home edit page issue tracker

This page pertains to UD version 2.

Treebank Statistics: UD_French-ALTS: POS Tags: NUM

There are 95 NUM lemmas (3%), 145 NUM types (2%) and 731 NUM tokens (1%). Out of 14 observed tags, the rank of NUM is: 6 in number of lemmas, 7 in number of types and 13 in number of tokens.

The 10 most frequent NUM lemmas: deux, trois, mille, quatre, cinq, six, sept, huit, quinze, vingt

The 10 most frequent NUM types: deulx, mille, troys, .iii., .vi., deux, quatre, chinq, .iiii., .viii.

The 10 most frequent ambiguous lemmas: neuf (NUM 11, ADJ 1), douze (NUM 7, PROPN 2), seize (NUM 7, ADJ 1), un (DET 744, PRON 83, NUM 4)

The 10 most frequent ambiguous types: .l. (NUM 4, NOUN 1), 13. (NUM 2, ADJ 1), 4. (NUM 2, ADJ 1), cinq (NUM 2, ADJ 1), .i. (DET 11, ADJ 2, NUM 1), 16. (ADJ 1, NUM 1), 20. (ADJ 1, NUM 1), neuff (ADJ 1, NUM 1), ung (DET 340, PRON 25, NUM 1)

Morphology

The form / lemma ratio of NUM is 1.526316 (the average of all parts of speech is 1.784739).

The 1st highest number of forms (6) was observed with the lemma “deux”: .ii., deulx, deulz, deux, dex, ij..

The 2nd highest number of forms (6) was observed with the lemma “trois”: .iii., .iiii., 3, 3., trois, troys.

The 3rd highest number of forms (5) was observed with the lemma “sept”: .vii., 7, 7., sept, vii.

NUM occurs with 1 features: NumType (719; 98% instances)

NUM occurs with 1 feature-value pairs: NumType=Card

NUM occurs with 2 feature combinations. The most frequent feature combination is NumType=Card (719 tokens). Examples: deulx, mille, troys, .iii., .vi., deux, quatre, chinq, .iiii., .viii.

Relations

NUM nodes are attached to their parents using 7 different relations: nummod (540; 74% instances), conj (67; 9% instances), flat (56; 8% instances), nmod (50; 7% instances), obl (9; 1% instances), obj (5; 1% instances), nsubj (4; 1% instances)

Parents of NUM nodes belong to 6 different parts of speech: NOUN (584; 80% instances), NUM (119; 16% instances), VERB (19; 3% instances), PROPN (5; 1% instances), PRON (3; 0% instances), ADJ (1; 0% instances)

489 (67%) NUM nodes are leaves.

214 (29%) NUM nodes have one child.

17 (2%) NUM nodes have two children.

11 (2%) NUM nodes have three or more children.

The highest child degree of a NUM node is 6.

Children of NUM nodes are attached using 12 different relations: conj (102; 35% instances), cc (66; 23% instances), flat (56; 19% instances), nmod (20; 7% instances), det (16; 5% instances), case (11; 4% instances), punct (7; 2% instances), advmod (5; 2% instances), acl:relcl (2; 1% instances), cop (2; 1% instances), mark (2; 1% instances), nsubj (2; 1% instances)

Children of NUM nodes belong to 12 different parts of speech: NUM (119; 41% instances), CCONJ (66; 23% instances), ADV (42; 14% instances), NOUN (20; 7% instances), DET (16; 5% instances), ADP (12; 4% instances), PUNCT (7; 2% instances), AUX (2; 1% instances), PRON (2; 1% instances), SCONJ (2; 1% instances), VERB (2; 1% instances), ADJ (1; 0% instances)