home edit page issue tracker

This page pertains to UD version 2.

Treebank Statistics: UD_Romanian-Nonstandard: Features: NumForm

This feature is language-specific. It occurs with 3 different values: Digit, Roman, Word.

5036 tokens (1%) have a non-empty value of NumForm. 531 types (2%) occur at least once with a non-empty value of NumForm. 360 lemmas (3%) occur at least once with a non-empty value of NumForm. The feature is used with 1 part-of-speech tags: NUM (5036; 1% instances).


5036 NUM tokens (97% of all NUM tokens) have a non-empty value of NumForm.

The most frequent other feature values with which NUM and NumForm co-occurred: Case=EMPTY (4384; 87%), Definite=EMPTY (4193; 83%), Gender=EMPTY (2952; 59%).

NUM tokens may have the following values of NumForm:

Paradigm treiRomanWord
Case=Acc,Nom|Definite=Def|Gender=Masc|Number=Sing|NumType=Ordtreile, treilea, triile
Case=Acc,Nom|Definite=Def|Gender=Fem|Number=Sing|NumType=Ordtrie, tree
Gender=Masc|Number=Sing|NumType=Ordtreile, treilea, triili
Gender=Masc|Number=Plur|NumType=Cardtrei, trii
Gender=Fem|Number=Sing|NumType=Ordtreia, trie, triia, tree, tria
Number=Plur|NumType=Cardtrei, trii, tri, trie

NumForm seems to be lexical feature of NUM. 97% lemmas (350) occur only with one value of NumForm.

Relations with Agreement in NumForm

The 10 most frequent relations where parent and child node agree in NumForm: NUM –[conj]–> NUM (339; 98%), NUM –[nummod]–> NUM (112; 82%), NUM –[nmod]–> NUM (18; 100%), NUM –[acl]–> NUM (3; 60%), NUM –[compound]–> NUM (3; 100%), NUM –[appos]–> NUM (1; 100%).