home edit page issue tracker

This page pertains to UD version 2.

Treebank Statistics: UD_Bulgarian-BTB: POS Tags: ADJ

There are 3122 ADJ lemmas (20%), 6323 ADJ types (23%) and 13591 ADJ tokens (9%). Out of 16 observed tags, the rank of ADJ is: 2 in number of lemmas, 3 in number of types and 5 in number of tokens.

The 10 most frequent ADJ lemmas: нов, друг, български, голям, народен, пръв, държавен, европейски, цял, втори

The 10 most frequent ADJ types: други, народното, българската, нова, другите, нови, европейската, последните, 2001, друг

The 10 most frequent ambiguous lemmas: нов (ADJ 301, PROPN 3), български (ADJ 229, ADV 2), голям (ADJ 229, PROPN 1), европейски (ADJ 124, ADV 1), политически (ADJ 107, ADV 3), икономически (ADJ 54, ADV 2), следвам (ADJ 52, VERB 14), мина-(се) (ADJ 50, VERB 37), стар (ADJ 49, PROPN 4), син (ADJ 44, NOUN 20)

The 10 most frequent ambiguous types: 2001 (ADJ 42, NUM 1), 2000 (ADJ 40, NUM 12, PROPN 4), български (ADJ 30, ADV 2), политически (ADJ 33, ADV 3), 1 (NUM 53, ADJ 29, PROPN 1), II (ADJ 18, PROPN 1), останалите (ADJ 10, VERB 1), европейски (ADJ 13, ADV 1), Южна (ADJ 14, PROPN 1), свързани (ADJ 13, VERB 3)

Morphology

The form / lemma ratio of ADJ is 2.025304 (the average of all parts of speech is 1.727904).

The 1st highest number of forms (25) was observed with the lemma “голям”: големи, големите, големия, големият, голям, голяма, голямата, голямо, най-големи, най-големите, най-големия, най-големият, най-голям, най-голяма, най-голямата, най-голямо, най-голямото, по-големи, по-големите, по-големия, по-голям, по-голяма, по-голямата, по-голямо, по-голямото.

The 2nd highest number of forms (21) was observed with the lemma “добър”: Най-добра, добра, добрата, добри, добрите, добрият, добро, доброто, добър, най-добрата, най-добри, най-добрите, най-добрия, най-добрият, най-доброто, най-добър, по-добра, по-добри, по-добрият, по-добро, по-добър.

The 3rd highest number of forms (17) was observed with the lemma “висок”: висок, висока, високата, високи, високите, високия, високо, високото, най-висок, най-високата, най-високите, най-високо, по-висок, по-висока, по-високи, по-високите, по-високо.

ADJ occurs with 10 features: Number (13504; 99% instances), Definite (13480; 99% instances), Degree (13274; 98% instances), Gender (9557; 70% instances), Aspect (1486; 11% instances), VerbForm (1486; 11% instances), Voice (1486; 11% instances), NumType (906; 7% instances), Tense (521; 4% instances), Case (24; 0% instances)

ADJ occurs with 19 feature-value pairs: Aspect=Imp, Aspect=Perf, Case=Voc, Definite=Def, Definite=Ind, Degree=Cmp, Degree=Pos, Degree=Sup, Gender=Fem, Gender=Masc, Gender=Neut, NumType=Ord, Number=Plur, Number=Sing, Tense=Past, Tense=Pres, VerbForm=Part, Voice=Act, Voice=Pass

ADJ occurs with 98 feature combinations. The most frequent feature combination is Definite=Ind|Degree=Pos|Number=Plur (1817 tokens). Examples: други, нови, различни, големи, български, народни, добри, подобни, политически, финансови

Relations

ADJ nodes are attached to their parents using 22 different relations: amod (11829; 87% instances), conj (454; 3% instances), root (404; 3% instances), obj (201; 1% instances), nsubj (165; 1% instances), nmod (143; 1% instances), ccomp (106; 1% instances), iobj (58; 0% instances), obl (50; 0% instances), advcl (47; 0% instances), acl (43; 0% instances), xcomp (19; 0% instances), parataxis (17; 0% instances), flat (14; 0% instances), csubj (13; 0% instances), nsubj:pass (13; 0% instances), discourse (6; 0% instances), csubj:pass (3; 0% instances), vocative (3; 0% instances), compound (1; 0% instances), nummod (1; 0% instances), orphan (1; 0% instances)

Parents of ADJ nodes belong to 10 different parts of speech: NOUN (11710; 86% instances), VERB (707; 5% instances), (404; 3% instances), ADJ (376; 3% instances), PROPN (321; 2% instances), NUM (24; 0% instances), DET (19; 0% instances), ADV (15; 0% instances), PRON (11; 0% instances), PART (4; 0% instances)

10784 (79%) ADJ nodes are leaves.

1491 (11%) ADJ nodes have one child.

443 (3%) ADJ nodes have two children.

873 (6%) ADJ nodes have three or more children.

The highest child degree of a ADJ node is 8.

Children of ADJ nodes are attached using 29 different relations: punct (1265; 21% instances), obl (862; 14% instances), advmod (663; 11% instances), cop (660; 11% instances), nsubj (546; 9% instances), conj (489; 8% instances), det (381; 6% instances), cc (375; 6% instances), case (281; 5% instances), mark (109; 2% instances), aux (93; 2% instances), advcl (63; 1% instances), expl (61; 1% instances), discourse (30; 1% instances), iobj (26; 0% instances), aux:pass (25; 0% instances), acl (22; 0% instances), nsubj:pass (20; 0% instances), flat (13; 0% instances), obj (5; 0% instances), vocative (3; 0% instances), amod (1; 0% instances), appos (1; 0% instances), ccomp (1; 0% instances), csubj (1; 0% instances), nmod (1; 0% instances), nummod (1; 0% instances), orphan (1; 0% instances), parataxis (1; 0% instances)

Children of ADJ nodes belong to 15 different parts of speech: PUNCT (1265; 21% instances), NOUN (1154; 19% instances), AUX (754; 13% instances), PRON (671; 11% instances), ADV (601; 10% instances), ADJ (376; 6% instances), CCONJ (375; 6% instances), ADP (282; 5% instances), VERB (190; 3% instances), PROPN (121; 2% instances), SCONJ (98; 2% instances), PART (96; 2% instances), INTJ (8; 0% instances), DET (6; 0% instances), NUM (3; 0% instances)