compound:prt (2; 0% instances), nmod:npmod (1; 0% instances)

Children of ADJ nodes belong to 17 different parts of speech: AUX (3326; 18% instances), PUNCT (3019; 16% instances), NOUN (2307; 13% instances), ADV (2281; 12% instances), VERB (1663; 9% instances), PRON (1633; 9% instances), CCONJ (994; 5% instances), ADJ (988; 5% instances), SCONJ (408; 2% instances), ADP (392; 2% instances), PART (376; 2% instances), PROPN (374; 2% instances), DET (363; 2% instances), INTJ (59; 0% instances), NUM (56; 0% instances), SYM (47; 0% instances), X (24; 0% instances)